Challenges of the Housing Economy
Jones_ffirs.indd i
2/7/2012 10:25:25 AM
Challenges of the Housing Economy An Inter...
15 downloads
1669 Views
2MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Challenges of the Housing Economy
Jones_ffirs.indd i
2/7/2012 10:25:25 AM
Challenges of the Housing Economy An International Perspective
Edited by Colin Jones Professor of Estate Management Institute for Housing, Urban and Real Estate Research Heriot-Watt University
Michael White Professor of Real Estate Economics School of Architecture, Design and the Built Environment Nottingham Trent University
Neil Dunse Reader in Urban Studies Institute for Housing, Urban and Real Estate Research Heriot-Watt University
A John Wiley & Sons, Ltd., Publication
Jones_ffirs.indd iii
2/7/2012 10:25:25 AM
This edition first published 2012 © 2012 by John Wiley & Sons, Ltd Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Registered Office John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Offices 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of the author to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Challenges of the housing economy : an international perspective / [compiled by] Colin Jones, Michael White, Neil Dunse. p. cm. Conference proceedings. Includes bibliographical references and index. ISBN 978-0-470-67233-4 (hardcover : alk. paper) 1. Housing–Prices–Congresses. 2. Real estate business–Congresses. 3. Real property–Ownership–Congresses. 4. Global Financial Crisis, 2008–2009–Congresses. 5. Housing policy–Congresses. I. Jones, Colin. II. White, Michael. III. Dunse, Neil. HD7286.C43 2012 333.3–dc23 2011045311 A catalogue record for this book is available from the British Library. Set in 10/13pt TrumpMediaeval by SPi Publisher Services, Pondicherry, India 1
Jones_ffirs.indd iv
2012
2/7/2012 10:25:25 AM
Acknowledgements The chapters of this book are selected from papers presented at a symposium organised by the editors and generously funded by the Royal Institution of Chartered Surveyors (RICS) and held in September 2010 in Edinburgh. It brought together a small number of experts from around the world to examine recent housing market trends and to share international experiences and policies. There were a number of interrelated themes: the challenges faced by policy makers following the post-credit crunch world and the implications for households, construction, the housing market and the economy. Twelve papers have been selected for this volume from the original twentyone, to create a focused but comprehensive analysis of these themes and also to provide a worldwide perspective. We are grateful to the other participants for their comments on the original papers.
Jones_ffirs.indd v
2/7/2012 10:25:25 AM
The Royal Institution of Chartered Surveyors is the mark of property professionalism worldwide, promoting best practice, regulation and consumer protection for business and the community. It is the home of property related knowledge and is an impartial advisor to governments and global organisations. It is committed to the promotion of research in support of the efficient and effective operation of land and property markets worldwide.
Real Estate Issues Series Managing Editors Clare Eriksson John Henneberry K.W. Chau Elaine Worzala
Head of Research, Royal Institution of Chartered Surveyors Department of Town & Regional Planning, University of Sheffield Chair Professor, Department of Real Estate and Construction, The University of Hong Kong Director of The Richard H. Pennell Center for Real Estate Development, Clemson University
Real Estate Issues is an international book series presenting the latest thinking into how real estate markets operate. The books have a strong theoretical basis – providing the underpinning for the development of new ideas. The books are inclusive in nature, drawing both upon established techniques for real estate market analysis and on those from other academic disciplines as appropriate. The series embraces a comparative approach, allowing theory and practice to be put forward and tested for their applicability and relevance to the understanding of new situations. It does not seek to impose solutions, but rather provides a more effective means by which solutions can be found. It will not make any presumptions as to the importance of real estate markets but will uncover and present, through the clarity of the thinking, the real significance of the operation of real estate markets.
Jones_ffirs.indd vi
2/7/2012 10:25:26 AM
Books in the series Greenfields, Brownfields & Housing Development Adams & Watkins 9780632063871
Management of Privatised Housing: International Policies & Practice Gruis, Tsenkova & Nieboer 9781405181884
Planning, Public Policy & Property Markets Adams, Watkins & White 9781405124300
Development & Developers: Perspectives on Property Guy & Henneberry 9780632058426
Housing & Welfare in Southern Europe Allen, Barlow, Léal, Maloutas & Padovani 9781405103077 Markets & Institutions in Real Estate & Construction Ball 9781405110990 Building Cycles: Growth & Instability Barras 9781405130011 Neighbourhood Renewal & Housing Markets: Community Engagement in the US and UK Beider 9781405134101 Mortgage Markets Worldwide Ben-Shahar, Leung & Ong 9781405132107 The Cost of Land Use Decisions: Applying Transaction Cost Economics to Planning & Development Buitelaar 9781405151238 Urban Regeneration & Social Sustainability: Best Practice from European Cities Colantonio & Dixon 9781405194198 Urban Regeneration in Europe Couch, Fraser & Percy 9780632058419 Urban Sprawl in Europe: Landscapes, Land-Use Change & Policy Couch, Leontidou & Petschel-Held 9781405139175 Transforming Private Landlords Crook & Kemp 9781405184151 Real Estate & the New Economy: The Impact of Information and Communications Technology Dixon, McAllister, Marston & Snow 9781405117784 Economics & Land Use Planning Evans 9781405118613 Economics, Real Estate & the Supply of Land Evans 9781405118620
Jones_ffirs.indd vii
The Right to Buy: Analysis & Evaluation of a Housing Policy Jones & Murie 9781405131971 Housing Markets & Planning Policy Jones & Watkins 9781405175203 Office Markets & Public Policy Jones & White 9781405199766 Challenges of the Housing Economy Jones, White & Dunse 978047062334 Mass Appraisal Methods: An International Perspective for Property Valuers Kauko & d’Amato 9781405180979 Economics of the Mortgage Market: Perspectives on Household Decision Making Leece 9781405114615 Towers of Capital: Office Markets & International Financial Services Lizieri 9781405156721 Making Housing More Affordable: The Role of Intermediate Tenures Monk & Whitehead 9781405147149 Global Trends in Real Estate Finance Newell & Sieracki 9781405151283 Housing Economics & Public Policy O’Sullivan & Gibb 9780632064618 International Real Estate: An Institutional Approach Seabrooke, Kent & How 9781405103084 Urban Design in the Real Estate Development Process Tiesdell & Adams 9781405192194 Real Estate Finance in the New Economic World: Development of Deregulation and Internationalisation Tiwari & White 9781405158718 British Housebuilders: History & Analysis Wellings 9781405149181
2/7/2012 10:25:26 AM
Contents Contributors Glossary 1 Introduction: The Housing Economy and the Credit Crunch Colin Jones
1
The Irish example International historical housing market context Dynamics of the housing market Housing market and the economy Origins and impact of credit crunch Objectives and structure of the book Summary
2 3 6 9 11 18 23
2 US Housing Policy in the Era of Boom and Bust Harry W Richardson, Gordon F Mulligan and John L Carruthers
25
House prices from 1995 to present Housing reforms for the future Conclusion 3 Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas Craig A Depken II, Harris Hollans and Steve Swidler Data and definition of property flipping Anatomy of boom bust market dynamics Foreclosure activity after the bubble burst Concluding thoughts 4 Unemployment Risk, Homeownership and Housing Wealth: Lessons from Bubble Aftermath in Japan Yoko Moriizumi and Michio Naoi Housing bubble and burst in Japan Unemployment risk and homeownership Housing wealth and consumption Conclusion Appendices: Details of statistical analysis
Jones_ftoc.indd ix
xiii xvii
27 35 44
47 48 50 51 55
58 60 68 74 81 83
2/9/2012 11:33:16 AM
x
Contents
5 The Changing Nature of Household Demand and Housing Market Trends in China Edward C Y Yiu and Sherry Y S Xu Introduction to the housing market Population growth hypothesis Income growth hypothesis Monetary policy hypothesis Conclusions 6 Structural Sustainability of Homeownership in Australia Judith Yates
90 91 94 96 100 105
108
Australia’s housing system Future projections of homeownership sustainability Conclusions
110 119 126
7 Impacts on Wealth and Debt of Changes in the Danish Financial Framework Over a Housing Cycle Jens Lunde
128
Dynamics of recent Danish housing market cycles Changing structure of owner occupation Changing mortgage finance and indebtedness Financial stability of owner-occupied households Conclusions
130 133 135 140 151
8 Market Stability, Housing Finance and Homeownership in Germany Peter Westerheide Characteristics of the German housing market German housing price trends: an international comparison Housing demand and housing investment The contribution of the financing system The impact of the financial crisis on the German housing market High down-payment constraints and stability: contradicting aims? The role of savings behaviour Conclusion and outlook
Jones_ftoc.indd x
153 154 158 160 162 164 164 168
2/9/2012 11:33:16 AM
Contents
9
The Responsiveness of New Supply to House Prices: A Perspective from the Spanish Housing Market Paloma Taltavull de La Paz Housing supply trends Spanish housing market cycle pre and post the financial crisis Estimation of the supply elasticity of new houses Summary and conclusion Appendix: Details of statistical analysis
10 The UK Housing Market Cycle and the Role of Planning: The Policy Challenge Following the Financial Crisis Colin Jones Housing market context Anatomy of the housing market boom and bust Evolution of the planning framework in the UK Recent planning policies The planning policy consequences of the recession and beyond New planning agenda Conclusions 11 Developments in the Role of Social Housing in Europe Christine M E Whitehead Looking back: social housing and the welfare state in Europe Outcomes Future developments Conclusions 12 Delivering Affordable Housing in the UK Kenneth Gibb Historic social housing context in Scotland Background to the current Scottish affordability challenge Assessing the emerging models Concluding discussion
Jones_ftoc.indd xi
xi
170 171 174 179 184 186
195 196 197 205 207 212 213 214
216
217 226 231 233 235 238 239 245 252
2/9/2012 11:33:16 AM
xii
Contents
13 The Private Rented Sector As a Source of Affordable Housing Michael Ball The private rented sector within the housing system Potential policy developments Potential attraction of large investors Summary and conclusions
255 256 265 272 278
14 Conclusions: The Challenges Ahead Colin Jones
282
The credit crunch Commonalities and challenges The future Concluding remarks
282 284 290 292
References Index
Jones_ftoc.indd xii
294 312
2/9/2012 11:33:16 AM
Contributors Michael Ball is Professor of Urban and Property Economics in the School of Real Estate and Planning, Henley Business School, University of Reading. His books include Markets and Institutions in Real Estate and Construction (2006) and the co-author of the textbook, The Economics of Commercial Property Markets (1998). He jointly chairs the housing economics group of the European Network for Housing Research; and led the expert advisory panel on housing markets and planning for the UK government’s Communities and Local Government department from 2007–2010. He authors the annual European Housing Review for the Royal Institution of Chartered Surveyors. John L Carruthers is the Director of the Urban Sustainability Program at George Washington University in Washington, DC. He was formerly a Senior Economist in the US Department of Housing and Urban Development. Craig A Depken II received his PhD in economics from the University of Georgia and has held tenured positions at the University of Texas at Arlington and the University of North Carolina at Charlotte. His research areas include industrial organisation, sports economics, real estate, and public choice. He has served on the editorial boards of several journals and as co-editor of Contemporary Economic Policy. Neil Dunse is a Reader in Urban Studies in the Institute for Housing, Urban and Real Estate Research at Heriot-Watt University, Edinburgh. He received his PhD from the University of Paisley and was formerly a senior lecturer at the University of Aberdeen. His research interests are in housing and commercial real estate economics focusing on office submarkets, valuation accuracy, market adjustment processes over time, obsolescence and housing density and planning policy. Recent articles have appeared in Urban Studies, Real Estate Economics, Journal of European Real Estate Research and Journal of Property Research, among others. Kenneth Gibb is Professor of Housing Economics at the University of Glasgow. Kenneth is editor-in-chief of Urban Studies and a non-executive director of Sanctuary Housing Association. His principal areas of research concern the economic treatment of social housing, the application of behavioural economics to housing questions, the economics of housing policy and applied market analysis. He recently co-edited Housing Economics, a five volume reader published by SAGE (with Alex Marsh) and the SAGE Handbook of Housing (co-edited with David Clapham and William Clark).
Jones_flast.indd xiii
2/14/2012 10:18:00 PM
xiv
Contributors
Harris Hollans received his PhD in Real Estate from the University of Georgia. He is currently an Associate Professor in the Department of Finance at Auburn University. Prior to beginning his academic career he held various senior level positions in the field of commercial real estate valuation and consulting. His professional credentials include membership in the Appraisal Institute, with whom he holds the MAI designation. His research interests include real estate market cycles and the financial regulatory structure and its impact on real property markets. Colin Jones is Professor of Estate Management in the Institute for Housing, Urban and Real Estate Research at Heriot-Watt University, Edinburgh. He has held academic positions formerly at the universities of Glasgow, Manchester and Paisley. Colin’s research interests span housing, urban and commercial real estate economics. He has undertaken a range of projects for the UK and Scottish governments including a review of regulation and redress in the UK housing market (2009). His books include the Right to Buy (2006), Housing Markets and Planning Policy (2009) and (co-editor) Dimensions of the Sustainable City (2010). Jens Lunde is Associate Professor in the Department of Finance at Copenhagen Business School and has specialised in the area of housing economics and finance. Prior to this he started his carrier at the Danish Building Research Institute and later in the planning office at the Ministry of Housing. He has published widely in the different fields of housing economics, finance and policy. Jens has researched extensively on the taxation of the private ownership of housing, owner-occupiers’ capital structure and debt, tenure neutrality, the special Danish private co-operative tenure and mortgage loans. Yoko Moriizumi is a Professor at the Department of Economics in Kanagawa University. Her research interests include housing finance, portfolio choices by homeowners, and tenure choice and housing demand. Recently she has become particularly interested in the international comparison of mortgage choice. Gordon F Mulligan is Emeritus Professor in the Department of Geography and Regional Development at the University of Arizona, Tucson. Michio Naoi received his PhD from Keio University and is currently Assistant Professor at the Tokyo University of Marine Science and Technology. His research interests include tenure choice and housing demand; house price dynamics; and housing market regulation and institutions. Recent articles have appeared in Regional Science and Urban
Jones_flast.indd xiv
2/14/2012 10:18:00 PM
Contributors
xv
Economics, Journal of Real Estate Finance and Economics, and Journal of Property Research, among others. Harry W Richardson holds the James Irvine Chair of Urban and Regional Planning in the School of Policy, Planning and Development at the University of Southern California, Los Angeles. Paloma Taltavull de La Paz received her PhD from the University of Alicante. She is senior lecturer in applied economics and researcher in the International Economic Institute at the University of Alicante. Her research interests are housing market and real estate economics, as well as Spanish and European economics. Paloma’s latest research is focused on housing supply determinants, price formation and demand analysis in Spanish housing markets. Another area of interest is the development of European real estate education initiatives. Steve Swidler received his Ph.D. from Brown University and is currently the J. Stanley Mackin Professor of Finance at Auburn University. He is a recent Fulbright Scholar and former financial economist with the Office of the Comptroller of the Currency in Washington, D.C. His research interests include economic cycles in the housing market, managing house price risk, and more generally, empirical studies in derivative markets. Peter Westerheide studied economics at Witten/Herdecke University and received his doctoral degree from the University of Muenster. From 1999 to August 2011 he was researcher/senior researcher at the Centre for European Economic Research (ZEW) in Mannheim. His main fields of research addressed the areas of real estate markets analysis and real estate markets finance, private old age pension funding, saving behaviour, as well as corporate financing. Peter was also coordinator of the Leibniz-Network “Real Estate and Capital Markets” (ReCapNet). Since September 2011 he is chief economist of BASF SE, the world largest chemical company. Michael White received his PhD from the University of Aberdeen and is now Professor of Real Estate Economics at Nottingham Trent University. His research interests are in housing and commercial real estate economics focusing on market adjustment processes over time, spatial interactions across locations, and the separation of long and short run influences on market behaviour. His work has analysed issues relating to housing affordability for government bodies, investment performance for private institutions, and policy and market interaction for quasi government institutions. Christine M E Whitehead is Professor in Housing in the Department of Economics, London School of Economics and Senior Research Fellow at the
Jones_flast.indd xv
2/14/2012 10:18:00 PM
xvi
Contributors
Cambridge Centre for Housing and Planning Research, University of Cambridge. She has conducted extensive research on the housing market, with special reference to housing finance and subsidies, social housing provision and land use planning, as well as on urban, industrial policy and privatisation issues. Major themes in her recent research have included analysis of the relationship between planning and housing; housing needs assessments; the role and financing of social housing and privately rented housing in the UK and Europe. Sherry Y S Xu is a PhD student at the Department of Real Estate and Construction, the University of Hong Kong (HKU). Her research interests include macro analysis of real estate markets and real estate finance including REITs. Sherry’s research focus includes Hong Kong, China, USA and Australia. Judith Yates is currently an Honorary Associate in the School of Economics at the University of Sydney after a long career in academia. Her research has been in the fields of housing economics, finance and policy and, in the past few years, has focused on various aspects of housing affordability. She has served on numerous government advisory committees and boards and is currently a member of the government’s National Housing Supply Council. She holds a Doctor of Economic Science from the University of Amsterdam and a Bachelor of Economics (hons) from the Australian National University. Edward C Y Yiu received his PhD from the University of Hong Kong (HKU) and is now an Assistant Professor at the Department of Real Estate and Construction, the University of Hong Kong. His research interests include macro and micro-analyses of housing markets, real estate economics and finance, and urban economics. He has recently carried out a study on the effects of socio-economic, demographic and mobility changes on housing price gradient changes between Hong Kong and Macau, funded by the Research Grants Council of Hong Kong. He is currently extending the study to China housing markets.
Jones_flast.indd xvi
2/14/2012 10:18:00 PM
Glossary Basel III The Basel international agreements relate to common global standards of capital adequacy and liquidity rules for banks. These were first introduced in 1988. Basel III significantly increases the amount of equity capital that banks are required to have from 2013. Case-Shiller Index This is the gold standard of house prices in the USA. It provides a month-to-month measure of house prices that started in 1987. The S&P/Case-Shiller Home Price Indices are resale measures for the USA tracking changes in the value of real estate, both nationally and in 20 metropolitan regions. It is based on actual sales, and widely publicised in Standard & Poors outlets. For further information see: http:// www.standardandpoors.com/indices/sp-case-shiller-home-price-indices/ en/us/?indexId=spusa-cashpidff–p-us–– CDO A ‘collateralised debt obligation’ is a marketable investment security backed by a pool of bonds, loans (not necessarily mortgages) and other assets, that can be bought and sold on international capital markets. CDOs are therefore a general class of securities. Covered Bonds A covered bond is a specific form of CDO and enables the investor to have a ‘dual’ credit claim. Residential mortgage-backed securities are supported by the particular pool of mortgages. However, the covered bond holder is backed not just by the pool of (mortgage) assets but also by the issuer. This means that the bond holder has greater protection. Covered bonds have been used to a great extent in European countries such as Denmark, Spain and Germany. Fannie Mae Fannie Mae is a US government sponsored enterprise originally set up in 1938. It operates in the ‘secondary mortgage market’ to increase the funds available for mortgage lenders to issue loans to home buyers. It buys up and pools mortgages that are insured by the Federal Housing Administration (see below). It finances this by issuing mortgage-backed debt securities in the domestic and international capital markets. FHA The Federal Housing Administration is a US government agency created in 1934. It insures loans made by banks and other private lenders for home building and home buying.
Jones_flast.indd xvii
2/14/2012 10:18:00 PM
xviii
Glossary
Freddie Mac Freddie Mac is a US government sponsored enterprise established in 1970 to provide competition to Fannie Mae and operates in the same way. GHLC Government Housing Loan Corporation (of Japan) was established in 1950 to ensure stable housing finance to support new house building and to improve the standard of stock destroyed and devastated by Wold War II. It was an independent government agency that issued mortgages at low fixed rates of interest. GHLC had a large market share but from 2002 the government began phasing out subsidised mortgages. The agency was replaced by the Japan Housing Finance Agency in 2007 with a different primary role of securitising mortgages from private banks, although securitisation had begun under the GHLC. OECD The Organisation for Economic Cooperation and Development (OECD) comprises a group of 34 countries that includes all western countries. It was set up in 1961 to promote policies that improve economic and social well-being in the world.
Jones_flast.indd xviii
2/14/2012 10:18:00 PM
1 Introduction: The Housing Economy and the Credit Crunch Colin Jones
The world is still trying to adjust to the international banking crisis of 2007 that wrought recession and a dramatic downturn in housing markets across much of the globe, followed by a fiscal crisis in many countries that has yet to be fully resolved. The consequences for national housing markets have varied dependent on the specifics of their housing system, the point in the economic cycle and the exposure to the banking collapses. This book examines the different experiences of countries in Europe, Australasia and the USA and draws out the challenges to housing markets in the short and long term. The credit crunch and its aftermath has also emphasised the importance of the housing market to the economy. The relationship between the housing market and the economy has always stimulated debate but it has recently risen to greater prominence as special editions of Journal of Housing Economics (13.4) and the Oxford Review of Economic Policy (24.1) bear witness. In the latter volume, Muellbauer and Murphy (2008) offer a survey of the multiple interactions between the housing market and the economy but they leave many questions unanswered. The events at the end of the last decade have given this issue even more momentum as Gabriel et al. (2009) note, writing about the position in the USA, ‘The far-reaching economic and social consequences of the housing crisis require nothing less than the wholesale evaluation and redesign of housing policy, regulation, and the finance systems.’
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c01.indd 1
2/8/2012 8:28:53 PM
2
Challenges of the Housing Economy
This introduction begins by illustrating the scale of the impact by reference to the extreme example of Ireland. It then sets out a statistical overview of international housing market trends over the past four decades by examining price cycles to provide a context for the book. These trends demonstrate that the links between the macroeconomy and the housing market vary between countries (Meen, 2002a, Demary, 2009). The next section reviews the fundamental dynamics of the housing market and the importance of tenure institutional structures and market parameters in each country. The following section highlights the relationship between the housing market and the economy. The chapter then outlines the origins and impact of the recent financial crisis on house prices in individual countries. The contents of the remaining chapters are then described explaining which aspects of the housing economy and market cycles are highlighted in the different countries studied.
The Irish example The Republic of Ireland is an extreme example yet its experience is instructive as a magnification what happened in many countries. During a decade-long economic resurgence from the mid 1990s, recognised by the term ‘Celtic Tiger’, per capita incomes in Ireland rose dramatically and the population increased by 17%. With no property tax, income taxes and interest rates falling, and the liberalisation of mortgage finance making it readily available, a housing market boom was generated which saw prices roughly quadruple to their peak at the end of 2006 (ESRI, 2011). In the process mortgage debt per capita more than doubled supported by easier access to international finance for Irish banks and foreign banks entering the market (Norris and Winston, 2011). Following the credit crunch, house prices fell by 38% from the end of 2006 to 2010 (ESRI, 2011). The parallel problems in the commercial property market, together with the collapse of the housing market, contributed to the subsequent collapse of the banking system and a bailout by the Irish government. This in turn led to a fiscal crisis, and the Irish government needed to be sustained by a bailout of €85bn in November 2010 by the other European governments and the IMF. The internal political fallout from these events led to a successful vote of no confidence in the government and a general election that returned the opposition. Further afield it has contributed to uncertainties on the international financial markets about the future of the euro currency. Beyond these headlines the impacts on the Irish housing system were more subtle. Historically the homeownership level was high in Ireland reflecting both low house prices and that it was the normal tenure in rural
Jones_c01.indd 2
2/8/2012 8:28:53 PM
Introduction: The Housing Economy and the Credit Crunch
3
areas. The boom priced out low-income households, particularly in urban areas, who had to look to the private rented sector while there was an increased concentration of high housing debt among successful young firsttime buyers. The downturn has hit this latter group worst with negative equity but they also tend to be relatively affluent, while older home owners, who represent a broad range of incomes and largely own outright, have also suffered capital losses (Norris and Winston, 2011). The full consequences of the credit crunch for the housing market in Ireland, as in other countries, have yet to fully work their way out, and that continues to create uncertainty. The recession in the house market has focused even more public attention on house prices and their future prospects. Their prominence reflects the importance of the housing market to household decision making and to the economy. With owner occupation the dominant tenure around the world such household pressures will have been felt in most countries. The inevitable and unanswered questions for which many are searching is where are these trends taking us and when (and if) the upturn will arrive.
International historical housing market context In the UK, where monitoring the housing market has become a national obsession since the credit crunch, newspapers have periodically and enthusiastically reported negative forecasts, by prophets of doom, of dramatic falls in house prices. These conclusions are normally based on a return to a market price level associated with house-price-to-income ratios last seen just after the millennium. It is a simple analysis, and a message that dampens market expectations and plays on fears about the fragility of the housing market and household wealth bound up in a home. The theoretical basis of these reports is questionable as they consider only one linkage of the complex forces that shape the housing market and its dynamics. It does not draw on the experiences of the many other countries that have been similarly influenced by the global downturn, or at least only those that are consistent with their message. Similar national parochial analyses are being repeated in other countries. It is important to remember that homeownership is effectively treated distinctively in different cultures with implications for how housing markets work and trends in house prices. For example, in southern European countries households rarely move from their home, which is often purchased with the help of the family. Often several generations live together under the same roof. In contrast, in many Anglo-Saxon countries owner occupiers have sought to purchase at an early age, and move to adjust their housing requirements through the family life cycle so there is a much more active
Jones_c01.indd 3
2/8/2012 8:28:53 PM
4
Challenges of the Housing Economy
housing market. It is interesting to note that rising unaffordability, particularly in the latter stages of the last property boom, led, as the book chapters show, not only to a rise in the age of first home purchase in many countries but also a move towards a return to several generations living together. Furthermore, in the UK first time purchasers increasingly received financial support from parents to raise the deposit. A national survey by the Alliance and Leicester (2007) at the time found that parents were paying on average £21,314 to help their children get on the property ladder, One of the major motivations for this book is to take an international perspective on the short-term impact of the financial crisis on the housing market and review the reasons for differences between countries. To gain some initial understanding on the current housing market it is useful to begin by taking an overview of long-term trends worldwide by reference to OECD house price data (it excludes some major economies such as China). Some care should be taken in the comparison of this data as they are constructed in different ways and as a result they can be biased, for example towards certain segments of the market (Eiglsperger, 2010). This issue is also discussed for the particular case of Germany in chapter 7. In the analysis below our interest is just to establish a broad overview. A useful way to compare international housing market experiences is to focus on the different cycles of countries. Table 1.1 reports on major upturns and downturns in real prices since 1970 for selected countries. André (2010) defines major cycles in the housing market as real house price changes of at least 15% within some of the largest economies and ignores more modest upward and downward adjustments. Judged by this table, the UK housing market is the most volatile followed by Spain, Italy and Denmark. Although not reported in the table, Finland and Sweden show a similar volatility to Denmark albeit the cycles are different. Many countries experienced significant upturns in the housing market at the beginning of the 1970s. In the case of France the house price boom lasted most of the decade but for Italy, the UK and Japan the tripling of oil prices in 1973 and the recession that followed was mirrored in a downturn in house prices that lasted up to four years. The beginning of the 1980s saw a divergence of house price trends into two camps. In much of continental Europe real house prices fell substantially – Germany, France, the Netherlands, Spain and Italy (as well as Ireland, Sweden and Canada) – for the first half of the decade if not longer before showing dramatic growth in the latter part of the decade (Germany excepted). In contrast the early 1980s saw the beginning of substantial upturns in real prices in Finland, Denmark, the UK and the USA. These upswings (with the exceptions of Norway and Denmark) lasted most of the 1980s. This almost universal (in terms of OECD countries) house price boom of the (later) 1980s falls away around the end of the decade, and the beginning
Jones_c01.indd 4
2/8/2012 8:28:53 PM
Introduction: The Housing Economy and the Credit Crunch
5
Table 1.1 Real house price cycles in selected countries. Country
Upturns
Duration % (quarters) Change
Downturns
Duration % (quarters) Change
United States 1982Q4–1989Q4 1995Q1–2006Q4
28 47
+15.3 +60.5
1989Q4–1995Q1 2006Q4–2010Q4
21 16*
−5.5 −17.1
Japan
1970Q1–1973Q4 1977Q3–1991Q1
15 54
+56.9 +80.6
1973Q4–1977Q3 1991Q1–2010Q3
15 78*
−29.1 −44.2
France
1971Q2–1980Q4 1984Q4–1991Q2 1997Q1–2007Q4
42 23 44
+40.6 +33.0 +117.7
1980Q4–1984Q4 1991Q2–1997Q1 2007Q4–2009Q2
15 26 6
−17.6 −17.6 −9.5
Italy
1971Q1–1974Q4 1978Q1–1981Q2 1986Q4–1992Q2 1997Q3–2007Q4
15 13 22 41
+44.8 +41.9 +58.0 +59.2
1974Q4–1978Q1 1981Q2–1986Q4 1992Q2–1997Q3 2007Q4–2010Q3
14 22 25 11*
−9.0 −35.6 −28.3 −9.0
UK
1970Q1–1973Q3 1977Q3–1980Q3 1982Q1–1989Q3 1995Q4–2007Q4
14 12 38 48
+68.8 +27.7 +96.2 +157.8
1973Q3–1977Q3 1980Q3–1982Q1 1989Q3–1995Q4 2007Q4–2009Q2
16 6 25 6
−33.3 −12.1 −29.8 −14.7
Australia
1970Q1–1974Q1 1978Q4–1981Q4 1983Q1–1989Q2 1991Q2–2010Q2
16 12 25 76
+33.9 +15.0 +44.0 +139.5
1974Q1–1978Q4 1981Q4–1983Q1 1989Q2–1991Q2
19 5 8
−16.6 −10.8 −6.9
Denmark
1970Q4–1979Q2 1982Q3–1986Q2 1993Q2–2007Q1
34 15 55
+35.7 +57.9 +177.7
1979Q2–1982Q3 1986Q2–1993Q2 2007Q1–2010Q3
13 28 14
−34.8 −34.1 −20.7
Netherlands
1970Q1–1978Q2 1985Q3–2008Q4
33 89
+95.0 +217.6
1978Q2–1985Q3 2008Q4–2010Q4
29 8*
−49.6 −7.9
Spain
1972Q2–1974Q3 1976Q2–1978Q2 1986Q1–1991Q4 1996Q3–2007Q3
9 8 23 44
+31.1 +27.9 +135.1 +121.5
1974Q3–1976Q2 1978Q2–1986Q1 1991Q4–1996Q3 2007Q3–2010Q4
7 31 19 13*
−10.6 −32.3 −18.1 −18.5
* represents downturns which were continuing at end of 2010. Source: OECD quarterly house prices data.
of the 1990s saw a period of modest negative price adjustment for most countries. In the case of France, Finland, Italy, Spain, Switzerland and the UK substantial downturns lasted approximately six years. A new significant upturn emerged for most OECD countries in the middle of the decade, some earlier (Ireland 1992 and Scandinavian countries 1993) and others later (Canada 1998, Italy 1998, New Zealand 1998 and Switzerland 2000). For virtually all these countries the house price continued until a sudden end with the credit crunch. These long upward movements in house prices were not only the longest in recent times but also led to the doubling of real prices in some countries. Yet despite the longevity of the upswing there was considerable variance in the scale of the upturn. At one extreme the highest real increases were
Jones_c01.indd 5
2/8/2012 8:28:53 PM
6
Challenges of the Housing Economy
Ireland 302%, Norway 199%, Denmark 174% and the UK 173%. On the other hand, the US and Canadian real rises were only 56% and 72% while in Switzerland it was only 20% although it was not affected by the latest downturn. There were also mini-cycles within the long upswing with periods when prices surged and other periods when prices plateaued. There are a mass of micro-detail differences between countries of which the above commentary has only begun to scratch the surface. Nevertheless the review emphasises the important role and scale of cycles in the housing market. It also establishes a degree of commonality especially in the past 15 years of the cycles, reflecting the significance and impact of globalisation. Yet there is sufficient variation to demonstrate the importance of national factors. There are countries that stand out in different ways. Australia has not had a significant downturn in real house prices over the whole 40-year period, while the USA suffered its first substantive collapse of prices in living memory after the credit crunch. Japan has perhaps the most distinctive pattern of real house prices with long upswings and downswings. First, house prises rose by 57% in real terms between 1970 and 1974 in just 15 quarters. Then after a downward adjustment over the next three years of 31%, rising by 77% through 54 quarters, almost 14 years, from 1977 until 1991. This is followed by almost two decades of prices declining in total by over half in real terms over the period to the present day. Germany too suffered a long downward movement in real house prices from 1994 to 2008 which saw a real fall of 26.5% when most European cities were living through a strong housing market. The Netherlands in contrast has the reverse experience with real prices consistently rising over a 23-year period from 1985 to 2008 by 233%.
Dynamics of the housing market A key influence on these house price cycles is the business cycle of the individual country with peaks in the two cycles broadly coinciding (Aherne et al., 2005, André, 2010). This is not always true as the Netherlands illustrates. The long expansionary phase of house prices around the world culminating in the credit crunch also continued through the relatively shallow macroeconomic downturn in many countries around the millennium. This period of expansion is also significant in proportional terms with real prices approximately doubling that experienced in previous upturns in most countries. The exceptions are Finland and Spain in the 1980s when there were particular individual economic circumstances that contributed to dramatic growth (André, 2010). A rise in real house prices can normally be partly attributed to a period of (initially) relatively low interest rates, and be associated with ease of
Jones_c01.indd 6
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
7
mortgage finance and a growth in real incomes (Adams and Fuss, 2010). The increasing availability of mortgage finance often coincides with deregulation of banking systems and increased competition between banks. Significant deregulation of mortgage finance began in the early 1980s, and by 1990 most constraints in the industrialised world had been lifted, supporting a major stimulus to national housing markets in that decade (Ahearne et al., 2005). The latest upswing in prices was associated with the introduction of more flexible mortgage products with more relaxed loan-to-value ratios and longer terms available (Scanlon and Whitehead, 2004). Housing market trends can also be attributed to and be shaped by longterm factors which will vary nationally rather than just these short-term macroeconomic forces. The availability of mortgage finance varies across countries causing variations in response to interest rate changes (Kasparova and White, 2001). Demographic trends through the age structure of the population, the rate of household formation and migration can have an important influence on demand, as the Irish example illustrates (although this is also a function of long-term economic prospects). House price trends over time are influenced by the supply response which in turn is a function of the land-use planning constraints and the economics of the house building industry. Barker (2003, 2004) has highlighted the importance of this issue in the UK relative to other countries. Jones and Watkins (2009) chronicle how the weak supply response in the noughties contributed to house price increases in every region of UK. However, in other countries with weak planning system there is the potential for overbuilding with consequences for the investment attributes of housing. Different tenure structures in countries may influence the relationships between the housing market and the economy. Owner occupation and renting are substitutes, and the interaction between these sectors can impact on market changes as the following examples illustrate. House price trends influence the relative attractiveness of different tenures. A private rented sector comprising a high proportion of landlords with high debt gearing is potentially very susceptible to housing market downturns and likely to amplify price cycles. Long-term tenants are likely to respond differently to house price cycles from short-term transient tenants who are saving to get access to the owner occupied sector. The degree of low-income owner occupation may be an important factor through its impact on house price volatility and hence the attractiveness of homeownership as an investment (Westerheide, 2009). Tenure systems in different countries are partly a policy variable as the role of the state is crucial and bound up in the realms of political economy. They are linked directly to housing policy including any rent controls/ regulation, the scale and allocation of social housing and welfare payments. Kemeny (2005), for example, distinguishes between the dual rental markets
Jones_c01.indd 7
2/8/2012 8:28:54 PM
8
Challenges of the Housing Economy
Table 1.2 Country Greece Iceland Portugal Belgium Australia UK USA Finland France Sweden Denmark Netherlands Germany
Households by tenure in selected countries. Owner occupation 80 78 76 74 70 70 68 64 56 55 53 53 41
Social housing 0 2 7 7 5 20 3 17 17 21 19 35 6
Private rented 20 5 15 16 20 10 30 15 21 24 18 12 49
Other 0 16 2 3 5 0 0 4 6 0 9 0 5
Year 2001 2003 1999 1999 1999 2001/02 2002 2001 2002 1997 1999 1998 2001
Source: Scanlon and Whitehead (2004).
of Anglo-Saxon countries and the unitary rental markets of some European countries dependent directly on government policies and social values (Hulse et al., 2010). In fact the influence of housing finance systems and subsidies/taxation on national patterns of homeownership and renting is a complex issue. Wolswijk (2010) gives a summary of the diverse taxing of housing in the EU including taxes on imputed rent and capital gains, the deduction of mortgage interest against tax and VAT on new housing. Table 1.2 illustrates the tenure patterns of selected cities and demonstrates the scale of variation between countries. With the exception of Germany the majority tenure is owner occupation but there are differences in the split between the proportions of social housing and private rented sectors. Social housing is important in many northern European countries although its role is in decline through sales to sitting tenants, and these figures exaggerate its current position (Jones and Murie, 2006). In many countries such as the USA social housing has been seen simply as welfare housing and is therefore on a very limited scale (Harloe, 1995). These countries have a long-established private rented sector that represents a substantial proportion of the stock. Significant changes in the long-term social policy and financial parameters of the housing market can impact on private housing market decisions abruptly but usually take some time. The turning points in house price cycles are almost always associated with the fundamentals of the economy, namely interest rates and real incomes. Downturns in housing markets are historically linked to the overheating of the macroeconomy and growing inflation, followed by a tightening of monetary policy and the raising of interest rates to correct this phenomenon (Aherne et al., 2005).
Jones_c01.indd 8
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
9
Housing market and the economy The relationship between the housing market and the economy is more complex than this simple focus of the macroeconomy on house prices. Much interest has centred on the role of the housing market on consumption. Housing is not just any good as it accounts for a high percentage of income for many households. The Keynesian ‘absolute income hypothesis’ with its close link between current expenditure/savings and current income or earnings was challenged in the 1980s. This is illustrated by the experience of the UK during this period; with savings falling dramatically and consumer spending rising faster than average earnings it suggested that the Keynesian view was incomplete. The consensus explanation was in the role of permanent income and wealth: the dramatic rise of house prices in the UK (and in other countries) significantly increasing the wealth of households and reducing the need to save for, say, retirement. Case and Quigley (2008) examine this relationship in the USA and suggest that the relationship is not so straightforward. They find that the wealth effect works only when the housing market is booming but the reverse does not happen when house prices fall. The reason for the asymmetry may be because unless they are selling their home people do not perceive that their own house price will fall in the medium to long term. An alternative perspective on this relationship between consumption and the economy sees the growth in consumer spending arising from, or at best encouraged by, positive prospects for the economy, and this process is underpinned by house price rises that permit mortgage equity withdrawal (where the finance system permits) as a means of funding such expenditure more cheaply than unsecured borrowing (Aoki et al., 2001). Similarly when house prices are rising this is usually associated with greater turnover or housing transactions which leads to a rise in expenditure on complementary goods such as furniture and carpets. A further dimension to this relationship between consumption and the housing market is outlined by Aoki et al. (2001). They develop a model that simulates the impact of a reduction in interest rates using the financial accelerator model of Bernanke et al. (1999) and applied to housing, in which imperfections in credit markets amplify and propagate shocks in the macroeconomy. The mechanics of this dynamic is broadly as follows: a positive shock in economic activity causes a rise in housing demand, which leads to a rise in house prices and so an increase in home owners’ net worth and their ability to borrow. Changes to the (de)regulation of banks over time have influenced these mechanisms. Some commentators have argued that the financial liberalisation policies gradually endorsed by most countries since the 1970s has led to excessive liquidity thereby increasing household debt.
Jones_c01.indd 9
2/8/2012 8:28:54 PM
10
Challenges of the Housing Economy
Arestis and Karakitsos (2010) propose that monetary policies should incorporate the targeting of net wealth (and hence the housing market) of the personal sector rather than just price stability via interest rates. The housing market also impinges on the economy via new investment in the form of building or more precisely housing completions. The value of this output is an injection into the economy with an associated multiplier effect. Case and Quigley (2008) examine the effect via a downturn in housing completions in the USA. This is a decline in investment or a reduction in injections into the economy. They note two induced effects: first the impact on associated businesses such as mortgage brokers including second-hand sales (to buy new houses) and second induced expenditure on furniture etc. They apply a multiplier of 1.4 to account for these impacts. Investment can also be seen as a major dynamic of the business cycle via the (Keynesian) accelerator. Within this perspective new house building as a component of investment (representing the order of say 20% of investment but variable over time and between countries) is a major driver of the economy. However, the supply of new housing is in turn determined by the operation of the housing market and in particular viability is a function of house prices. Within this theoretical framework a housing ‘crisis’ through weakening housing prices and residential investment/construction could lead through to a subsequent recession. Leamer (2007) studies residential investment cycles and their importance for the economy in the US and concluded: ‘Eight of the ten recessions were preceded by sustained and substantial problems in housing and there was a minor problem in housing prior to the 2001 recession. The one clear exception was the 1953 recession, which commenced without problems from housing.’ (Leamer, 2007, p 164). By use of similar techniques Danish residential investments were shown to contribute to weakness in GDP growth one year before an actual recession.1 Notwithstanding the arguments about the potential leading role that the housing market can play in contributing to an economic downturn the evidence presented in Table 1.1 is that for most countries the fall in real house prices followed the world global crisis. The downturn since 2007 has a very different origin, effectively an exogenous event in the form of the subprime banking crisis creating a financial collapse with consequences for housing markets around the world. Even in countries already suffering real price falls the crisis probably magnified this market downturn. Similarly in countries with historically high real house prices prior to the credit crunch then this is likely to have had a detrimental impact on the subsequent performance of the economy for the reasons outlined above. The chapter now examines the long-term roots of this financial crisis and its specific short-term impacts on the housing market in different countries.
Jones_c01.indd 10
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
11
Origins and impact of credit crunch The credit crunch had a dramatic impact on banks and the supply of mortgage credit but the origins of the problem can be traced back to the deregulation of the banks in the 1980s. It stems from banks seeking to maximise their lending by borrowing. It is easiest to illustrate for mortgage banks in the UK (or building societies) but the story applies more generally with different timings and institutional arrangements. Mortgage banks originally grew by attracting savers with funds. Thirty years ago building societies in the UK would require only a small number of savers for each mortgage offered. As they sought to expand they needed to attract more savings but to do this the additional savers tended to save less per person and so more savers were needed per each mortgage granted. This was achieved through attracting savings customers through their branches and hence the funds are described as ‘retail’ (Jones, 1984). Eventually the additional growth from this source became exhausted as it was more and more expensive to attract marginal savers who only had limited funds. The building societies/mortgage banks turned to ‘wholesale’ funds, i.e. borrowing from institutions and large investors around the world. This was already well established in the USA through publicly sponsored agencies such as Fannie Mae (see glossary, and chapter 10 for a UK bank case study) by securitising (publicly insured) mortgages. In the lead-up to the credit crunch the balance of funds raised from retail and wholesale sources in the UK varied by mortgage lender, with some almost entirely dependent on the latter. With constraints on the degree to which building societies can access such funds those institutions that wished to use wholesale funding extensively had to convert to being a bank. The cheapest way of borrowing was undertaken by issuing bonds backed by the mortgages already issued which were bundled up together. In this securitisation process investors who bought the bonds were assured of a given rate of interest for a fixed number of years (with a set redemption date) supported by the income. These are known as ‘residential mortgage backed securities’. However, there is the risk of default by the original mortgagors. In the original American model all this risk was underwritten by the government. With deregulation the use of these bonds in the USA was extended beyond the inclusion of only publicly insured mortgages based on conservative lending criteria. The noughties saw a dramatic increase in the global issue of these bonds, roughly quadrupling between 2000 and 2006 (Crosby, 2008). The international extent of the use of residential mortgage backed securities including ‘covered bonds’ (see glossary) is demonstrated in Table 1.3. These bonds were increasingly backed by subprime mortgages, amounting to
Jones_c01.indd 11
2/8/2012 8:28:54 PM
12
Challenges of the Housing Economy
Table 1.3 International residential mortgage backed securities and covered bond balances outstanding in 2008Q3. Residential mortgage backed securities € bn Austria Denmark France Germany Ireland Italy Netherlands Spain UK USA Multinational Other
2.3 0.3 13.6 5.9 27.3 65.4 130.5 135.2 406.9 4574.0 2.9 58.2
Covered bonds € bn 4.1 335.8 63.6 206.5 13.6 — 15.7 267.0 82.0 12.9 — 156.3
Notes: Multinational bonds includes all deals in which assets originate from a variety of jurisdictions. Covered bonds refer only to those where the underlying asset is mortgages and relate to balances outstanding at the end of 2007. Source: Crosby (2008).
around one-third of such bonds issued in 2005 (Crosby, 2008), but the system collapsed when there was a ‘subprime’ mortgage crisis. Subprime mortgages are loans that are offered to house buyers who have poor credit histories and are offered on higher than usual interest rates to reflect the higher risk to the bank. They can be a bone fide way of offering people with past debt problems a chance to rebuild their credit risk and ultimately transfer after a number of years to a standard mortgage. Subprime mortgages also provide opportunities for households excluded by mainstream lending. In the UK, for example, many low-income households who took up the opportunity to purchase their council home under the government ‘Right to Buy’ scheme in the noughties bought with a subprime mortgage (Jones, 2009). The long rise in house prices from the mid 1990s encouraged a dramatic increase in subprime lending in many countries (see Jaffee, (2008) for USA) as households were attracted by the potential investment return and lenders forgot that price booms never last. Subprime mortgages are inevitably more prone to foreclosure. In 2007 these foreclosures rose dramatically in the USA and this coincided with a severe downturn in house prices (see below). It is probable that the subsequent rise in foreclosures was exacerbated by the fall in house prices (Jaffee, 2008), and the decline in values also undermined the fundamental values of the assets secured by these loans. However, there are also allegations that in the heat of the property boom mortgages were sold to people who would never have had the financial capacity to pay. On the face of it this
Jones_c01.indd 12
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
13
seems simply an American banking issue created by the deregulation of the secondary mortgage market at the beginning of the noughties. The problem mushroomed into a global financial crisis because the US banks with these subprime mortgages had also parcelled them up (just as in the UK above) with other mortgages to support the issue of bonds. Furthermore the scrutiny of the risk associated with these bonds was extremely lax and many received the rating of triple A from credit agencies. Banks from around the world had bought these bonds as part of a global financial capital market which encompasses equivalent products (not just linked to residential mortgages) in other countries. This interbank market of bonds allows banks to manipulate/control their liquidity and so minimise the cash they need to hold to be available to customers. Bonds can be sold quickly to improve liquidity/cash holdings. In this case the bonds were ultimately found to be supported by assets that were worthless and so the bonds in turn were worthless. While the subprime crisis was at its extreme in the USA, and arguably confined to that country, the impact on the international mortgage backed securities (bonds) market and their values was devastating. These securities no longer had credibility as assets as no one could be confident about the underlying assets. It was impossible to distinguish between bonds secured by good or bad assets because of the failure of the credit rating system. No one would buy them but the real problem for the banks was just beginning. The biggest problem occurred when these bonds were to be redeemed and refinanced. In normal circumstances new similar bonds would be issued to pay off the first set of creditors but this was no longer an option. Banks in western economies were also more generally unwilling to lend to other banks because of uncertainties about creditworthiness. This was partly in the light of the failure of the Lehman Brothers bank in September 2008 but also because of potential hidden bad debts. It was unclear who owned these worthless subprime assets because of the maze of the world capital markets, and as commercial and residential assets fell in value this led to the failure of property companies and more foreclosures. Some banks were facing this problem in late 2008 and, with insufficient funds, banks had to issue more shares to make up the shortfall in funds. With few takers, these shares have been bought by governments who have become the major shareholders, i.e. banks have been partially nationalised. This is referred to as the re-capitalisation of the banks as effectively taxpayers’ money has been used to rebuild the capital base of the banks to replace assets that have become almost worthless. This period of rebuilding by the banks (where necessary) has also been a period of retrenchment in terms of lending, generally reflected in less liberal lending criteria exacerbated by expected falls in house prices and economic uncertainties. The reappraisal of risks led typically to stricter loan-to-value
Jones_c01.indd 13
2/8/2012 8:28:54 PM
14
Challenges of the Housing Economy
Table 1.4 Falls in mortgage finance in selected countries between late 2007 and late 2008†. Country Ireland Iceland Denmark UK Sweden Portugal USA Finland Spain France New Zealand Australia
% fall in number of new mortgages
% fall in value of new mortgages
50 52 — 59 — — — — 29 9 22 24
59 51 49 46 44 43 42 40 38 30 28 19
†
Selected countries where price falls had occurred. Source: Scanlon et al. (2011).
ratio requirements and severe reductions in the availability of mortgage finance. Scanlon et al. (2011) report on these reductions from late 2007 to late 2008 for 12 countries that suffered price falls and their results are summarised in Table 1.4. The declines ranged from, at the extreme, Ireland with a 59% fall to only 19% in Australia where the house price fall was only a modest 9%. For most of these countries the fall in mortgage finance was of the order of 40% or more. The short-term impact of the credit crunch on the housing market is best seen through the prism of nominal house prices. Nominal prices are important for housing market dynamics in terms of influencing household decisions, and falls can create negative equity with the threat of foreclosure. Falls in nominal house prices can also be critical for the profitability and output of the house building industry. An overview of nominal house price trends since the global financial crisis for OECD countries is given in Table 1.5 and shows a common fall in prices in the four quarters to the end of 2008. Even where there is not an actual fall in house prices there is a slowdown in growth. The only exception to this pattern is Switzerland. The pattern of price change also identifies Ireland and the USA as leading this downturn; Japan too stands out as experiencing consistent annual falls in house prices over the whole of the five-year period, 2006–2010, albeit with a deterioration from 2008 on. A closer view of the response of national housing markets to the global financial shock is provided by Table 1.6 which details the precise turning points and price paths for selected countries. The market outcomes can be characterised as price waves that peak broadly at the same time but follow
Jones_c01.indd 14
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
15
Table 1.5 Annual change in nominal house prices 2006–2010 in OECD countries†. Country
2006
2007
2008
2009
2010
United States Japan Germany France Italy United Kingdom Canada Australia Belgium Denmark Finland Greece Ireland Korea Netherlands Norway New Zealand Spain Sweden Switzerland
4.7 −2.1 0.5 9.8 6.0 9.2 9.8 9.8 10.7 15.0 5.4 13.0 11.9 9.2 4.4 16.7 9.6 8.3 10.5 1.8
−0.3 −0.7 0.8 5.6 4.7 9.9 12.3 14.1 7.8 1.1 4.0 2.7 −6.8 5.3 4.3 7.7 7.7 3.9 11.3 2.0
−4.9 −2.6 −0.6 −3.1 −1.2 −8.7 −10.9 −4.1 3.2 −10.5 −3.9 0.8 −8.7 3.9 1.6 −6.9 −9.0 −3.9 −2.0 3.7
−4.3 −4.0 2.4 −4.6 −3.4 0.3 19.4 14.0 1.1 −5.1 7.9 −4.2 −18.6 0.7 −5.1 11.6 5.4 −6.5 7.0 6.5
−1.3 −2.7* 2.4 9.3 −1.1* 4.3 1.3 5.0 5.9 2.1 5.2 −6.0 −10.8 1.7 −1.0 6.6 −1.7 −3.2 5.3 4.2
†
Annual growth to the fourth quarter. *Up to third quarter of 2011. Source: OECD Quarterly House Price Data.
Table 1.6 Paths of change in nominal house prices 2006–2010 in selected OECD countries. Country United States France Italy United Kingdom Canada Australia Belgium Denmark Finland Greece Ireland Korea Netherlands Norway New Zealand Spain Sweden
First peak quarter
First trough quarter
Fall from peak to trough
Change from peak to 2010Q4
2007.2 2008.2 2008.1 2008.1 2007.4 2008.1 2008.3 2007.1 2008.2 2008.3 2006.4 2008.3 2008.3 2007.4 2007.3 2008.1 2008.1
2010.4* 2009.2 2010.3* 2009.2 2008.4 2008.4 2009.2 2009.2 2009.1 2010.4* 2010.4* 2009.2 2011.1* 2008.4 2009.1 2011.1* 2009.1
−10.8 −9.2 −6.5 −13.3 −10.9 −5.1 −2.1 −17.3 −5.4 −10.1 −38.1 −1.7 −6.9 −6.9 −11.2 −15.9 −2.3
−10.8 +0.6 −6.5 −4.7 7.7 12.7 6.6 −13.6 9.6 −10.1 −38.1 3.8 −6.9 14.0 −5.9 −15.9 10.5
*downturn continuing. Source: OECD Quarterly House Price Data.
Jones_c01.indd 15
2/8/2012 8:28:54 PM
16
Challenges of the Housing Economy
subsequent divergent ripples. From this table it can be seen that the housing markets in Ireland, Denmark and the USA had already turned down before the international crisis had begun because of internal domestic forces. Most other countries experience house price peaks from the fourth quarter of 2007 onwards. Some care needs to be taken in the interpretation of these turning points because the timing depends to some extent how, or more precisely when, the house price data is collected. As a result, alternative sources of data for the same country could come to different conclusions. For example, official statistics in the UK, collected at the formal completions of a house sale put the peak quarter at the beginning of 2008 whereas the Nationwide Building Society’s seasonally adjusted index identifies the fourth quarter of 2007 and the Halifax Bank’s equivalent index reports it earlier, as the third quarter. The differences partly reflect the fact that latter two are based on successful mortgage applications, and there is a lag to the formal completion of the transaction. Whichever index is used, the peak of the market in the UK, as in most other countries, is in September 2007. This retrenchment in national housing markets reflects the dramatic cuts in mortgage lending that followed the aftermath of the bail-out of the banking system across much of the western world. Indeed, it can be seen as the clear primary causal factor, because other main macroeconomic forces on the housing market, interest rates and economic growth, were not (yet) dampening demand. There was a concerted effort by governments around the world to reduce interest rates in the wake of the banking crisis. The US base rate rapidly fell from 5.25% from August 2007 to zero% in December the following year before stabilising at 0.25%. In the UK, bank base rates also began to be reduced in December 2007, falling eight times over 16 months from 5.75% to 0.5% in March 2009. They have continued at this level for over two years. The European Union base rate was similarly reduced in stages. At the time of writing, in July 2011, interest rates around the globe remain at historically low levels although the spring of this year saw a very modest upward movement in some countries. This reduction of interest rates has therefore supported housing markets since 2007 by substantially easing the cost of borrowing. Major economies have suffered recessions with knock-on effects for housing demand, but the brunt of the falls in real output (GDP) occurred from 2009 onwards, after the housing market had already turned down. Some countries such as Australia and China experienced only a slowdown in economic growth and this is reflected in the housing market trends discussed in subsequent chapters. However, it is interesting to note that the countries that led the international reverses in housing prices, such as Ireland and Denmark, generally experienced earlier recessions, reinforcing the arguments discussed earlier of the reverse link between housing market and the economy.
Jones_c01.indd 16
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
17
Perusal of Table 1.6 reveals that in many countries house prices began to recover approximately only a year – or in some cases less – after a sharp fall. But in some cases at least three years on from their respective market peak house prices continue to decline. These countries are Greece, Ireland, Italy, the Netherlands, Spain and the USA. Even where the immediate price downturn was stemmed prices had not necessarily recovered the ground lost by the beginning of 2011. Countries in this position are Denmark, New Zealand and the UK. Furthermore even after three years there are doubts about the sustainability of the market recovery in these countries, with prices dipping again or stuttering. The persistence of mortgage lending constraints combined with fiscal cutbacks still stifles demand. The legacy of the credit crunch three or four years on is a continuing and widespread international dampening impact on housing markets. In OECD countries the exceptions to this general pattern are Australia, Norway and Sweden. Where there is recovery it is interesting to note that in many countries the rise in house prices has not been accompanied by the normal dynamics of a housing market upturn. Housing transactions and new housing supply have been at a low ebb across Europe for example. These market conditions suggest fragility even in some countries where there has been a revival of house prices (Ball, 2011). An important short-term question about the significant house price falls and mortgage lending cutbacks (where they have occurred) is to what extent it matters beyond the impact on an individual household’s wealth. Some households who bought at the top of the market will have negative equity in their home although they may have been cushioned by the fall in interest rates. For households seeking to purchase their first home the impact has been bitter-sweet, for while the fall in real prices has ostensibly made owner occupation more affordable the tightening of credit availability outlined above has left many of them ultimately frustrated. Their prospects depend on the easing of lending criteria but this is not on the immediate horizon. There are also potentially severe ramifications for the house building industry with the flow of potential buyers turned off. The distributional impacts are only part of the policy dimension. There are two wider but opposite perspectives that are espoused: the performance of the housing market can be viewed as the pulse of the economy while the alternative sees rising house prices as unproductive financial investment. The latter view is also associated with a belief in unsustainable housing bubbles created by easy availability of credit, low interest rates, and even 125% mortgages. The recipe for economic growth and housing policy of these two conflicting views is diametrically opposed: supporting the housing market through perhaps help to first time purchasers, reductions in stamp duty, etc. versus allowing house prices to fall and increasing housing
Jones_c01.indd 17
2/8/2012 8:28:54 PM
18
Challenges of the Housing Economy
taxation to discourage speculative investment in bricks and mortar, and in turn to encourage capital toward productive investment. These alternatives focus on linkages between the business cycle and housing that revolve simply around house prices but the reality of the relationship between the housing market and the economy is more complex as noted above. This book refers to this as the housing economy and it is dependent on national housing market structures encompassing tenure, taxation, mortgage finance and the role of the planning system in influencing supply. The challenges of the housing economy are not just the shortterm practical ‘firefighting’ to address the immediate problems created by the credit crunch including minimising mortgage foreclosures and homelessness. There are also longer-term underlying forces building up in the housing economy that have been exposed or highlighted by the financial collapse and the subsequent fiscal crisis. These include the distribution of wealth between generations, the constraints on housing supply and the financing of social housing.
Objectives and structure of the book This book explores these challenges by taking an international perspective that draws on chapters from a range of countries with different experiences of the housing market over recent decades. It attempts to draw on the broad spectrum of international circumstances with individual chapters emphasising particular issues to a country. There are chapters on rising housing markets that have been relatively immune to the fallout from the credit crunch, namely Australia and China. There are also chapters on Germany and Japan that have very distinctive house price trends, and did not participate in the almost global house price boom, reflecting institutional differences and macroeconomic performance. The countries with the most volatile housing markets identified above, namely Denmark, Spain and the UK are considered in separate chapters. The USA is discussed in two chapters which draw out the implications of the aftermath of the housing boom and the pain of local experiences beneath national statistics that hide substantial spatial variation. Finally there are three chapters that centre on the role of rented housing and the role of the state. The book begins with the two chapters on the USA. In the first, US housing policy in the era of boom and bust, Richardson et al. describe and explain the geography of the major US housing market trends of the past two decades. The chapter argues that the boom was unsustainable and emanated from federal housing policies over three administrations through the relaxation of mortgage deposit constraints to promote homeownership. Combined with long-term public policy that offered federal mortgage
Jones_c01.indd 18
2/8/2012 8:28:54 PM
Introduction: The Housing Economy and the Credit Crunch
19
insurance and tax relief on mortgage interest payments, this was a toxic policy mix. The authors then analyse a range of housing policy reforms that might have avoided, or at least mitigated, the bust that happened, and what might ensure that these problems will not recur. In the following chapter, Housing bubbles and the foreclosures that follow: the case of Las Vegas, Depken et al. undertake a case study of Las Vegas that illustrates an extreme local outcome of speculative investment within the recent housing market boom and bust in the USA. It also introduces to many readers a new term, flip, as any property that sells within two years. The impact of the extensive occurrence of this short-term investment strategy in Las Vegas is shown to challenge the normal stickiness of house prices in a downturn as sales fall away rather than prices. Rapidly falling prices in Las Vegas have arisen from flipping activity. The chapter demonstrates how, as flippers could not find buyers in the downturn for houses bought near the peak, this led to mortgage defaults. Foreclosure activity then picks up and becomes the majority of transactions, so prices fall dramatically. In fact, the authors argue that such speculation was not forever sustainable as eventually housing stock growth must mirror population growth. In the meantime, there are many thousands of homes in foreclosure that lenders have not sold, with a consequent shadow on the future of the local market. The links between the labour market and homeownership and the implications for housing wealth are the focus of the next chapter, by Moriizumi and Naoi, drawing on the experience of Japan. Owner occupation is the majority tenure (61%), private rented housing accounts for (27%) and there is only very limited social housing (6%). The chapter, entitled Unemployment risk, homeownership and housing wealth: lessons from the bubble aftermath in Japan, examines the consequences of the housing boom of the late 1980s, and subsequent collapse in the early 1990s that was followed by a prolonged recession. The chapter considers the impacts of this housing boom and bust on households with different housing tenure status in Japan, and the consequences for the timing of homeownership by firsttime buyers. It outlines the housing price fluctuations and deteriorating labour market conditions in the long post-bubble period. The continuous decline in house prices after the bubble burst is shown to bring reverse wealth effects on consumption expenditure by homeowners. The chapter demonstrates how typical young first-time buyers suffered from rising housing prices in the bubble era, and from the higher unemployment rate in the post-bubble period. Positive effects of declining house prices in the 1990s on affordability were outweighed by the increased unemployment risk in the same period, resulting in a virtually stable homeownership rate among young households. Housing wealth effects, they argue, are substantially larger for households with losses in housing wealth/price falls than those with gains.
Jones_c01.indd 19
2/8/2012 8:28:54 PM
20
Challenges of the Housing Economy
China has seen dramatic rising housing prices over the past decade and did not suffer a downturn in the economy and housing prices after the financial crisis of 2007/08. Yiu and Xu note, in their chapter, The changing nature of household demand and housing market trends in China, that the boom is generally perceived to be driven by fundamental socio-economicdemographic demand factors, such as a substantial increase in disposable incomes in recent decades and the growing up of ‘baby-boomers’ from the 1960s, etc. However, this chapter puts forward evidence to refute this emphasis on these forces, noting the apparent existence of an oversupply yet there is a house-price-to-income ratio in Shanghai of 20. Instead, they emphasise regional differences, monetary policy and the operation of capital markets in China as the root of the house price boom. In particular with very limited investment vehicles in China the housing market has become almost a pure investment medium or even a speculative commodity market for inflation hedging. Australia is another country that suffered only the slightest impacts from the international financial crisis. Yates in her chapter, The structural sustainability of homeownership in Australia, draws out the long-term consequences of house prices rising faster than earnings on the structural sustainability of homeownership. She shows how it has added to barriers for aspiring first-time house purchasers in Australia who are unable to compete with increased demand from established owner-occupiers and investors, so that access to homeownership for young households moved higher up the income scale. There are therefore limited housing options for low- to middleincome households, and reduced access to the tax advantages of homeownership adds to wealth inequalities and limits asset accumulation that protects housing costs in old age. In fact, while there has been a plateau in the level of homeownership in Australia since the early 1960s Yates demonstrates that owner occupation rates in young households have been falling since 1976. The final part of the chapter projects forward current trends and the underlying forces and suggests that the existing level of homeownership is unsustainable. This raises significant issues about intergenerational equity and long-term government policies toward the support of the elderly. The intergenerational theme is also taken up by Lunde in his chapter, Impacts on wealth and debt of changes in the Danish financial framework over a housing cycle, for a country which has the highest level of household debt in the world. He dissects how different age cohorts fared over stages in the recent housing cycles. Lunde notes that there has been a fall in the owner occupation rate among the younger generation. The analysis shows that young owner occupiers who succeeded in buying are the most heavily in debt, and that they gained least from the long rise in house prices. Following the credit crunch they are also shown to be the most vulnerable to any future market instability. Overall Lunde finds that the long housing
Jones_c01.indd 20
2/8/2012 8:28:55 PM
Introduction: The Housing Economy and the Credit Crunch
21
market upturn from the mid 1990s benefitted the elderly at the expense of the young, creating barriers to owner occupation and financial strains on many who did succeed in buying a home. The intergenerational issues have been highlighted in many western countries by the increased deposits required by banks in the post credit crunch era resulting in greater barriers to homeownership for young people. The experience of Germany is instructive here as this is the long-term model of housing finance there. The next chapter, Market stability, housing finance, and homeownership in Germany, by Westerheide speculates whether the remarkable stability in the housing market in that country can be attributed to the operation of this housing finance model or the low income growth. The theoretical literature notes that the impact of down payment constraints on volatility is ambiguous as it partly depends on the potential indirect effects on saving levels. However, Westerheide finds that the impact on saving in Germany is minimal and that the answer partly lies in the availability of comparable houses to own and rent and these will vary across countries. The scale of the exogenous shock caused by the credit crunch is demonstrated by two chapters that take a supply perspective. Taltavull in her chapter, The responsiveness of new supply to house prices: a perspective from the Spanish housing market, shows how housing construction activity in Spain expanded from the late 1990s. The strong new supply growth was the development sector’s reaction to market impulses stemming from a rising number of new households due to the young structure of the population in Spain, the attraction of labour immigrants (about 4 million in seven years), a growth of retired/semi-retired buyers from other European countries that were very focused in a few regions, and very strong internal movements of the domestic population in those areas with rising economic activity. All these flows coincided with exceptional good economic conditions (job creation, low interest rates, low relative housing prices and an expansive flow of liquidity within the financial system encompassing both Spanish and European banks). These trends coincided with a relaxation of the planning system. Housing construction reacted positively because of the compliant planning system ensuring land availability and the immigration of construction workers. She demonstrates an increasing supply elasticity during this period, although it begins to reduce in the noughties. The high supply elasticity explains the strong contraction in construction activity that occurred after 2008 when the liquidity flows towards the financial markets stopped and Spanish banks and saving banks reduced the amount of finance to developers, first, and then to households. The aftermath of the credit crunch has been both a dramatic reduction in housing starts (reaching an historical minimum) and the level of housing market transactions. Nevertheless the housing market still has a high level of affordability and prices have not fallen substantially.
Jones_c01.indd 21
2/8/2012 8:28:55 PM
22
Challenges of the Housing Economy
The post credit crunch story is much the same in the UK although there were differences in terms of the supply response to the housing price boom of the noughties. In the next chapter, The UK housing market cycle and the role of planning: the policy challenge following the financial crisis, Jones quotes a government discussion paper in 2007 that acknowledges that for a generation the supply of new housing has not kept pace with rising demand. But the number of houses built in 2010 was the lowest since 1924. House prices from their peak around the end of 2007 showed an initial relatively quick recovery before falling away moderately. They have now broadly stabilised and are back at their 2006 levels after reaching a trough in the spring of 2009. The major constraint on recovery has been the mortgage famine leaving transactions at a low ebb. Jones argues that this initial relatively quick upturn compared with the previous recession was partly down to a different policy response (constraints on foreclosures and stamp duty holiday) but also down to the long-term housing supply constraints that are a function of planning policies. These policies have also contributed to higher density development and an emphasis on the construction of flats. The overhang from these policies, Jones argues, will shape and constrain a housing market recovery. The impact of the credit crunch in terms of the decline in private house building together with the long-term barriers to homeownership leads to a major policy challenge of the provision of affordable housing particularly for low-income households. The traditional solution for many countries has been social housing. Whitehead charts Developments in the role of social housing in Europe and the long term move away from housing as a core element in the welfare state, and the shifting of subsidies away from bricks and mortar towards income-related subsidies. Over time there has been increased private sector involvement (and rent increases) with the development of sophisticated financial instruments to support this process and the transfer of ownership to individuals and private and non-profit landlords. The most extreme changes have been in ex-socialist countries. Since the credit crunch, social housing investment has been seen as an important part of stimulus packages: examples include in England ‘Kickstart’ – additional capital funding for ‘shovel ready’ development; in France additional funding made available to increase social sector output; and in Norway – counter-cyclical funding policies to assist affordable housing. Whitehead notes that the longer-term position is much less clear with cutbacks in public expenditure Europe wide. At the same time there is likely to be an increasing demand for rented housing as the economies stutter forward over the next few years and governments will have to continue to be involved. In his chapter, Funding affordable housing in the UK, Gibb illustrates the options for funding models in the UK, given the government’s commitment to reducing public expenditure substantially. He notes how value for money
Jones_c01.indd 22
2/8/2012 8:28:55 PM
Introduction: The Housing Economy and the Credit Crunch
23
is now a central criterion in public policy to ensure that public funding goes further. This will undoubtedly be an international theme. He draws on the Scottish government’s approach that is now comparing different funding models that explicitly assess the public cost of each additional unit (standardised for years of life) and the relative benefits, whereas before these choices may have been determined more by political influence. As a result, the public sector is promoting the part funding of ‘intermediate renting’ at rent levels above existing social housing. The retreat of the role of the state in many countries leaves a question about whether the private rented sector can fill the gap. Ball in his chapter, The private rented sector as a source of affordable accommodation, reviews the role of the private rented sector in the UK. He finds that while there has been a substantial growth in private individual landlords over the past decade it is unlikely to see a further significant short- or long-term expansion because of the ageing of the population. Nevertheless, young households rationed out of homeownership because of the current restrictive loan-tovalue ratios applied by banks will continue to look to this sector and this would lead to more crowding within existing housing. Ball considers whether the private rented sector could expand its role to be a permanent tenure for some households by providing greater security of tenure for tenants or introducing ‘second generation’ rent controls. He argues that these policies are unlikely to improve – and will probably worsen – the housing situations of low-income households. Drawing on international experience Ball also does not foresee any large scale institutional investment in this sector because of the weak economies of scale of private landlordism. He concludes that is impossible to compensate for general supply-side-induced housing shortages through adjustments in the operation of specific tenures. The final chapter will reprise the principal themes presented in the book and compare and contrast the key findings from the contributions. It will draw conclusions about what the contributions tell us about the impact of the financial crisis on the housing economy and long-term housing market issues, drawing out international similarities that transcend institutional/ cultural differences between countries as well as explaining significant variations.
Summary The chapter has developed a framework to examine the short-term links between the housing market and the macroeconomy together with more long-term underlying influences that shape a nation’s housing market. It has also established the international context for the analyses presented,
Jones_c01.indd 23
2/8/2012 8:28:55 PM
24
Challenges of the Housing Economy
both highlighting the exceptionally long house price boom, beginning in the mid 1990s, that was shared by many countries, but also pointing to major exceptions to this cycle such as Germany and Japan. It then developed an overview of the lead-up to and the impact of the credit crunch on housing markets around the world. It noted the development of mortgage backed securitised bonds designed to increase bank lending and profitability. The selling and buying of these bonds and other sophisticated instruments on international capital markets with the aim of improving banking liquidity ultimately almost brought the global financial system to its knees. In its wake there were major bank casualties and the consequent bailouts by national governments wrought fiscal crises and in some cases rejection at the ballot box in subsequent elections. While the importance of the role of credit in the housing market has been dramatically emphasised by the financial crisis the role of interest rates and household incomes remain crucial short-term underlying influences. There are also a wide range of long-term factors on national housing markets including demographic trends and what can be described as ‘institutional’ contextual determinants such as tenure, planning and taxation. This chapter has demonstrated that many housing markets were at different points in the cycles when the global crisis occurred, and that these institutional factors are likely to be a strong influence on the outcomes and policy response. The chapters partly explore these short-term consequences of the financial crisis but contributions also reveal common long-term global trends in housing markets.
Note 1. I am grateful to Jens Lunde for these points.
Jones_c01.indd 24
2/8/2012 8:28:55 PM
2 US Housing Policy in the Era of Boom and Bust Harry W Richardson, Gordon F Mulligan and John L Carruthers
There are now many wide-ranging plans being hatched in the Obama Administration to prevent a recurrence of the recent housing crisis. These include cutbacks on mortgage subsidies and a proposed shift in focus to affordable rentals, but these plans have not yet been discussed in much detail. Other proposed changes are likely to involve more tightening of lending practices by the banks backed up by the government to save bailouts. The impact of these efforts on the revival of the housing market inevitably remains unclear at the moment, but they are unlikely to be very favourable. However, they should lead to more stability in both the housing and financial markets. The deadline for implementation of new proposals is the end of 2011. The focus is intended to be on ‘sustainable homeownership’, whatever that may mean, rather than on promoting homeownership for its own sake. The US housing crisis of recent years reflects a toxic mix of aggressive lending, pie-in-the-sky borrowing and weak financial regulation. US society has been built upon the dream of homeownership which peaked at 68 per cent of households in 2007, but has fallen a little in the past few years. What has happened in recent years is a deviation from this system in a somewhat reckless attempt to boost national homeownership. The stimulus to growth goes back to the concept of Federal government mortgage insurance in the 1930s, subsequently reinforced by the establishment of Fannie Mae (Federal National Mortgage Association) and Freddie Mac (Federal Home Loan
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c02.indd 25
2/8/2012 9:10:01 PM
26
Challenges of the Housing Economy
Mortgage Corporation) (see glossary). Until recently, most mortgages were 30-year loans, with some principal payments each month (before the 1930s five-year rollover loans with interest-only payments were the more common type). The standard loan system goes back to the New Deal in the 1930s when the Home Owners’ Loan Corporation introduced the 30-year amortised loan system. The system worked well for as long as homeowners could meet their payments, a precondition being keeping their jobs. The introduction of innovative but risky lending practices combined with growing unemployment risks disturbed any market equilibrium. Although the 30-year mortgage with a fixed interest rate and a 20 per cent down payment became standard, in recent years this standard was dramatically relaxed with a variety of mortgage instruments with adjustable mortgage rates and lower (even zero) down payments in an attempt, especially during the Clinton Administration (1992–2000), to expand homeownership to lowincome households. Generous mortgage interest tax deductions (up to $1 million for one or two houses) and unlimited property tax deductions have been in place for decades, and no capital gains taxes have to be paid on primary residence sales (up to $500,000 per couple). This stimulated excessive housing consumption reinforced by relatively cheap home equity loans while house prices were rising. This in turn created strong ‘wealth effects’ in which residential equity was transferred into many types of consumption. During the boom, mortgage rates were low (hovering around 5 per cent); back in the late 1970s they had been in the mid teens. Everyone forgot that housing was a cyclical industry, however. Adjustable interest rates began to climb, often rapidly, and the economy began to teeter. It became more difficult for borrowers to meet payments, especially those who had purchased second properties for quick profit turnover (the so-called ‘flipping’, typically defined as a 180-day turnover – see chapter 3). Mortgages had been bundled for sale to investors, but these bundles became problematic as more and more mortgage payments were unpaid. House prices began to tumble, the construction industry declined as did general consumption expenditures. Unemployment rose into the double digits nationally (with regional variations), and even in mid 2011, after a very modest recovery, it remains around 9 per cent. Foreclosures have become rampant with almost 832,000 sales in 2010 (at an average discounted price of 28 per cent), and 8 per cent of these were resold within 180 days. There has been a recent decline, but not because of improvement in the housing market but rather because of delays in processing and the high supply inventory. Nationally, as of April 2011 9 per cent of mortgages had been delinquent for more than 90 days, and the proportions in some states were much higher (20 per cent in Florida, 17 per cent in Nevada and 10 per cent in California and Arizona; data from Realty Trac).
Jones_c02.indd 26
2/8/2012 9:10:01 PM
US Housing Policy in the Era of Boom and Bust
27
These trends and policies are the backdrop to this chapter, which has two main sections. The first examines in detail how the geography of the ups (approximately 1995–2007) and the downs (from 2007 to the present) in house prices varied in cities across the country, and the reasons for these differences. The second section looks at the potential menu of reforms in housing policies and practices that may prevent a recurrence.
House prices from 1995 to present National Association of Realtors vs Case-Shiller data As a backcloth to the discussion that follows, Carruthers and Mulligan (forthcoming) outlined an argument that might sustain a differential house price revival. The cities that will do best as house prices begin to recover are those that benefit from a combination of social and natural amenities. Metropolitan areas can generate social amenities that improve the quality of life via public investment and civic leadership. Strengthening natural amenities is a much tougher proposition, although there are feasible actions such as preventing environmental deterioration at national parks and other recreational areas. We have pursued this line of analysis in examining house price increases in a wider array of cities over the period 1995–2007. An alternative to the Case-Shiller house price index (see glossary), one with much more geographic detail is provided by the National Association of Realtors (NAR, 2011). Instead of focusing on just 20 places as the Case-Shiller index does, NAR traces the median price for existing single-family houses across many hundreds of metropolitan areas. However, this is a quarterly and not a monthly index and the numbers are not seasonally adjusted. Moreover, a certain amount of accuracy is sacrificed because the alternative index is not based on repeat sales. Nevertheless, study of the NAR data sheds light on nationwide trends that cannot be addressed with the better known data. As pointed out, the Case-Shiller index only measures house prices in 20 of the nation’s largest and most visible cities. However, across the many hundreds of other metropolitan areas – those places having populations exceeding 50,000 – the pattern of rising house prices during recent decades was much less dramatic. To illustrate, we consider the 50 quarters following 1995Q1, where the initial house prices in 370 cities have all been set at a base value of 100. While places like Miami (247 per cent), Los Angeles (240 per cent) and Phoenix (193 per cent), all included in the Case-Shiller index, exhibited some of the fastest rising prices during those 12-plus years, other relatively large places like Milwaukee (93 per cent), St Louis (91 per cent) and Louisville (69 per cent) exhibited much more modest growth in their median house prices. And some smaller cities like Abilene, Texas (67 per cent),
Jones_c02.indd 27
2/8/2012 9:10:01 PM
28
Challenges of the Housing Economy
Appleton, Wisconsin (62 per cent) and Decatur, Illinois (55 per cent) showed even slower rises in their house prices. It is arguable that the Case-Shiller index with its 31 per cent decline in its 20 cities somewhat exaggerates the housing price crisis, given the less than 25 per cent in the much broader database from the National Association of Realtors. However, this is easily explained by the loading of very large cities in the Case-Shiller index with their generally faster price increases during the boom phase and the evening out of price changes a little across the country because of regional differentials in the extent of the housing crisis. The upcoming analysis focuses on two aspects of the rapid house-price shifts experienced by these metropolitan areas: the amount of the price rise and the pattern of the price rise. Initial concern is devoted to the overall magnitude of the price rise, which is simply house price in 2007Q3 minus house price in 1995Q1. This measure is the usual concern of house-price studies. Then a short analysis focuses on the longitudinal composition of the metropolitan house price increases, looking at the entire distribution of the various quarter-to-quarter price changes in each place. These changes are also used to cluster together cities into homogeneous groups that share the same pattern of house-price increases.
Amenities and other influences on house prices Three metropolitan attributes addressed in the Places Rated Almanac (Savageau, 2007) have been chosen to shed light on the issue of house price change. These attributes capture many of the human-created externalities and environmental advantages that differentiate large cities in postindustrial society. First, social ambience is considered. This factor relates to such things as the quality and quantity of restaurants, the availability of book stores and the incidence of national landmarks. These are goods and services widely demanded in the contemporary consumer city. Second, education is addressed. This attribute is measured by assessing the numbers of children in private and public schools, turnover rates in public libraries and the incidence of colleges or universities. These are all features of an affluent society that places a premium on widening and deepening human capital. Thirdly, natural amenities are examined. This factor is restricted here to climate, where mildness, brightness and hazard-free days are all preferred. Natural amenities are demanded by a mobile and educated workforce. Finally, using the REIS data (Bureau of Economic Analysis, 2011), metropolitan areas are assessed in terms of their per capita incomes in 2005. On all four attributes cities are first rank ordered and then these ranks are standardised to a percentile distribution having a lowest score of 0, a highest score of 100, and a mean score of 50. As Table 2.1 indicates three of these four factors are highly correlated, where natural amenities are the anomaly.
Jones_c02.indd 28
2/8/2012 9:10:01 PM
US Housing Policy in the Era of Boom and Bust
29
Table 2.1 Correlations among the (ranked) variables/characteristics of US cities.
Price rise Social Human Natural Income
Price rise
Social
Human
1.00 0.00 0.46 0.00 0.00
0.18 1.00 0.00 0.55 0.00
0.04 0.73 1.00 0.20 0.00
Natural 0.60 0.03 −0.07 1.00 0.02
Income 0.31 0.58 0.44 0.12 1.00
Note: Upper triangle figures are correlation coefficients; lower triangle figures (italicised) indicate significance levels.
Table 2.2 Mean percentile scores for the two groups of US Cities. Variable House price rise Social amenities Human capital Natural amenities Per capita income
Case-Shiller cities
Other cities
All cities
71.9 88.7 75.0 65.0 86.2
48.9 47.9 48.7 49.3 48.1
50.0 50.0 50.0 50.0 50.0
Note: There are 20 Case-Shiller cities and 350 other cities.
Table 2.2 indicates that the 20 Case-Shiller cities are very different from the 350 others included in the NAR data set. The Case-Shiller cities not only experienced higher house-price increases than the others during the 50 quarters of the study period but those highly visible places ranked much higher on all four metropolitan attributes. The very largest Case-Shiller cities like New York, Los Angeles, and Chicago rated especially high on per capita income and on social amenities (for an even broader view of these factors see Florida (2002), Glaeser (2011) and Bolton and Westlund (2003) ).
Magnitude of the house price changes The intent here is not to develop a causal model but simply to indicate which broad urban features seem most associated with the nation’s recent run-up in house prices. For Table 2.3 the 370 cities were divided into 10 groups of identical size where these groups were ranked from the fastestgrowing (category 1) house prices to the slowest-growing (category 10) house prices. In every growth category the mean value for each metropolitan attribute was next calculated and these are shown in that table. Moreover, it is noted if any cities in that category are included in the Case-Shiller list. So, to clarify, in the 37 cities with the fastest-rising prices the average score on natural amenities was an impressive 82.8 (where the overall mean is 50) while in the 37 cities with the slowest-rising prices the average score on
Jones_c02.indd 29
2/8/2012 9:10:02 PM
30
Challenges of the Housing Economy
Table 2.3 Mean levels of urban attributes classified by house price increases. Category
#Ca-Sh
Price
Social
Human
Natural
Income
1 Fastest 2 3 4 5 6 7 8 9 10 Slowest
8 4 1 2 1 1 0 1 0 2
95.0 85.0 75.0 65.0 55.0 45.0 35.0 25.0 15.0 5.0
57.8 55.8 59.3 53.6 44.8 50.7 47.9 44.7 44.8 42.0
52.7 49.7 54.0 46.7 54.4 49.9 47.5 49.1 49.4 48.0
82.8 73.0 69.0 57.3 41.3 45.9 35.7 34.1 31.0 31.1
72.8 60.3 58.9 48.2 45.3 46.8 44.0 41.2 41.4 42.3
natural amenities was only 31.1. Eight of the 37 cities in the former category were on the Case-Shiller list (40 per cent of that total) but only two of the 37 in the latter category were on that same list, again demonstrating that the Case-Shiller cities represent a very biased sample of the nationwide metropolitan trends. In any case, there was much more variation in the stratified mean scores for natural amenities and per capita income than for either social or human amenities. Although each of the four attributes is probably related to the nationwide rise in metropolitan house prices, for present purposes this variation merely indicates that the four effects can be rankordered just as they were by the sizes of their correlation coefficients in Table 2.1: natural > income > social > human. In summary, those cities enjoying higher (lower) levels of income and more (fewer) natural or humancreated amenities clearly experienced larger (smaller) house-price increases between 1995 and 2007. However, uncovering the precise nature of this relationship would require further spatial-econometric research.
Patterns in the house price changes The nature of the house price changes, as they unfolded over time, can be analysed in several ways. One simple way involves first calculating the Z-score1 for each quarter-to-quarter relative shift in house prices across the 370 cities. Next these 49 Z-scores are summed to arrive at a standardised overall figure for longitudinal price change in each city. Those places having rapidly growing prices in any time period have positive scores during that same time period, so places that consistently have high house-price increases – relative to all cities in the study – necessarily have high positive overall scores. Likewise those places having slowly growing prices receive negative scores and, when summed, necessarily have high overall negative scores. Of course, this score is highly correlated with the score developed earlier; in fact the rank correlation between the two is r = 0.98.
Jones_c02.indd 30
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
Table 2.4 Group 1 2 3 4 5 6
31
Numbers and attributes of the six price change clusters.
Number
Price rise
Social
Human
Natural
Income
96 136 20 50 31 27
36.2 31.3 75.9 67.7 87.2 92.4
34.2 54.0 75.2 52.6 43.9 62.7
37.7 56.7 63.8 47.8 40.2 61.7
44.1 32.9 66.5 66.4 76.3 81.6
35.1 51.6 71.3 42.3 58.6 78.0
But there are some subtle differences between the two approaches. Some cities have very dramatic price changes confined to a few time periods, and this phenomenon can affect the cumulative magnitude greatly, even though the overall pattern of house price changes might be very similar to that of other places. This is notable in places like Wheeling, West Virginia, a place that experienced a single price spike between 1996Q3 and Q4 or in Fairbanks, Alaska, a place that experienced fairly flat house prices during several quarters when the rest of the nation generally had rapidly growing prices. Of more interest, though, is using these 49 Z-scores to cluster together those places having similar longitudinal house price profiles. Ward’s grouping algorithm (Ward, 1963), which compares the within-group and betweengroup variances, was used to cluster the 370 city observations into relatively homogenous groups. Ward’s method usually produces groups with relatively similar numbers of members, in contrast to some of the simpler grouping algorithms that incrementally attach new members to existing groups. Results for a number of different cluster numbers were examined and a decision was made to analyse the run having six clusters in more detail. The results of this six-group classification are shown in Table 2.4. Group 1 includes 96 cities whose house prices grew slowly during the study period. These places, such as Birmingham and San Antonio, were generally located throughout the nation’s south and along the MississippiOhio valley up into the states of Pennsylvania and New York. These cities generally rated poorly on the four urban attributes of interest. However, the 136 Group 2 cities had even slower growing house prices, even though these places rated higher on all of the urban attributes except natural amenities. The Group 2 places, which included Cleveland and Detroit, were generally located to the north of the Group 1 places. In fact, Groups 1 and 2 together formed a much larger cluster of 232 smaller cities found throughout the nation’s interior whose membership in either group seems to have been determined by each city’s balance between natural and human-created amenities. Another 70 cities experienced house prices that grew moderately fast and this large cluster was also comprised of two separate groups. The 20
Jones_c02.indd 31
2/8/2012 9:10:02 PM
32
Challenges of the Housing Economy
cities (e.g. Baltimore, New York) of Group 3 were largely confined to the heavily urbanised north-eastern seaboard. These cities ranked high in both social amenities and income. The 50 Group 4 cities (e.g. New Orleans, Phoenix), on the other hand, were found along the south-eastern seaboard and throughout the west. These places had natural amenities that were comparable to the Group 3 places but were much poorer, and ranked much lower in social and human amenities. Finally, 58 cities exhibited remarkably high increases in house prices during the study period. One cluster, Group 5, was comprised of 31 places (e.g. Fresno, Miami, Washington) that rated low on income and on social and human amenities. Most of these places were found in the Gulf or inland from the Pacific coast. The remaining 27 cities were much richer in terms of both income and social and human amenities (e.g. Boston, Los Angeles, San Francisco). These places were largely confined to coastal California and to Massachusetts. The mapping exercise clearly indicates the geography of the pre-crisis run-up in US house prices. Most of the metropolitan areas of the nation’s interior experienced much slower growth in house prices than the coastal metropolitan areas. Most of the nation’s largest cities – those included in the Case-Shiller index – experienced much higher house-price hikes than did their smaller and more numerous counterparts.
After the bubble The preceding discussion of house price increases between 1995 and 2007 illustrates the importance of the well-known fact that all housing markets are local and cannot be understood solely in national terms. House price increases since 1995 have varied widely from one city to another and not merely randomly. Social factors, education, per capita income and natural amenities all played a role. In addition, city size (large cities tended to experience faster house price inflation), region (in general, house prices in coastal regions grew faster than in inland locations) and economic structure (specialisation in resilient industries or degree of sectoral diversification) were all important. Turning to house price declines after the housing bubble burst (primarily in 2006, but varying between September 2005 (Boston) and August 2007 (Charlotte) ), similar patterns can be observed. Table 2.5 ranks Case-Shiller’s 20 cities by their peak-to-trough percentage declines and also shows their rank in terms of peak house price levels. In most cases, the trough is December 2010, but this is not necessarily the actual trough because continued price declines are expected in many cities into 2011. Also, a few cities troughed early (Washington, DC, in September 2009 and San Diego in April 2009). The most expensive cities at the peak experienced the sharpest price declines – eight of the latter nine were ranked #9 or above in terms of
Jones_c02.indd 32
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
33
Table 2.5 Percentage peak-to-trough decline in house prices in Case-Shiller cities. City Las Vegas Phoenix Miami Detroit Tampa San Diego San Francisco Los Angeles Washington, DC Minneapolis Chicago Seattle Atlanta Portland New York Cleveland Charlotte Boston Denver Dallas Composite-10 Composite-20
% Decline 58.5 54.7 49.0 48.1 45.3 42.4 37.8 37.6 33.9 31.6 30.0 27.9 27.3 25.9 22.2 20.9 17.1 16.4 11.5 9.4 30.9 31.0
Rank in peak house prices 6 7 1 18 5 4 8 2 3 13 14 10 16 11 9 19 17 12 15 20
Source: Case-Shiller Index reported by Standard and Poors.
peak house prices; Detroit was the only exception. Apart from Detroit and Washington, DC, the prices fell most in Florida and California, plus Las Vegas and Phoenix. The top place of Las Vegas perhaps suggests that the gambling spirit of the casino drifted over into the housing market. As demonstrated by chapter 3 ‘flipping’ began to dominate the market with ‘investors’ looking at housing not in terms of personal consumption but as a profit opportunity (hoping to resell very quickly). After the market peaked, sales fell off rapidly followed by a tumble in prices because the ‘investors’ could not afford to extend the carrying costs. Table 2.6 shows the three-year decline in house prices in a second tier of relatively large cities monitored by the National Association of Realtors. The declines are less extreme than in the Case-Shiller cities (only five out of the 20 cities suffered price declines in excess of 25 per cent), but a similar pattern is found. The largest declines are found in Florida, California and Arizona while the cities in Texas are untouched. This marked difference further undermines treating America’s current housing problems as a nationwide phenomena. Table 2.6 also lists the 20 cities by metropolitan area (MA) size rank (the MAs fall within the 1–4 million size range). The relationship between
Jones_c02.indd 33
2/8/2012 9:10:02 PM
34
Challenges of the Housing Economy
Table 2.6 Percentage change in house prices (2007Q3–2010Q3) in 20 sample US cities. City Orlando Sacramento Tucson Jacksonville San Jose Baltimore Salt Lake City Milwaukee Philadelphia St. Louis Columbus Nashville New Orleans Kansas City Austin Raleigh Indianapolis Louisville San Antonio Fort Worth US
% Change
Metro size rank
−48.7 −38.0 −33.0 −32.2 −26.6 −17.1 −14.6 −10.9 −8.4 −7.6 −7.5 −6.3 −6.1 −4.2 −2.7 −2.0 −1.9 +0.5 +0.5 +1.6 −24.7
8 4 20 15 9 3 18 14 1 2 10 13 17 7 12 19 11 16 6 5
Source: National Association of Realtors.
decline rates and MA size is not fully observable in this case, with four out of ten of the cities with the largest declines ranked 14, 15, 18 and 20 in size terms. To analyse the effects of a ‘rolling recession’2 in the housing industry is a complex task, much beyond the scope of a single paper. There is no such phenomenon as a national housing market in the USA, only a proliferation of local markets that vary widely by state and metropolitan area. Mortgage rates are very similar across the country, but local economic conditions and state policies are very different. Here we will focus on one spatial issue: how did Texas manage to avoid the bursting of the bubble? Lower house prices even in peak times curbed speculative fever and lower unemployment dampened foreclosure tendencies. Houston is the only very large city excluded from the Case-Shiller sample. Houston’s price path was somewhat exceptional, peaking late (in 2008) and reviving early (in late 2009) with a narrow amplitude as elsewhere in Texas. Texas avoided much of the housing downturn as the result of many factors: relatively low unemployment; initially low house prices that changed little during the housing boom because of the lack of speculation; and a much less restrictive housing policy than in other states (Park, 2011). To comment on these in turn, Texas is among the fastest-growing economies in the USA, and during the recession it continued to grow at a rate
Jones_c02.indd 34
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
35
70 per cent faster than California while Florida did not grow at all. The unemployment rate remained about 2 per cent below the national average and 3 per cent below California and Florida. In the two years (between June 2009 and June 2011) Texas alone accounted for 37 per cent of the national job growth (265,000 out of 716,000 jobs). Texas’s house prices barely increased during the bubble, and median prices started much lower (the price in the year 2000 was only 77 per cent of that in the USA as a whole) and at the peak this ratio had fallen to about 58 per cent (the corresponding proportion of Texas to California prices was 26 per cent – Brash, 2009). The median multiple (median house prices divided by median household income) peaked in Texas at 3.2 compared to the national average of 4.5 and more than 5.0 in Florida. Low and stable prices in Texas meant that it escaped the speculative fever of California, Florida and selected cities elsewhere such as Las Vegas and Phoenix. They also meant close to negligible declines when the bubble burst. The less restrictive land policies would take up too much space to document. It is not merely the absence of zoning in Houston. Most local jurisdictions are more likely than elsewhere to grant planning permission, NIMBY opposition is much less vocal and development impact fees are lower or non-existent. Other development constraints such as urban growth boundaries and concurrency agreements (infrastructure to be in place before development is permitted) are rare. This environment has resulted in relatively unconstrained housing supply and modest housing prices in most Texas locations. In other words, although the housing bubble was primarily the result of conditions on the demand side (exacerbated by relaxed mortgage standards), the price escalation was aggravated by supply constraints. In particular, as pointed out by Glaeser and Gyourko (2003) among others, the supply of housing was restricted in the 1990s by obstacles to the granting of building permits and by the proliferation of regulations impeding residential construction. These factors pushed up prices dramatically on the supply side and this fuelled rather than dampened housing demand, primarily because of speculative forces. How this worked itself out is difficult to quantify, but it deserves further research.
Housing reforms for the future Prospects There is significant controversy about 2011. Some analysts believe that the housing recession has levelled out, while others are forecasting a ‘double dip’. There are many variables in play. Unemployment appears to have turned a corner, if modestly. The foreclosure crisis has not yet been resolved because of shoddy legal processing (such as the ‘robot-signing’ scandal, in
Jones_c02.indd 35
2/8/2012 9:10:02 PM
36
Challenges of the Housing Economy
which banks signed large numbers of foreclosure documents daily without paying attention to the requirements of the legal and judicial system), and another round of new foreclosures is expected. Nationally, close to one-third of homeowners3 are ‘underwater’, i.e. have negative equity (with mortgages larger than the anticipated sales price), giving them a possible incentive to walk away. Concerns about oil and food prices are creating an environment unfavourable to potential homebuyers. Formerly liberal lending rules are being tightened, and mortgage rates are slowly moving upwards. The National Association of Realtors is forecasting a decline of 5.5 per cent in house prices from 2010Q3 to 2011Q3, with higher forecast declines in Orlando (14.0 per cent), Sacramento and San Jose (8.3 per cent) and Jacksonville (6.6 per cent). The latest Case-Shiller index (at the time of writing for May 2011) shows a continued decline (4.2 per cent since January 2011) for the 20-city composite index. House prices are at their lowest level since 2002, and in some cities below the 2000 level. Only Washington, DC, is avoiding the decline, and seven declined by more than 7 per cent. Another analysis (by Zillow, a US housing statistical database) has revealed a 3 per cent decline in the first quarter of 2011. The ‘experts’ are predicting a possible bottoming-out later in the year but without a significant housing recovery, if at all, until 2012, and perhaps beyond. No one expects house prices to revive quickly; the 2006–07 levels are unlikely to be reached in the current decade. There are many reasons for this: a very slow recovery in the economy, especially the unemployment rate; an anticipated increase in mortgage rates fuelled by incipient inflation and rising commodity prices; a tightening up in the previously liberal mortgage lending policies; a lingering of the foreclosure problem; the slow decline in housing inventories combined with a sluggish supply response by developers; and continuing uncertainty among potential house buyers.
Promotion of home ownership In considering reforms to the system a first question to ask is, ‘Has the distribution between homeownership and renting been out of balance?’ In periods of rising house prices, if the costs of buying and renting are not too far apart, buying seems a much better deal because it can build up equity and create important household ‘wealth effects’. Via refinancing, home equity loans and second mortgages households can generate cash for consumption of cars, remodelling, vacations and other items. However, the fact that housing is a cyclical sector is often forgotten as it was in the first decade of this century (up to 2007) by mortgage brokers, realtors, banks, other financial institutions and even professors of economics and finance. There is a case for renting just as there is a case for car leasing. Home buying has its pros and cons. Mobility costs are high, sometimes prohibitively, and
Jones_c02.indd 36
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
37
maintenance costs and property taxes add substantial sums to mortgage payments. The costs of buying are destined to move much higher than in the past. Take away the prospects of a significant wealth effect, and especially for households in the early stage of their careers, homeownership may be less appealing. A major feature of US housing policy over the years has been building the ‘American Dream’ that includes a nation of homeowners. Formerly targeted at the relatively well-off, in recent years there have been attempts to reach down the income ladder. As shown by Fischel (2001), homeownervoters can make a great difference to the degree of political participation so homeownership has many desirable characteristics. However, it is possible that the federal government has taken homeownership too far, especially during the Clinton Administration and its aftermath when Franklin Raines, a strong Clinton proponent, became CEO of Fannie Mae. Risk-taking was thrown to the wind and low, even zero, down payments became common, and a variety of mortgage instruments were introduced with adjustable rates geared, at least initially, to keeping monthly payments as low as possible. The homeownership rate gradually crept up to 68 per cent of households (the latest estimate during the foreclosure crisis is that it has fallen back to 64 per cent), but this is not much higher than in Canada where there is no mortgage interest tax deductibility or in the UK where mortgage relief was abolished in 2000. This suggests that there is a case for cutting back on this subsidy to the middle and the wealthier classes (its total abolition was recommended by the 2010 Federal Debt Commission, but is very unlikely that the strong real estate lobbies will allow it). Given that a taxpayer can deduct up to $1 million of mortgage interest on two dwellings (plus unlimited property tax deductions) some restrictions seem appropriate.
Policy recommendations It is clear that the liberal mortgage lending policies that became more common from the late 1990s cannot be restored. The problem with reforms is that they may slow down the revival of the housing market by dampening demand, but that may be an inevitable price. Around the world most governments focus their housing policies on help for low- and moderate-income households leaving private markets to deal with the housing demands of the better off. The USA has been different. Despite forays into subsidies for the poor, most resources (primarily in the form of tax relief) have been allocated to wealthier households, i.e. the homeowners. The attempt to extend these benefits to lower-income groups did not succeed because the lending criteria were too lax and the assumption that house prices would continue to rise (perhaps forever) was false. Here are some policy reforms that merit consideration but not necessarily adoption.
Jones_c02.indd 37
2/8/2012 9:10:02 PM
38
Challenges of the Housing Economy
Loan modification schemes.4 There have been many schemes (either in operation or in the proposal stage) to stave off foreclosures by modifying loan conditions to keep people inside their homes. Modification may take several forms: a reduction in the mortgage rate, a discount of the outstanding principal or conversion of an adjustable to a fixed-interest rate. These have hitherto had little impact because of the slow rate of implementation. Also, although hard data are hard to find, anecdotal evidence suggests that in many cases they have not even staved off eventual foreclosure.
Mortgage lending policies ●
●
●
●
●
●
Jones_c02.indd 38
Restoring the 20 per cent down payment requirement as a general rule. Although the national average decline in prices during the recent housing recession was somewhat higher than 20 per cent (about 25 per cent), this was a very rare event, and the 20 per cent down payment is usually a reasonable level of insurance against ‘underwater’ mortgages. However, if promoting homeownership among moderate-income households remains an important policy goal, some exceptions may be needed, and current private mortgage insurance (PMI) levels may be insufficient. No interest only loans. Interest only loans make no sense now that we all recognise that the housing sector is cyclical. Allowing longer repayment periods is a better option although the principal payments in the early years of the mortgage may be quite small. Abolition of no income certification (NIC) loans. NIC loans, which carry a somewhat higher interest rate, are taken out only by borrowers who need a quick escrow or who doubt whether they would qualify under normal underwriting standards. The former group can easily be handled by prequalifying or negotiation, while the latter should have been shut out by not allowing NIC mortgages. Limited adjustable interest schemes (e.g. 1/5 or 1/7 programmes with rate ceilings).5 In recent months, these have been growing in popularity again, accounting for about 10 per cent of new mortgages. Provided that down payments are reasonable and other underwriting standards are satisfied, there is nothing wrong with this. No refinancing with subsidised loan programmes6 (for three or five years). If taxpayer resources are being used, refinancing (especially with cash out) should be restricted. Prepayment penalties for refinancing (again, for three or five years).7 Zero prepayments were very popular with borrowers but they had the unfortunate consequence of encouraging refinancing, sometimes over and over again. Only the reintroduction of sizable prepayments can slow this
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
●
●
39
process down in periods of sharply rising house prices (not very likely in the near future). The purpose here is also to inhibit refinancing which has been a major contributor to the housing price collapse. 40-year amortisations in special cases. This is not an ideal remedy (because the principal payments in the early years would be very small), but may be justifiable as part of a loan modification for current mortgages. Restrictions on second mortgages and home equity loans. Limits on second mortgages are difficult to control, but constraints on home equity loans might be easier via government negotiations with the banking industry. However, the banks may be resistant because home equity loans are profitable.
Reduction or elimination of mortgage interest and property tax deductibility There is a strong rational case for reducing these tax benefits, although their total elimination is very unlikely for political reasons. Absorption of too many resources into housing deprives other more productive sectors of scarce investment and can undermine economic growth. If mortgage credit becomes more expensive, other forms of credit, such as loans for small business, will become cheaper – a plus for the economy. The tax deductions encourage excess housing consumption in the form of homeownership at the expense of renting. According to some analysts, this also encourages more suburbanisation (most new housing is located in the suburbs) with repercussions on the economic welfare of the central city and on commuting behaviour (especially in those metropolitan areas where suburb-central city work trips remain important).8 Another problem is the highly regressive consequences of this type of tax deduction.
Reform of Fannie Mae and Freddie Mac Although Fannie Mae’s and Freddie Mac’s quarterly losses have been declining for more than a year, their combined net losses to taxpayers total about $154 billion since being taken into government conservatorship in September 2008. The forecast is that this could more than double before the crisis is over. Adding in the role of the Federal Housing Administration (FHA), the federal role in housing finance has continued to grow (and now backs about 95 per cent of new mortgages) as the private banks continue to remain somewhat hesitant about new lending. The share of the FHA – only 2 per cent of the mortgage market at the peak of the boom – is now more than 35 per cent, despite its less stringent loan requirements than the two FMs. Hitherto, the federal agencies have escaped the reforms of the
Jones_c02.indd 39
2/8/2012 9:10:02 PM
40
Challenges of the Housing Economy
Dodd-Frank Act,9 which among other restrictions impose a 5 per cent risk-retention requirement10 for securitisation of loans from which the government agencies are exempt. A first step towards equalising the playing field would be to withdraw that exemption. Perhaps the private sector could be encouraged to be more innovative. For example, mortgages could be issued in which a bank takes an equity share in the home, thereby sharing in the risks and rewards of price declines and increases. Also, pools of mortgages could be created and managed by the banks with the capacity to modify the loans to maintain the quality of the bonds behind the mortgage pools. The European Union has outlawed governments from backing mortgage companies via guaranteeing the loans. The big banks have not objected because it has been profitable to securitise mortgages guaranteed by the FHA or to sell mortgages for the two FMs. While the private banks were primarily responsible for the subprime mortgage epidemic, the federal agencies placed no obstacle in their way. Both Fannie Mae and Freddie Mac are so-called government-sponsored enterprises (GSEs). They are technically private companies, designed to promote home mortgages but enjoying government insurance of their liabilities without cost, the right to borrow at little more than the Treasury rate, and with little oversight. It was a win-win-win: high returns for investors, excellent salaries for senior managers and campaign contributions for the politicians, balanced by huge costs for the taxpayer. Using a metaphor borrowed from Mary Shelley, Dr Frankenstein (the US Government) created a monster (Fannie Mae and Freddie Mac) that they could not control (Acharya et al., 2011). In 1970, the two FMs accounted for only 4.4 per cent of mortgages; by 1991 it was 28.4 per cent and by 2009 it had risen to almost 45 per cent. Currently, the Federal Reserve Bank is responsible for $1.5 trillion of GSE debt and backed securities. The two FMs have $3.5 trillion of mortgage guarantees, a $1.7 trillion mortgage portfolio and $2.2 trillion of derivatives. This is the socialisation of what should be a private mortgage market. Can the home mortgage market stand on its own without federal support? Probably, yes. Accounting irregularities at the two FMs in 2003 that affected their business had no effect on mortgage rates. The jumbo mortgage11 market has operated efficiently without government intervention and little penalty in terms of higher rates. Also, the auto loan market is partially securitised and has performed very well without the government’s involvement. Because the two FMS, given their size, would be difficult to close overnight, if they were prohibited from buying new mortgages they would be gradually phased out as the existing mortgages were paid off. Similarly, the imposition of higher fees would facilitate the transfer of the securitisation function to the private sector. The US Treasury-HUD Report of February 2011 (US Treasury-HUD, 2011) raised the possibility of abolishing Fannie
Jones_c02.indd 40
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
41
Mae and Freddie Mac reeling under the pressures of the $154 billion taxpayer bill, a view that is supported by most Republicans. Regardless of the decision, it remains unclear whether government support to the middle class should continue or whether every responsibility should be transferred to private mortgage markets. There is a broad consensus that reforms are badly needed. In the ‘good old days’ the two agencies performed by securitising mortgages with 20 per cent down payments (or a 10 per cent down payment backed up by PMI). By returning to that approach combined with standard fixed-interest mortgages, without any involvement in low down payment loans, their survival might be acceptable. Private markets could handle all the other loans such as investment properties, home equity loans, second mortgages and jumbo loans. Some analysts argue that complete privatisation would increase risk and that a degree of government involvement is needed to reduce risk (or, at a minimum, providing enough information to make credit risks transparent, via standardisation of mortgage products). Improved mortgage insurance schemes and government guarantees are useful mechanisms of intervention, although higher fees might encourage the substitution of private insurers for the government. Also, they might depress house prices still further, but not by much. As for government guarantees, some have argued that shifting more responsibility to private banks in a world without the FMs would lead to even more government outlays. Insured banks (or those who take them over in times of stress) enjoy government guarantees via the Federal Deposit Insurance Corporation, which can result in very costly bailouts (not only recently, but in the 1980s’ savings and loans crisis). There is certainly a severe ‘moral hazard’ problem because guarantees encourage more risk-taking. Mortgage rates would vary somewhat more widely without the FMs, according to the mortgage menus and, of course, the borrowers’ income and credit situation. In the absence of government guarantees (almost universal since the 1950s) private lenders would be much more careful in accounting for differential risks. Denmark and some other European countries offer long-term, fixed-rate mortgages without government support. Also, some countries do well with some types of variable rate mortgages; for example, in Canada five-year rollover mortgages are quite popular. The two FMs did not allow for much rate variation, preferring bureaucratic simplification. Also, because they are not direct lenders but bundle mortgages and, via securitisation, sell them to third parties, they are not as concerned as private markets by risk issues. For example, they disallowed prepayment penalties. However, a counter argument to criticisms of FM securitisation is that ‘private-label’ securitisation is even riskier with higher default rates.
Jones_c02.indd 41
2/8/2012 9:10:02 PM
42
Challenges of the Housing Economy
Thus, there are two countervailing extreme views about how to deal with the two FMs. One is to abolish them altogether and fully privatise the mortgage market. The other is to retain them in close to their current role but to refocus their attention to those more in need, leaving the well-off to work out their best deals with private sector actors. The first approach implies getting rid of the government’s role altogether once the rules are decided and implemented. The second probably requires a further downward revision in the conforming loan limit far below the newly proposed maximum of $625,500 and some reinforcement of mortgage insurance. An intermediate step to achieve the latter, with the advantage of reducing bureaucratic costs, would be to amalgamate the two FMs with the FHA and consolidate their roles, but much would depend on the terms of the amalgamation.
Other proposals There are other options for alleviating the current problems. For instance, the FHA which insures 34.5 per cent of all borrowers might permit its borrowers who become unemployed to shift temporarily to lower payments. However, the large differences between paid wages and unemployment compensation might make this an impractical solution for most households. To the extent that it is viable, it might both satisfy the banks and stave off foreclosure. Another idea is to introduce a wage-based housing fund (several countries around the world have versions of this) in which workers (and possibly employers) can make tax-free contributions to an account that can be drawn upon for housing payments in times of need; if unused they could be paid out in full at Social Security age. These are certainly not panaceas or a substitute for more orthodox efforts to reform housing finance.
The renting option – introduction There have been some recent discussions within the Department of Housing and Urban Development (HUD) of shifting federal housing policies more in favour of renters, but any detailed proposals have yet to be announced. There is a Rental Policy Working Group in HUD, but its members had to sign a non-disclosure agreement in February 2011 so (at the time of writing) its deliberations are secret. However, rental policy reform is under serious consideration. In the words of Raphael Bostic (a senior official at HUD and an academic colleague of Richardson): ‘In previous eras, we haven’t seen people question whether homeownership was the right decision. It was just assumed that’s where you want to go. You’re not going to hear us say that. … (W)hat we’ve seen in the last four years is that there really is an underside to homeownership.’
Jones_c02.indd 42
2/8/2012 9:10:02 PM
US Housing Policy in the Era of Boom and Bust
43
This decision to focus more on rental housing and less on homeownership differs in many ways from the Bush and Clinton administrations. President George W Bush promoted an ‘ownership society’ that sought to increase homeownership rates, specifically for low-income people. President Clinton had a ‘national homeownership strategy’ that advocated a specific homeownership rate. HUD Secretary Shaun Donovan has been the most vocal advocate of the new rental-oriented approach: ‘While we continue to promote affordable homeownership, for many Americans renting will continue to be the only or preferred option’ (Congressional testimony). Supporters of rental housing believe that this represents the view of the Obama administration as a whole: ‘My impression is that the administration is serious about a balanced policy … Their purpose is to look at and make more workable rental-housing programs.’ (Vincent O’Donnell, Local Initiatives Support Corporation). However, there is concern from both sides of the debate about turning away from homeownership. For example, ‘This is confusing to me– the view that the best way to help someone accumulate savings over time is to subsidise their rent now’ (Keith Hennessey, former Director of the National Economic Council under Bush). Or, ‘(w)e have to define more clearly what we mean by homeownership for low-income families and make sure we don’t … come to a very simplistic reading of our recent history that it was simply lending to low-income families that got us into trouble’ (Janis Bowdler, La Raza, a Latino civil rights group).
Homeownership for low-income households To follow up on this last point, it does not follow that there is no longer any hope for low-income households to buy their own home. As pointed out earlier, housing markets are local not national, and there are some cities with relatively low house prices where the costs of buying do not differ very much from the costs of renting. Many cities away from both coasts fall into this category including much of a large state such as Texas. The only major obstacle in such cases is probably the down payment requirement, and a federal matching grant might be a one of several options worth consideration.
Rental subsidies Rental subsidies have had a chequered history in the USA from the long-abandoned public housing projects (apart from mixed-income redevelopment schemes) to a variety of demand-side rent subsidies, of which the most successful and most popular has been the Section 8 programme (which pays a low-income household the difference between 30 per cent of income and the fair market rent, and currently costs about $8.2 billion per annum).
Jones_c02.indd 43
2/8/2012 9:10:02 PM
44
Challenges of the Housing Economy
However, in recent years the Section 8 programme has been running into trouble, especially in high-rent cities. The demand for the Section 8 apartments has far surpassed the supply because of the reluctance of many landlords to participate in the programme. One palliative may be to supplement the demand subsidies with some assistance on the supply side, perhaps in the form of a cash payment to landlords, an upgrading allowance or improved depreciation rates. A more direct way to even the playing field between owners and renters is to offer all renters a tax credit. There are precedents for this. Although it is a pittance, California offers renters a tax credit of $30 per year to offset their state tax payments. Over the years, a useful rent subsidy strategy has been to give developers planning permission to build apartments (sometimes on favourable terms, e.g. with density bonuses) in return for allocating a proportion of the units (perhaps 10, 15 or 20 per cent) for low-income households at subsidised rents while the rest of the units are market rent (this procedure usually involves a degree of cross-subsidisation). This is almost always a local government initiative, reflecting the long-established principle that local jurisdictions have the ‘police power’, i.e. planning authority. How the federal government might fit into this strategy is unclear and perhaps unanswerable, but there might be some scope for ‘piggy-backing’ with additional incentives.
Conclusion The housing market in the USA is very complex because of the extremely large geographical size of the country and the wide differentials in regional economic performance. There is no uniformity in the housing markets although mortgage rates are more or less the same nationwide. Hence, geographical variations in the housing boom or bust are a big part of the story. The housing crisis is not yet over, but hopefully is in its later stages. There is more or less universal agreement that policy reforms are needed to avert major problems in the downward phase of the next cycle, probably many years away. It is unclear which of the recommendations in this chapter will be proposed and adopted. Certainly, mortgage lending policies are already tightening, if not in specific ways, the future of Fannie Mae and Freddie Mac is in doubt and there will be some changes in rental housing policies. Hitherto, the Obama administration has offered only the most vague of generalisations. Its specifics will probably not be announced until very late in 2011. Certainly, the past five years have been a very challenging period in US residential real estate. No one expects a quick recovery although opinions
Jones_c02.indd 44
2/8/2012 9:10:03 PM
US Housing Policy in the Era of Boom and Bust
45
vary widely as to when the market will return to 2006 levels. The most optimistic is 2015, the most realistic perhaps 2020, and the most pessimistic perhaps never.
Notes 1. The Z-score is a simple and well-known statistical measure that compares the mean of any particular variable with the mean of an overall normally distributed population, expressed in terms of the standard deviation of that population. In other words, in this case it examines whether any quarter over 49 quarters is above or below average. See Wikipedia for a more precise definition. 2. The ‘rolling recession’ concept can be applied to state, regional or metropolitan economies as a whole and is not limited to construction. Downturns begin in different locations at dissimilar dates and reverse in the same way. The standard, but not uniform, phenomenon is if an area leads in it leads out, while if it lags in it lags out. 3. In some states, the numbers are even worse, e.g. Nevada (65 per cent), Arizona (48 per cent), Florida (45 per cent), Michigan (37 per cent) and California (35 per cent). These statistics refer to the third quarter of 2009; they have probably deteriorated since then. They are based on a comprehensive database produced by First American CoreLogic (www.corelogic.com). More recent data would have to be purchased commercially. 4. Loan modification schemes are mechanisms to reduce monthly mortgage payments in order to allow people to keep their homes. Hitherto, government-initiated measures have not fared well. The bailout of the banks authorised $46 billion for housing assistance; but as of May 2011 only $1.85 billion had been spent. The target of the Home Affordable Modification Program, begun in early 2009, was up to 4 million homeowners; so far, 670,000 have been helped. The Hardest Hit Fund has a budget of $7.6 billion, but hitherto $455 million has been spent. 5. These schemes place limits on the amount by which mortgage rates can be adjusted after, say, 5 or 7 years. A major impact will be that the initial rates will be much lower than the ‘teaser’ rates offered in the past. 6. Subsidised loan programs are measures aimed to promote homeownership among lowincome households by subsidising mortgage rates and/or by allowing zero or very low down payments. 7. Prepayment penalties used to be standard: paying a substantial amount for ending your mortgage early. They have disappeared, more or less, in recent years. This encouraged both ‘flipping’ and refinancing combined with cashing out sizable sums. Reviving prepayment penalties for the first few years could dramatically reduce these speculative aspects of the housing market. 8. This view has been challenged in the recent crisis because sales in some suburban areas have been even slower than in the central cities. There is anecdotal, and in a few cases from the 2010 Census of Population more solid, evidence of a ‘return to the city’ movement. However, analysis is a little complicated because of cheaper dwelling units (especially condominiums) in the central city and higher petrol prices. 9. The Todd-Frank Act (the Wall Street Reform and Consumer Protection Act) was passed by the US Democratic Congress in July 2010. It is a 2,319 page law that covers a wide range of banking, financial and consumer reforms (many of them affecting mortgages) via rules that will take years to implement. The regulators have already fallen so far behind in their rulemaking that the law is now being derided as the ‘Dawdle-Frank Act.’ It remains unclear what the final outcome will be. See Acharya et al. (2010) for more analysis.
Jones_c02.indd 45
2/8/2012 9:10:03 PM
46
Challenges of the Housing Economy
10. The risk-retention requirement compels lenders to retain a financial stake in the eventual outcome of loans. 11. Jumbo mortgages refer to mortgages that are too large to be guaranteed by Fannie Mae and Freddie Mac, so that borrowers have to rely on private banks that will charge them a higher mortgage rate at least 0.5 per cent higher. The maximum is $729,750 to be cut down this year (2011) to $625,500. This threshold applies only in very expensive markets. The national lower threshold is $417,000.
Jones_c02.indd 46
2/8/2012 9:10:03 PM
3 Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas Craig A Depken II, Harris Hollans and Steve Swidler
A key issue in the almost ubiquitous global housing boom of the past two decades has been the role and importance of speculation and the existence of bubble effects driving house prices to unsustainable levels. This chapter focuses on this issue and illustrates the nature of speculative investment and local market dynamics in an extreme case study of Las Vegas within the recent housing market boom and bust of the USA. During the upturn in numerous local US housing markets, property ‘flipping’ became a popular investment strategy whereby houses where bought and quickly resold for a profit. The argument is that this process became self-fulfilling by creating a bandwagon and led to inflated unsustainable home prices, and what is now frequently referred to as a ‘housing bubble’ (see Wheaton and Nechayev, 2008). Ultimately, with the collapse of the US housing market, a national mortgage foreclosure crisis ensued as detailed in chapter 2, part of which stemmed directly from this property flipping. To analyse this process the chapter starts from the premise of a threestage boom to bust housing cycle as set out in Depken et al. (2011) that explicitly incorporates the role of flipping. The first stage is a period of rapidly increasing prices (both absolute and relative) driven by a large volume of speculative purchasers entering and exiting the market over a relatively short time horizon. The second stage of the bubble is a period of declining
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c03.indd 47
2/7/2012 10:17:54 AM
48
Challenges of the Housing Economy
profitability for those following this short-term flipping strategy. As the late buyers find that they are unable to turn an earlier purchase for a significant profit due to the declining pool of buyers, the market experiences an incipient decline in the percentage change in home prices. This middle period is when flippers are left ‘holding the bag’. The third stage of the bubble is the foreclosure phase which is marked by precipitously declining house prices and a rapidly rising number of mortgage foreclosures. This period of growing foreclosure activity leads to the presence of an increasing shadow inventory, or a significant overhang in supply, which must be absorbed for the housing market to have any opportunity to recover. This final stage, the resolution of homes in foreclosure, is still playing itself out in many local markets across the USA in 2011. The analysis in this chapter describes in detail the Las Vegas, Nevada, housing bubble, a market that witnessed annual returns as high as 40%, and more recently double-digit negative returns. It is a city that exemplifies the extremes of the housing market in the USA and so is a useful case study of speculation, to assess the underlying conditions that enable it to flourish and the ultimate consequences in the aftermath of the bust. As chapter 2 notes, Las Vegas experienced the highest fall in house prices from peak to trough of any city in the USA, and Nevada has the second highest level of mortgage arrears of states, after Florida, in 2011. In the early part of the past decade, Las Vegas was the fastest-growing major city in the USA which a priori is a major factor underpinning the start of the local housing boom. The structure of the chapter is as follows. First, it examines the market data on which the research is based and looks more precisely at the definition of property flipping. The study is based on the market for single family home sales in the Las Vegas metropolitan area. The next section examines how the type and composition of housing market transactions, from free market to forced sale, changed over the boom bust cycle 1994–2009. The analysis then focuses on the third stage of the boom to bust cycle outlined above, namely the foreclosure phase. It considers the changing position of foreclosures in the local market and discusses the implications of widespread foreclosures for valuation/appraisals, constructing house price indices and the consequences for an ultimate recovery. Finally some general conclusions are drawn.
Data and definition of property flipping The data employed here are derived from the tax assessor’s records of all single-family detached home sales in Clark County, Nevada, which encompasses the Las Vegas metropolitan area. The data stretches over a 15-year
Jones_c03.indd 48
2/7/2012 10:17:54 AM
Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas
49
period from 1994Q1 through 2009Q4 that encompasses the housing boom and bust, and are described in full in Depken et al. (2009, 2011). At the heart of the empirical analysis is the Clark County tax assessor’s use of a multi-criteria coding system to identify various types of property transfers. Three primary types of transaction codes are of interest in this analysis: the arms-length transaction, denoted as an ‘R transaction’, where the sale price is that determined between two independent parties; a trustee’s deed, denoted as a ‘T transaction’, where the sale price is the highest price bid on the property at foreclosure auction during a trustee sale; and a resale of the property subsequent to foreclosure, denoted as an ‘F transaction’, where the sale price is that which the lender obtained upon reselling the property after foreclosure. To provide context about the distribution of these transaction types, prior to the market peak in 2006, the percentage of transactions that were arms-length sales consistently exceeded 95%. However, as home prices began to fall after the peak, the arms-length (R) transactions gave way to trustee transactions (T) and foreclosure (F) influenced sales by an ever-increasing margin. By the end of 2009, the arms-length transactions had fallen to 35% and trustee and foreclosure influenced transactions had risen to 34% and 31%, respectively. For the purpose of collecting data to be used in tax appraisal models, it is important for the county assessor to distinguish between arms-length sales and transactions which are influenced by a home foreclosure. The R transaction represents a sale between a willing buyer and willing seller. However, the T and F transactions are influenced by the conditions dictated by a mortgage foreclosure. In the early years of our analysis the F transactions represented a Deed in Lieu of foreclosure, where the property owner simply hands over title and all equity interest to the lender. This process was made easier by changes in Nevada foreclosure law and provisions of the Trouble Asset Relief Program (TARP) and the American Recovery and Reinvestment Act (ARRA).1 In the later years of the sample period, the county assessor changed the definition for an F transaction to include any distressed sale which occurs between a lending institution holding the house and a new owner. As the next section shows, a significant contributor to the bubble story is property flipping. Depken et al. (2009) first define a property flip as a dyad of two arms-length (R) transactions within two years of one another, a time frame largely motivated by tax considerations. The front-end of the dyad has been termed the ‘buy flip’, whereas the back-end transaction has been coined the ‘sell flip’. Employing hedonic models, Depken et al. (2009) find that buy flip transactions, on average, occur at a discount from otherwise similar homes, whereas a sell flip sells at a premium.
Jones_c03.indd 49
2/7/2012 10:17:54 AM
50
Challenges of the Housing Economy
0.5
250,000
0
200,000 –0.5
Median price
Per cent/Pct. change
300,000
150,000
100,000
–1
1994Q1 1996Q1 1998Q1 2000Q1 2002Q1 2004Q1 2006Q1 2008Q1 2010Q1
Per cent flips Pct. change med. price
Per cent foreclosures Median price
Figure 3.1 Flips, foreclosures and prices in Las Vegas through the boom bust cycle, 1994–2010. Source: Depken et al. (2011).
Anatomy of boom bust market dynamics The three phases of the housing bubble of Depken et al. (2011) projects a changing role for flipping and speculation through the boom bust cycle and this is depicted in vivid detail for Las Vegas in Figure 3.1. The figure overlays four different time series: the quarterly median transaction price, the quarterly percentage of all transactions that were either the front end (buy side) or back end (sell side) of a house flip, the quarterly percentage of all transactions that were some form of foreclosure and the quarterly percentage change in median transaction price of all house sales in Las Vegas. As can be seen, early in the upturn of the cycle the number of foreclosures was scant (as a percentage of overall transactions) and the quarterly percentage change in median house prices was rather nominal. However, as depicted by the dashed line in Figure 3.1, the median house price experienced a continual rise over the early part of the study period with a marked increase in the slope of the median price trend starting approximately in 2002Q1. The median price trend has a shape that is similar to the trends experienced in other US cities, with the median price peaking in 2006Q2. Soon after the median price experienced its most dramatic increase, the number of foreclosures goes to zero, as any mortgage holder who was at risk of default could simply sell the property for more than they owed. However, this honeymoon from default was short-lived; the time of zero foreclosures
Jones_c03.indd 50
2/7/2012 10:17:54 AM
Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas
51
lasted from 2004Q2 through 2006Q1, after which the number of foreclosures once again increases such that by 2007Q4 more than 50 per cent of all recorded house transactions are some form of foreclosure. On the other hand, as the median house price enjoyed a steady if not meteoric rise during the period between 1994Q1 and 2002Q1, the number of transactions that were part of a flip, either the buy side or the sell side, were somewhat nominal. Nevertheless they were always slightly more than foreclosures as a percentage of all transactions. However, after 2002Q1 when the median house price increased substantially more, from quarter to quarter, than in the past, the percentage of all transactions that were part of a flip also increased. Flipping reached a peak in 2004Q1, but remained a plurality of all transaction types until market prices topped out in 2006Q2. At this point, flippers could no longer (easily) sell their home at a higher price. A primary reason for this is quite simply the increasing number of houses on the market relative to the number of potential home buyers. For much of the period from 2002 through 2005, flippers were literally selling to flippers, but by 2006, the available housing stock largely satiated owner-occupied demand for homes. In fact, from 2000 to 2008, the population of Las Vegas grew by 34%, but the housing stock almost doubled. By 2007, foreclosures began to increase noticeably. We describe this last phase of the bubble in more detail in the next section. What is interesting is that while median prices were falling throughout 2007 and the first half of 2008, a non-trivial number of transactions were still part of a flip, although during this time the vast majority of the flip transactions were on the sell side. Even as foreclosures mounted in 2009Q1, there were still some transactions that were the back end of a flip transaction.
Foreclosure activity after the bubble burst While the previous discussion describes the basic phases of the housing bubble in Las Vegas we now explore the metropolitan area’s recent foreclosure experience following the beginning of the fall in nominal prices. Table 3.1 lists the number of trustee (T) transactions for each quarter beginning with 2007Q1. This period marks the beginning of significant foreclosure activity in Las Vegas. A T sale is the most common (first) foreclosure transaction and transfers the property to the highest bidder, likely the primary lender. As Table 3.1 illustrates, the number of T transactions rapidly increased during the sample period, starting at 646 in 2007Q1 and rising to levels of four to five thousand per quarter by 2008 and 2009.2 To assess how these foreclosures influenced the structure of the housing market we analyse what price ranges they are to be found in – low, mid or
Jones_c03.indd 51
2/7/2012 10:17:54 AM
Jones_c03.indd 52
2/7/2012 10:17:54 AM
1 2 3 4 1 2 3 4 1 2 3 4
2007 2007 2007 2007 2008 2008 2008 2008 2009 2009 2009 2009
3053
646 980 1432 2231 3060 4586 5108 4473 4319 3591 4431 1787
Total T trans in quarter
197071 103131 52.33%
10.16%
6.50% 5.71% 7.12% 7.66% 8.27% 9.44% 10.40% 11.58% 11.16% 13.03% 10.31% 11.75%
Decile 1
244757 129630 52.96%
9.65%
5.57% 7.76% 7.75% 7.44% 9.28% 10.07% 9.81% 9.93% 10.81% 10.14% 10.25% 9.51%
Decile 2
266931 142034 53.21%
9.96%
9.29% 10.00% 9.01% 9.01% 8.07% 9.40% 10.92% 10.31% 10.35% 10.41% 10.38% 10.30%
Decile 3
285449 154725 54.20%
10.12%
10.68% 8.47% 9.85% 9.55% 10.16% 10.25% 9.98% 9.84% 9.42% 10.33% 11.06% 11.36%
Decile 4
302553 164052 54.22%
9.59%
9.91% 9.29% 9.22% 9.14% 9.22% 9.49% 8.91% 10.53% 9.52% 9.77% 10.11% 9.51%
Decile 5
Highlighted cells are the three top deciles for T transactions in a specific quarter. *Peak value is a home’s hedonic value at the peak of housing prices in 2006Q2. **Average is the three-year total for a decile’s T transactions divided by all T transactions for years 2007–2009. ***Ave actual/ave peak is a decile’s average T price divided by average value of homes in 2006, quarter 2.
Average peak value Average actual price Ave actual/ave peak***
Average**
Quarter
Year
319860 175314 54.81%
9.22%
10.68% 10.41% 7.82% 8.25% 9.38% 9.46% 9.01% 9.30% 9.08% 8.94% 9.84% 9.23%
Decile 6
Table 3.1 Las Vegas trustee’s (t) transactions in each quarter by peak value deciles, 2007–2009*.
341128 192596 56.46%
10.23%
11.30% 10.20% 11.03% 10.76% 10.56% 10.29% 10.20% 9.59% 10.49% 10.11% 9.86% 10.13%
Decile 7
368878 207499 56.25%
9.70%
10.68% 10.71% 11.59% 11.16% 9.80% 10.23% 9.46% 9.57% 8.61% 9.69% 8.82% 9.79%
Decile 8
412208 243554 59.09%
9.85%
12.07% 12.76% 10.13% 11.79% 10.92% 10.12% 9.49% 9.03% 9.82% 8.24% 9.61% 9.18%
Decile 9
522584 338236 64.72%
11.52%
13.31% 14.69% 16.48% 15.24% 14.35% 11.25% 11.82% 10.33% 10.74% 9.33% 9.75% 9.23%
Decile 10
Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas
53
high priced segments. Table 3.1 segments the homes according to their value in 2006Q2, the period for peak housing prices in Las Vegas. By anchoring values to one point in time, the analysis circumvents the problem that a home’s value will have fallen if it goes into foreclosure later in the period. To determine every home’s value, we estimate a hedonic model and substitute the relevant values as of 2006Q2. The hedonic model’s attributes include age, square footage, number of bedrooms and number of bathrooms and adjusts for the home’s tax district.3 Estimated home values are then partitioned into deciles. The lowest decile has an average 2006Q2 value of $197,000, while the top decile’s average value is more than $522,000. Looking at deciles 5 and 6, the median home value looks to be roughly $310,000 for 2006Q2 and mirrors Las Vegas housing prices as reported by Case-Shiller (see glossary). The results in Table 3.1 show that foreclosures were not distributed uniformly across home values. Initially foreclosures predominantly came from the high end of the spectrum. In 2007Q1, decile 10, with the highest home values, had the greatest number of foreclosures. Moreover, in 2007, the three top deciles for foreclosures always came from deciles 7–10. Thus, in the early part of the sample, foreclosures were predominantly centred in the high end of the value distribution. But by mid 2008, circumstances dramatically changed, and the majority of the foreclosures switched to the low end of housing values. In fact, for four of the last six quarters, the highest percentage of foreclosures came from decile 1. Finally, Table 3.1 shows that the ratio of average T price to average peak value almost monotonically increases with home value. This may be a vintage effect, with low-end homes defaulting later in the sample. Still another possibility is that given the terms of their mortgage, low-end homes can be purchased at auction with lower bids, on average. Without mortgage information, one can only speculate as to the types of loans that have defaulted in the study period. It seems reasonable to hypothesise that many of the homes initially going into foreclosure were high-end homes with jumbo mortgages (see chapter 2), adjustable rates and/ or high loan-to-value ratios. This intuition is supported by recent research by Amromin et al. (2011) who find that complex mortgages taken by those with higher income and better credit scores had higher default rates than fixed-rate mortgages taken by those with similar characteristics. Alt-A loans likely funded a number of these properties. An Alt-A mortgage (Alternative A-paper), is riskier than prime and less risky than subprime and typically characterised by borrowers who have less than full documentation, lower credit scores or require higher loan-to-value ratios or relate to investment properties. However, once housing prices began to fall significantly, and then the unemployment rate shot up in Las Vegas, foreclosures became dominated by
Jones_c03.indd 53
2/7/2012 10:17:54 AM
Jones_c03.indd 54
2/7/2012 10:17:54 AM
1 2 3 4 1 2 3 4 1 2 3 4 2007–09
2007 2007 2007 2007 2008 2008 2008 2008 2009 2009 2009 2009 Total
646 980 1432 2231 3060 4586 5108 4473 4319 3591 4431 1787 36644
Total T trans in quarter 256 595 1163 2009 2806 4232 4736 4104 3778 2591 1512 16 27798
F trans follows T trans 39.63% 60.71% 81.22% 90.05% 91.70% 92.28% 92.72% 91.75% 87.47% 72.15% 34.12% 0.90% 75.86%
Percentage F trans 332 318 191 111 87 114 131 63 48 76 27 2 1500
R trans follows T trans 51.39% 32.45% 13.34% 4.98% 2.84% 2.49% 2.56% 1.41% 1.11% 2.12% 0.61% 0.11% 4.09%
Percentage R trans 21 16 16 25 41 34 24 13 15 7 27 0 239
T trans follows T trans
*Shadow inventory are homes that had an initial trustee (T) transaction, but have not subsequently sold before the end of 2009.
Qtr
Las Vegas transactions following trustee’s (t) transaction by quarter, 2007–2009.
Year
Table 3.2
3.25% 1.63% 1.12% 1.12% 1.34% 0.74% 0.47% 0.29% 0.35% 0.19% 0.61% 0.00% 0.65%
Percentage T trans
37 51 62 86 126 206 217 293 478 917 2865 1769 7107
No sale after T trans
5.73% 5.20% 4.33% 3.85% 4.12% 4.49% 4.25% 6.55% 11.07% 25.54% 64.66% 98.99% 19.39%
Percent shadow inventory*
Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas
55
low-value homes. In 2007 and early 2008, the unemployment rate hovered between 4% and 5%.4 Then in August 2008 unemployment reached 7% and by the end of the year stood at 9.3%. The unemployment rate continued to rise throughout 2009 reaching 14.5% at year’s end. These foreclosed properties were probably originally bought with subprime mortgages. Once in foreclosure, what then happened to these mainly real estate owned (REO) properties, i.e. houses owned by the original mortgage lender, usually a bank? Table 3.2 reports the sale following the T transaction in a specific quarter. In the first quarter of 2007, more than half the subsequent sales are arms-length, R, transactions. But this quickly changes and F becomes the predominant second transaction after foreclosure beginning in 2007Q2. For the entire 2007–2009 period, an F transaction follows more than 75% of all homes in foreclosure. During this period, arms-length transactions following a T include only 4% of the sample. With the exception of a small number of T following a T transactions (presumably one investor selling to another), the remaining homes (19.39%) did not resell before the end of 2009. These 7107 REO homes thus became a net addition to a shadow inventory and pointed towards further declining prices in 2010 and beyond. The figures in Table 3.2 point to several important implications for Las Vegas housing prices starting in 2007. First, county tax assessment figures based on arms-length transactions were constructed on a dwindling number of R transactions. Second, housing price indexes like S&P/Case-Shiller are also based on arms-length transactions and were thus estimated on a small and declining sample. Finally, to the extent that foreclosure prices are not being taken into account directly through the use of F transactions, the actual decline in value may be greater than the estimated decline either by the county or S&P/Case-Shiller. This last effect is also a concern for private appraisals/valuations which currently may include distressed properties. Pending legislation in Nevada would prohibit the use of foreclosures and short sales in the estimation process.5 The fear of some legislators is that these transactions depress appraisal figures, thereby increasing the likelihood that the appraisal will come in under the offer price and scuttle the real estate sale. This in turn will exacerbate the downward spiral in house prices.
Concluding thoughts In hindsight, it is easy to explain the recent housing bubble in Las Vegas. The national rise in house prices, especially the acceleration in the first half of the last decade, combined with a rapidly growing city and the expansion of the local housing market created conditions ripe for speculation. Flipping
Jones_c03.indd 55
2/7/2012 10:17:55 AM
56
Challenges of the Housing Economy
activity fuelled the increase in demand for housing. At one point, more than 40% of all home transactions were a part of a flip. While prices rose, people continued to buy under the Greater Fool theory that they could sell the home for even more. At some point, though, the supply of housing stock outstripped the demographic need, and flippers could no longer find someone to purchase their home at a higher price. As price increases ceased, flippers were caught owning a depreciating investment. Profits turned to losses and falling prices led some homes to go into foreclosure. Immediately after the Las Vegas bubble popped, foreclosure activity centred on homes at the upper end of the market. It was not until mid 2008 that low-end homes and the collapse of the subprime market became the primary driver of downward spiralling prices. This timing closely followed the path of unemployment in Nevada. With double-digit unemployment rates, many people could no longer afford to make mortgage payments, and foreclosures, especially in the subprime market, escalated. There may also be feedback effects as home prices fell, construction also declined pushing the unemployment rate up. With foreclosure activity becoming the predominant sale type, one question is how to incorporate this information in home valuations. Traditionally, distressed sales are not included in either county assessments or house price indices like Case-Shiller. Thus, foreclosures only enter home valuation indirectly through depressed price levels of arms-length transactions. From a practical standpoint, perhaps a more important issue is whether to include distressed properties in private appraisals. To do so may lower appraised values and potentially cancel the sale of the property. Not to include these properties means that there is still disequilibrium between price and value. Until this disequilibrium is remedied, there can be no turnaround in the housing market. Finally, it should be pointed out that flipping activity as a driver of a housing bubble is not limited to the Las Vegas market. Clark (2007) described the ‘simplicity’ of flipping in London, pointing out how easy it is to make money in a rising real estate market. More recently, China experienced double-digit annual returns, in part caused by flipping activity and speculative investment behaviour. In fact, in response to fears that an unsustainable bubble had formed, the Chinese government introduced a tax to discourage flipping activity (Anderlini, 2010). Thus, while we have only formally discussed the real estate bubble in Las Vegas, there is evidence that similar forces caused other bubbles around the world. This is important because when attempting to reconcile the factors which led to the most recent global recession, most discussions draw the conclusion that the housing market played a primary role in precipitating the rapid fall of the US and European economies (Shiller, 2009).
Jones_c03.indd 56
2/7/2012 10:17:55 AM
Housing Bubbles and Foreclosures that Follow: The Case of Las Vegas
57
Notes 1. TARP was passed by the US government in October 2008 and allowed large financial institutions access to direct assistance from the US Treasury. The programme included a provision for banks to sell distressed residential mortgage debt back to the US Treasury. ARRA was a stimulus package passed by the US Congress in February 2009. One provision was a federal income tax credit for first-time homebuyers that ultimately expired in April 2010. 2. Data for 2009Q4 are incomplete and cover only a portion of the quarter. 3. A complete description of the hedonic model and coefficient results may be obtained from the authors and is similar to that used in Depken et al. (2011). 4. Unemployment figures from the US Department of Labor, Bureau of Labor Statistics. 5. See SB 330 in Nevada’s 76th (2011) legislative session.
Jones_c03.indd 57
2/7/2012 10:17:55 AM
4 Unemployment Risk, Homeownership and Housing Wealth: Lessons from Bubble Aftermath in Japan Yoko Moriizumi and Michio Naoi
The recent financial crisis, initially emanating from the US subprime loans, had a drastic impact on the housing markets elsewhere in the world. In Japan, the early symptoms of the crisis were not so serious because Japanese financial institutions are relatively less exposed to mortgage-backed securities (MBS). However, the global downturn triggered by the subprime loan crisis ultimately hit the Japanese economy strongly. As a result, the number of loan defaults and foreclosures rose sharply in 2008. Obviously, there are various channels through which the global financial crisis can influence the local housing market and consumer behaviour. Among these, we will focus on two potential channels. First, the fallout of the global financial crisis on the labour market that was dramatic. During the global financial crisis, unemployment rose from 3.9% in 2007 to 5.1% in 2009, a relatively high figure in Japan where the historical high of 5.4% was registered in 2002. A rising unemployment rate and the resulting income uncertainty may restrict access to homeownership particularly among young renters. Second, the large swings in housing prices can affect homeowners’ consumption and saving. In fact, housing prices in Japan dropped by about 10%, and several other housing indicators showed a similar tendency in 2009. Yet, as the recent financial crisis is still ongoing, its long-term consequences for the housing market remain unclear. Instead, this paper draws upon Japan’s Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c04.indd 58
2/8/2012 8:25:08 PM
Unemployment Risk, Homeownership and Housing Wealth
59
experience of housing boom (bubble) and subsequent burst in the late 1980s and early 1990s, respectively. We can obtain lessons from the bubble aftermath that Japan experienced in the 1990s in order to understand the impact of the subprime loan crisis on the housing market. In particular our purpose is to draw lessons from this experience in order to consider the potential impact of the current crisis on households with different housing tenure status. During the bubble period the average price of a house was about eight times the average annual income, which was beyond the reach of younger renter households. On the other hand, since housing equity is the important component of owner household wealth, the rise and fall of housing prices might well have had a substantial positive and negative effects on households’ consumption. The Japanese economy suffered from a surge in asset prices (housing and stock prices) during the bubble era (1987–1991) and subsequent sharp decline in prices after the bubble burst, which resulted in a decade-long persistent recession as well as deflation, usually referred as the ‘lost decade(s)’. In this study, we first provide an overview of key trends in housing and labour markets, the macroeconomy and so on during and after the bubble era. Second, we estimate the timing of homeownership of younger households applying a survival analysis approach to investigate the impact of the bubble burst on household tenure choice. We use the retrospective panel data of Keio University in Japan (KHPS). According to Moriizumi and Naoi (2011), income uncertainties expressed by unemployment probability and income volatility impact homeownership negatively. In fact, although housing prices and mortgage rates continuously declined throughout the 1990s after the bubble burst, young renters could not purchase a house due to a high level of unemployment risk. The estimation results suggest that the unemployment risk that a household expects to face decreases homeownership significantly, and this effect is especially serious for younger renters. Applying the estimation results, we conduct simulations under several scenarios using the actual unemployment rates and prices during and after the bubble period, where the effect of unemployment risk on homeownership prevailed over those of both housing price and mortgage rate. These results could be extended to the case of the recent subprime crisis, where the unemployment risk was extraordinarily high in historic terms since the end of World War II (since 1953) and it was exceedingly serious for younger households. Finally, we estimate a consumption function to assess the significance of housing and financial wealth. The estimation result shows that the wealth effect and the reverse wealth effect have a serious impact on consumption in Japan. Households facing greater income risks are less likely to own their home due to credit rationing. To the extent that public housing is scarcely provided in Japan they have no choice but to select private rental housing whose rental price is high for low-standard housing services. Given that rental housing is of inadequate quality in Japan, the subprime loan crisis and
Jones_c04.indd 59
2/8/2012 8:25:09 PM
60
Challenges of the Housing Economy
the subsequent deepest recession might well reduce household welfare seriously because a large number of have-not households emerge. Since especially young households have delayed the tenure transition from renting to owning under the recent economic deterioration, housing assistance toward younger households including perpetual renters and prospective owners must be considered. Besides these issues, declining housing prices after the subprime crisis impaired housing wealth seriously and the reverse wealth effect of housing wealth on household consumption is not negligible from the macroeconomic policy perspective. This chapter is organised as follows: the next section presents a brief description of the economic situation in Japan during and after the bubble. The following section examines the effects of housing and labour market conditions on the access to homeownership by young renters. The chapter then focuses on homeowners, examining housing wealth effects on household consumption. Finally, some general conclusions are drawn.
Housing bubble and burst in Japan Housing market and economy during the bubble era The Japanese economy experienced an asset price bubble during the late 1980s and early 1990s (broadly 1987–1991). Although there may be many factors that increased asset prices, the main factors were excessive liquidity caused by massive expansion of money aggregates and credit, and extremely optimistic towards the future (Shiratsuka, 2003). Relaxed monetary policy implemented by the Bank of Japan against the backdrop of a high saving ratio of households increased the money supply during the bubble period. Figure 4.1 shows that the rate of increase in the money supply is higher than 10% while the official discount rate declined to the historically low level of 2.5% at that time. Expansion of the money supply with the low discount rate resulted in excessive liquidity. A relatively high proportion of the loan finance provided by private banks was taken up by the real estate industry compared to other industries, which caused the large number of non-performing loans after the bubble burst. In fact, as Figure 4.2 indicates, a large amount of these loans was allocated to the real estate industry until 1998; therefore households could easily borrow from banks or non-bank private lenders. Owing to the cut in the official discount rate, the Government Housing Loan Cooperation’s (GHLC) mortgage rate came down (see Figure 4.1). However, because of the relative ease of access of funds to both households and real estate firms from private financial institutions (banks and non-banks) the share of GHLC in the housing loan market declined. During the bubble period the number of housing starts was the second highest since the end of World War II as Figure 4.3 shows. Furthermore the extremely
Jones_c04.indd 60
2/8/2012 8:25:09 PM
Unemployment Risk, Homeownership and Housing Wealth
61
12
14
12 10 10 8 8 Mortgage rate (private) 6
6
4
Mortgage rate (GHLC)
4
2 2 0
Official discount rate / Mortgage rates (%)
Money supply (M2, year-on-year changes, %)
Money supply
Official discount rate 0
19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10
–2
Money supply (M2)
Mortgage interest rate (private bank, ARM)
Official discount rate
Mortgage interest rate (GHLC, FRM)
Figure 4.1 Money supply and interest rates in Japan, 1976–2010. Soure: “Financial and Economic Monthly”, 1976–2010.
Outstanding of bank lending (1980 = 100)
600
500
400
300
200
100
19 8 19 0 8 19 1 82 19 8 19 3 84 19 8 19 5 86 19 8 19 7 8 19 8 8 19 9 9 19 0 91 19 9 19 2 93 19 9 19 4 9 19 5 9 19 6 9 19 7 98 19 9 20 9 00 20 0 20 1 0 20 2 03 20 0 20 4 05 20 0 20 6 07 20 0 20 8 09 20 10
0
Manufacturing
Construction
Real estate
Figure 4.2 Outstanding bank loans by industry in Japan, 1980–2010. Source: “Financial and Economic Statistics Monthly”, Bank of Japan. Notes: The data are not comparable before and after 1993 because of major changes in the industry classifications.
Jones_c04.indd 61
2/8/2012 8:25:09 PM
62
Challenges of the Housing Economy
Total housing starts (in 1,000 units)
2000
1500
1000
500
19
48 19 50 19 52 19 54 19 56 19 58 19 60 19 62 19 64 19 66 19 68 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08
0
Total housing starts
Built-for-sale homes
Figure 4.3 Housing starts in Japan, 1948–2009. Source: “Building Construction Survey”, Ministry of Land, Infrastructure, Transport and Tourism. Note: Housing starts for built-for-sale homes are not reported before 1951.
optimistic expectation towards the increase in asset prices pushed them upward sharply. Especially because of the ‘land myth’ (widespread since World War II) – the belief that land price would continue to rise prevailed among households and firms – they rushed into purchasing property (house or land), which again resulted in an extraordinary rise in land prices as well as housing prices. The average stock market Nickei index increased 2.07 times, while land prices rose 2.68 times in the six major cities during the 1986–1991 period, and average land prices grew only by 1.57 times (Figure 4.4). Slightly later, the stock price increased and land price rose sharply during these period. Figure 4.4 also shows that commercial land prices in six major cities increased markedly after 1986, peaking in 1991, and they were generally higher than those of local cities. A similar movement occurred in residential land prices. Since land price constitutes a large portion of housing price in Japan the average price of a house in Tokyo was about eight times the average annual income, which is beyond the reach of an average household, especially a young household in Japan. House prices had markedly and rapidly risen but income per worker had remained rather stable over a long time (Figure 4.5). However, this affordability gap varied across areas. As Figure 4.6 indicates, asset prices (land and stock prices) were increasing from 1986 to 1991, while the consumer price index (CPI) remained stable. This bewildered the Bank of Japan making it difficult to take an appropriate shortterm policy decision with the right timing.
Jones_c04.indd 62
2/8/2012 8:25:09 PM
Unemployment Risk, Homeownership and Housing Wealth
63
1800
Land and stock prices (1970 = 100)
1600 1400 1200 1000 800 600 400 200
19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 0 20 2 0 20 3 0 20 4 0 20 5 0 20 6 0 20 7 0 20 8 09
0
Commercial land price (nationwide) Commercial land price (six major cities) Nikkei stock average
Residential land price (nationwide) Residential land price (six major cities)
Figure 4.4 Trends in land and stock market prices in Japan, 1970–2010. Source: “Real Estate Related Statistics”, Mistusi Fudosan Co., Ltd.
1985 = 100 (constant 2005 prices)
150 140 130 120 110 100 90
90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08
19
88
89
19
19
86
87
19
19
19
85
80
House price
Household income
Figure 4.5 House prices and incomes in Japan, 1985–2008. Source: “Family Income and Expenditure Survey”, Statistics Bureau/“Annual Survey of Statistics on Construction”, Ministry of Land, Infrastructure, Transport and Tourism/ “Survey on Land Prices”, Ministry of Land, Infrastructure, Transport and Tourism.
Jones_c04.indd 63
2/8/2012 8:25:09 PM
64
Challenges of the Housing Economy
50 40
Year-on-year changes (%)
30 20 10 0
–10 –20 –30
19
8 19 0 8 19 1 8 19 2 83 19 8 19 4 8 19 5 86 19 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 95 19 9 19 6 97 19 9 19 8 9 20 9 0 20 0 0 20 1 02 20 0 20 3 0 20 4 0 20 5 0 20 6 0 20 7 08
–40
CPI
Residential land price
Nikkei stock average
Figure 4.6 Changes in inflation (CPI) land prices and stock market indices in Japan, 1980–2008.
Table 4.1
Housing tenure in Japan, 1978–2008.
Owned houses (%) Year
Total Age < 40
1978 1983 1988 1993 1998 2003 2008
59.9 62.0 61.1 59.6 60.0 60.9 60.8
40.7 42.0 36.9 28.8 25.8 26.6 27.8
Public rental housing (%) (by local government)
Public rental housing (%) (by public corporation)
Private rental housing (%)
Company houses (%)
5.3 5.4 5.3 5.0 4.7 4.6 4.2
2.2 2.2 2.2 2.1 2.0 2.0 1.8
26.0 24.3 25.7 26.3 27.3 26.7 26.8
5.7 5.2 4.1 5.0 3.9 3.2 2.8
Source: “Housing and Land Survey”, Ministry of Land, Infrastructure, Transport and Tourism.
In Japan, the government housing policy has aimed to support a household home acquisition for several decades, which has resulted in a high rate of homeownership. As evidenced in Table 4.1, owner-occupation dominates the housing tenure in Japan. Owner-occupied housing constitutes a large portion of household wealth and it provides better quality housing services than rental housing in Japan. Homeownership also compensates for an insufficient social support system for the elderly because there is discrimination against them in the renting of housing accommodation. In contrast, policies aiming to improve the quality of rental housing or to increase the
Jones_c04.indd 64
2/8/2012 8:25:10 PM
Unemployment Risk, Homeownership and Housing Wealth
14
65
6
12 5
4
8 6
3 4 2
2
Unemployment rate (%)
Real GDP growth (%)
10
0 1 –2 0
19 6 19 8 6 19 9 7 19 0 7 19 1 7 19 2 73 19 7 19 4 7 19 5 7 19 6 7 19 7 78 19 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 90 19 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 00 20 0 20 1 0 20 2 0 20 3 0 20 4 05 20 0 20 6 0 20 7 0 20 8 09
–4
Real GDP growth
Figure 4.7
Unemployment rate
Real GDP growth and unemployment rate in Japan, 1968–2009.
overall stock of public housing is limited. This finding implies that public housing may not work as a safety net in Japan. The government boosts homeownership by providing housing loans directly to home buyers with a lower rate and easier terms than the market. The GHLC has played this role for more than half a century. The mortgage market has therefore been dominated by public corporations and subprime lending was rare. GHLC, and the current Japan Housing Finance Agency (JHFA), has recently changed its role to securitise housing loans originated by private lenders.
Bubble burst and the prolonged recession In 1990 the Bank of Japan shifted its monetary policy from relaxed to tight, together with raising the official discount rate to 6%. The government then introduced land value tax in 1992 to depress the total amount of money flowing into real estate firms. In addition, a property tax was also raised. Besides these measures, the introduction of consumption tax in 1989 dampened housing construction and overall economic activities. Housing starts plunged 19.4 per cent in 1991 and although they subsequently rose and fell cyclically they did not rise above the 1990 level again as is shown in Figure 4.3. After the bubble collapsed, Japan suffered stagnant economic growth and a series of recessions for more than 10 years (the so-called ‘lost decade’). Japan is still suffering from an economic deterioration (Hayashi and Prescott, 2002). The rates of economic growth during this period (1991–2000) are shown in Figure 4.7. They were much lower than the previous decade, averaging 1.1
Jones_c04.indd 65
2/8/2012 8:25:10 PM
66
Challenges of the Housing Economy
90% 1983
1993
2003
80%
Homeownership rate (%)
70% 60% 50% 40% 30% 20% 10% 0% Less than 25 to 29 25 years years
30 to 34 years
35 to 39 years
40 to 44 years
45 to 49 years
50 to 54 years
55 to 59 years
60 to 64 years
65 to 74 years
75 years and over
Age of householder
Figure 4.8 Homeownership rates by age of householder in Japan, 1983, 1993 and 2003. Source: “Housing and Land Survey”, Ministry of Land, Infrastructure, Trasnport and Tourism.
per cent per year, and with several negative rates. The unemployment rate rose sharply after the bubble burst, in particular doubling for the youth (Figure 4.9). Figure 4.7 depicts the unemployment rate from 1968 to the present. Although the unemployment rate is a lagging indicator we can see the upward movement of unemployment rate in the 1990s. Despite low mortgage rates and the drop in housing prices after the bubble period (see Figures 4.1, 4.4 and 4.5), homeownership rate declined (Table 4.1) due to the rise in unemployment and fall in income and wealth. Furthermore Figure 4.8 exhibits a steep decline in the rate of homeownership for each age group and it especially shows a rapid decrease for households less than 35 years of age. During the period 1991–2001 household wealth, including both financial and housing wealth, decreased rapidly due largely to the sharp decline in land and stock prices as is shown in Figure 4.4, where the land prices declined monotonically and sharply, while the stock prices were highly volatile. The steep decline in wealth impacted household consumption negatively (a reverse wealth effect). Owing to the pervasive asset deflation and the sharp curtailment of bank lending under the critical erosion of the banking system following the credit crunch, the serious delinquent (and default) has increased much more rapidly than we have ever experienced. There are no details available for delinquent payments or mortgage defaults from private banks, but we know that there is a serious situation from the data of the GHLC. It is
Jones_c04.indd 66
2/8/2012 8:25:10 PM
Unemployment Risk, Homeownership and Housing Wealth
67
noticeable that although there is no subprime lending available in Japan, a large amount of non-performing loans have accumulated. The White Paper of 2010 (Annual Report on the Japanese Economy and Public Finance, Cabinet Office, Government of Japan) reports that Japan is still suffering from economic deterioration and it has consecutive ‘lost decades’. There may be several causes for the prolonged recession: the critical erosion of capital of the banking system, a decline in total factor productivity, a shortage of demand, the inadequacy of policy, sluggish government response, increased uncertainty about future prospect and so on (Ogawa and Wan, 2007). We focus on a shortage of demand, which has continued until the present and is a major driver for the deflation that has persisted over a long period until now (Figure 4.6). Horioka (2006) finds that the stagnation of private fixed investment, especially sluggish private housing investment, was a major culprit of the prolonged slowdown of the Japanese economy, and the decrease in household income and wealth, and increased uncertainty about the future were the causes of stagnant household consumption. To sum up, looking at the prolonged recession from the housing economic perspective, the following two points are worth noting. (i) The protracted unemployment delayed homeownership strongly even if housing prices and mortgage rates were lower, and (ii) the reverse wealth effect on household consumption cannot be negligible. Further explanations are as follows: (i) Despite decreased mortgage rates and housing prices, the effect of unemployment on housing purchase prevailed over the price-lowering effects, therefore homeownership was not easily attainable and housing investment declined strongly (rate of homeownership 61.1%→59.6%). This led to the recession, which in turn raised unemployment. A vicious circle occurred. As a result the homeownership rate, especially that of the youth who were seriously impacted by high unemployment, declined and the economy plunged into recession for a longer period than was anticipated. This effect is mainly related to tenant households, especially younger households. (ii) The rise and fall in asset prices influences the macro economy through wealth effects on consumption. During the bubble period the inflated wealth had a positive effect on household consumption, while after the bubble burst deflated wealth had a negative effect on it (a reverse wealth effect). The negative wealth effect depressed household consumption which in turn deteriorated the economy. Taking into consideration the fact that more than 60 per cent of households on average own their homes in Japan, this effect is important. Furthermore this is especially relevant for the elderly because more than 80 per cent of the elderly
Jones_c04.indd 67
2/8/2012 8:25:10 PM
68
Challenges of the Housing Economy
household own their houses. This effect is directly related to owner households but it also impacts on motivations for home purchase. While falling housing prices reduce the purchasing cost it induces some renters not to own because of the potential capital risk and acts as a drag on homeownership.
Unemployment risk and homeownership Empirical analysis In general, a household’s decision to own their home depends not only on the individual/household characteristics but also on the overall economic conditions. As discussed in the previous sections the Japanese housing bubble, which comprised a dramatic rise and fall in housing prices, influenced the homeownership decision, particularly the timing of homeownership, significantly. For example, in a rising market, prospective owners may postpone home purchase due to high housing prices, while in a declining market falling house prices may facilitate homeownership but other factors such as a high unemployment rate and a decline in real income could deter ownership as well. In the following analysis we investigate the timing of homeownership in Japan using a survival analysis approach, with a special emphasis on the role of unemployment risk and key housing market conditions. Our purpose is to identify the individual effect of macroeconomic factors (such as the unemployment rate, housing prices, mortgage interest rates etc.) on the overall homeownership in Japan, and to derive a relevant policy implication based on the Japanese housing boom and bust of the early 1990s. Our retrospective panel data tracks the same household for a fairly long period covering both the pre- and post-bubble eras; thereby it enables us to identify the long-term impact of housing boom and bust on the subsequent tenure choice. The data used in this analysis are taken from the Keio Household Panel Survey (KHPS) conducted by Keio University first in January 2004 and subsequently in 2005. The KHPS provides information not only on the current state of the individuals but also on the history of each individual’s characteristics starting from the age of 15. The individual history includes retrospective information pertaining to the individual’s employment status and previous housing status. Using this information, we construct a (person-year) retrospective panel data set in order to examine the timing of the first-time home purchases of Japanese households. There are several empirical studies that investigate the timing of homeownership using a survival analysis approach. Moriizumi and Naoi (2011) examine the effect of unemployment risk and income variability using a split population duration model – a variant of a two-component
Jones_c04.indd 68
2/8/2012 8:25:10 PM
Unemployment Risk, Homeownership and Housing Wealth
69
Table 4.2 Descriptive statistics and definitions of panel data variables. Variables
Mean
Baseline hazard Length of spell (in years) Demographic characteristics Age Married (1 = yes) Number of children Household characteristics Household wealth (predicted) Job-related characteristics Unemployed in year t (1 = yes) Ever been unemployed (1 = yes) Labour market characteristics Male unemployment rate Female unemployment rate Housing market characteristics Relative price of owner-occupied housing Mortgage interest rate
(S.D.)
Min
Max
9.344
(6.202)
1
31
28.813 0.577 0.644
(6.406) (0.494) (0.898)
18 0 0
53 1 5
269.888
(245.856)
0
1099
0.055 0.312
(0.227) (0.463)
0 0
1 1
3.238 3.166
(1.153) (0.956)
1.4 1.3
5.5 5.1
14.755 3.443
(5.098) (1.659)
2.332 −13.752
28.646 6.374
finite mixture model. Guiso and Jappelli (2002) examine the relationship between a private transfer and the timing of tenure transition while Deutsch et al. (2006) and Tiwari et al. (2007), focusing on wealth accumulation and liquidity constraints, estimate the duration until home purchase. Focusing on a sample of prospective first time buyers, we model the hazard of home purchase as follows: h(t|X it ) = L(φ(t) + α Pr(U i ,t +1 = 1) + X itβ ),
[4.1]
where t is the spell duration until a home purchase, L(·) is the logistic distribution function, f(t) is a parametric baseline hazard as a function of spell duration, and a and b are the parameters to be estimated. As for the observed covariates, Pr(Ui,t+1 = 1) represents the husband’s unemployment risk and is constructed in the following way. Our data provides the complete history of individual employment status. Using this information, we first estimate the probability that a male household head (husband) will be unemployed in the next year, and then use this predicted probability as an explanatory variable in equation 4.1. Xit includes person-specific characteristics such as marital status, number of children and the individual’s birth cohort, as well as regional and time-series characteristics such as house prices and mortgage interest rates. The household’s total wealth is also used as an explanatory variable and is estimated from our dataset. See Appendix A for the details of our explanatory variables. Table 4.2 presents the definitions and summary statistics of the variables.
Jones_c04.indd 69
2/8/2012 8:25:10 PM
70
Challenges of the Housing Economy
Table 4.3
Estimates of Timing of Home Ownership. Homeownership in year t (hit)
Outcome: Baseline hazard (length of spell, spline variable) 1 to 5 years 6 to 10 years 11 to 15 years 16 to 20 years 21 years and above Demographic characteristics Married Number of children Household characteristics Household wealth (in 10,000JPY) Unemployment risk Pr(Ui,t+1 = 1) Housing market characteristics Relative price of owner-occupied housing Mortgage interest rate log(Number of housing starts/population) Log-likelihood Number of observations
Coef.
(SE)
0.1423 0.1442** 0.1384** 0.1056** 0.0732**
(0.1188) (0.0467) (0.0321) (0.0246) (0.0200)
2.1838** 0.3882**
(0.3762) (0.0673)
0.0012**
(0.0003)
−1.4396*
(0.6625)
−0.0655** −0.2038** 0.4502
(0.0195) (0.0317) (0.2990) −1,196.613 8,388
Notes: Family characteristics are also controlled but omitted from the results. A set of dummy variables for birth cohort (with a 10-year interval) and regions are also included to control for cohort and region-specific effects. **, * and + indicate that the estimated coefficient is significant at the 0.01, 0.05 and 0.10 levels, respectively. Robust standard errors are presented in parentheses.
Empirical results The estimates of the discrete-time logit model of homeownership (equation 4.1) are presented in Table 4.3. From the estimation results we first find that a higher unemployment risk (probability) has a strongly negative impact on homeownership. It is demonstrated that on average a 10% increase in the unemployment probability decreases the annual probability of home purchase (i.e. the hazard rate) from 5.1% to 4.5%.1 The negative relationship shown in our estimation is consistent with the results of previous empirical studies such as Haurin (1991), Robst et al. (1999) and Diaz-Serrano (2005), which use income volatility as a measure of income uncertainty. With regard to the other variables, the following results are obtained: ●
●
Jones_c04.indd 70
wealthier households have a higher probability of homeownership, as expected; housing market characteristics such as the relative housing price (price of owner-occupied housing/rent) and mortgage interest rate have significant
2/8/2012 8:25:16 PM
Unemployment Risk, Homeownership and Housing Wealth
●
71
effects on homeownership with the expected signs, i.e. relatively expensive owner-occupied housing and a higher mortgage interest rate discourage households from purchasing a house; and marriage and a larger number of children promote housing purchases.2
To evaluate and express the results in a quantitative way, a simulation exercise is conducted in order to examine the long-term impacts of housing boom and bust and associated changes in labour market conditions on the access to homeownership by young tenants. Historically, the homeownership rate of young households in Japan has been held at a relatively low level. For example, the homeownership rate of young households (i.e. age of householder less than 40) was 28.8% in Japan (1993, see Table 4.1), compared to the USA where this rate was 45.5% (Current Population Survey/Housing Vacancy Survey, 1993). It is often pointed out that this gap can be partly explained by a relatively high level of housing price in Japan. The purchase price of typical built-for-sale housing is approximately 7.2 times the average worker’s pre-tax annual income in Japan (2010), while this ratio is about 4.0 in the USA (2007). This is not the whole story, however. Although housing prices steadily declined throughout the 1990s, the homeownership rate of young households continued to remain at a low level and even decreased further. While housing prices declined by more than 20% between 1993 and 2003, the homeownership rate slightly declined (28.8% in 1993 to 27.8% in 2003). Similarly, the real mortgage rate also decreased sharply during this period, and it too is supposed to promote homeownership (see Figure 4.1). Coincident with this housing price decline was the poor labour market conditions during the 1990s. As shown in Figure 4.9 the unemployment rate rose drastically during the 1990s, reaching a historically high level in the early 2000s. Such an upward trend was particularly marked for young workers. Altogether for these observations to be consistent, it is quite natural to think that the rising unemployment rate and the resultant income uncertainty restricted access to homeownership among young tenants. To examine whether this is the case, predicted homeownership probabilities under various scenarios are calculated for a typical household based on our empirical results.3,4 The results are summarised in Figure 4.10. Our benchmark case assumes that all macro-indicators (i.e., housing price, housing starts, mortgage interest rate and male and female unemployment rates) are fixed at their 1990 levels. In comparison, the probabilities of homeownership in other scenarios are based on the observed actual values of each macro-indicator. The long-term impact of each macro-indicator is evaluated by the difference between the benchmark homeownership probability and
Jones_c04.indd 71
2/8/2012 8:25:16 PM
72
Challenges of the Housing Economy
10.0% 9.0%
Unemployment rate (%)
8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0%
70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08
19
19
68
0.0%
Less than 30 years 50 to 59 years
30 to 39 years 60 years and over
40 to 49 years
Figure 4.9 Age specific unemployment rates in Japan, 1968–2009. Source: “Labour Force Survey”, Statistic Bureau.
90% 80% 70% 60% 50% 40% 30% 20% 10%
19
85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05
0%
Fixed all macro variables after 1990: benchmark
Observed housing price
Observed mortgage rate
Observed male unemployment rate
Observed female unemployment rate
Figure 4.10
Jones_c04.indd 72
Estimated home ownership probabilities under various scenarios.
2/8/2012 8:25:16 PM
Unemployment Risk, Homeownership and Housing Wealth
73
250
(1985 = 100)
200
150
100
50
86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05
19
19
85
0
Real housing price Mortgage interest rate
Figure 4.11 1985–2005.
Male unemployment rate Female unemployment rate
House prices, mortgage rate and unemployment rates in Tokyo,
that in the respective scenario. Observed trends in macro-indicators are presented in Figure 4.11. In the benchmark case where all macro-indicators are fixed at their 1990 levels (i.e. at the peak of the bubble), homeownership probability of a typical household reaches 0.66 in 2005 (i.e. at the head’s age of 40). Compared with this case, we find that, holding other variables constant, male unemployment rate has a fairly large and negative impact on the homeownership during the post-bubble period.5 Our simulation shows that homeownership probability of a typical household in 2005 is only 0.45 with the observed unemployment rates, suggesting that a sharp increase in male unemployment rates in the post-bubble period reduces homeownership probability by more than 20 percentage points (i.e. 0.66–0.45). On the other hand, our simulation shows that falling house prices and mortgage rates during the 1990s significantly promoted homeownership. In the case of observed housing prices, the homeownership probability of a typical household in 2005 is about 0.74. Similarly, with the mortgage rates applicable in 2005, the homeownership probability is about 0.83. As a result, the observed declines in the housing price and mortgage interest rate increase the homeownership probability by about 8.6 percentage points and 17.4 percentage points, respectively, from the benchmark case in 2005. Hence, if unemployment rates had been held constant at the 1990s level, these two factors would strongly promote homeownership among young
Jones_c04.indd 73
2/8/2012 8:25:17 PM
74
Challenges of the Housing Economy
tenants. In reality, however, these positive effects are outweighed by the poor labour market conditions in the 1990s, resulting in a virtually stable homeownership rate among young households.
Housing wealth and consumption Empirical analysis The relationship between household wealth and consumption has been an important topic in macroeconomics and finance. In this line of research, it is found that unexpected wealth-shocks change the permanent income of households and therefore affect the life cycle pattern of savings and consumption. More recently, the topic received renewed attention by housing economists focusing on the housing wealth effect on consumption. Obviously, the empirical analysis of housing wealth effects has an important policy implication. As discussed earlier, land prices in Japan showed considerable fluctuation from the 1980s, increasing by more than 300 per cent during 1983–1991, and then falling throughout the 1990s and the 2000s. Housing wealth effects have therefore important policy implications for the prolonged slowdown of the Japanese economy. While there has been a large body of literature, focusing mainly on US and UK markets, empirical studies do not seem to have reached a consensus on the existence and magnitude of the housing wealth effects. Of the US studies, Bostic et al. (2009), using matched datasets based on Survey of Consumer Finance and Consumer Expenditure Survey, find that there are relatively large housing wealth effects. They estimate that the elasticity of total consumption with respect to housing wealth ranges from 0.042 to 0.060, whereas the elasticity with respect to financial wealth ranges from 0.018 to 0.024. In comparison, Levin (1998), using micro data from the Retirement History Survey, finds no effect of housing wealth on consumption for the elderly. Of the UK studies, Campbell and Cocco (2007) find that the housing price elasticity of consumption varies with age of the household head and housing tenure; the largest house price elasticity is estimated for older homeowners, while the smallest elasticity, not significantly different from zero, is estimated for younger tenants. Disney et al. (2003), focusing on saving behaviour, find that the marginal propensity to consume from real housing gains ranges from 0.09 to 0.14 for the average UK household. They also find that the elasticity of consumption is likely to be different in rising and falling housing markets – considerably stronger effects of housing price shocks are found when house prices are rising. In comparison, there is only a limited empirical evidence for Japan. An exception is Hori and Shimizutani (2004) who find that the marginal propensity to consume with respect to
Jones_c04.indd 74
2/8/2012 8:25:17 PM
Unemployment Risk, Homeownership and Housing Wealth
75
house and stock prices are roughly comparable, and are estimated to be 0.05 to 0.10 for Japanese households. Our benchmark model of household consumption takes the following form: log Cit = β H H it + β F Fit + γ X it + ε it ,
[4.2]
where log Cit is a log of total consumption spending of household i in year t, Hit is the housing wealth, and Fit is the financial wealth. In our benchmark estimation, we use self-reported house value as the housing wealth measure, but we also use net home equity (i.e. self-reported house value minus mortgage outstanding) in the robustness checks. Xit includes various household and individual characteristics that are controlled for by the standard consumption models (e.g., age, marital status, family size, family structure, education, income etc.). In order to compare our estimates with those in the previous studies, we examine two alternative measures for the housing and financial wealth effects based on equation 4.2. One is the marginal propensity to consume (MPC) with respect to housing and financial wealth measures, and the other is the consumption elasticity with respect to housing or financial wealth. To test the robustness of our benchmark result, we also conduct a number of sensitivity analyses. Using total consumption expenditure as the dependent variable might suffer from a reverse causality problem. For example, expenditures for housing investments such as repair and renovation would increase the (self-reported) housing values, leading to a spurious correlation between total consumption and housing wealth measures. To cope with this problem, we also use non-housing expenditure as the dependent variable. For further robustness checks see Appendix B. In addition, empirical evidence from the US literature suggests that there seems to be an asymmetry of response between gains and losses in housing wealth. To see whether this is also the case for Japan, we will choose an alternative specification of equation 4.2 as follows. log Cit = β G I[ ΔH it ≥ 0] × H it + β LI[ ΔH it < 0] × H it + β F Fit + γ X it + ε it , [4.3] where I(·) is an indicator function taking the value 1 if the bracketed logical condition holds and 0 otherwise, and DHit is the changes in the housing wealth level from the previous year. Therefore, bG and bL represent consumption responses for gains and losses in housing wealth, respectively. We use household-level longitudinal data on consumption and wealth holdings in order to identify the household consumption response to the asset price movements over time. The KHPS provides longitudinal information on the household consumption as well as financial and housing wealth
Jones_c04.indd 75
2/8/2012 8:25:17 PM
76
Challenges of the Housing Economy
Table 4.4 Descriptive statistics of characteristic variables of home owners in 2004–2009 KHPS panels. Variables
Mean
S.D.
Min
Max
Monthly consumption (in 1,000JPY) Age of household head Married Years of education Number of household members Number of children Age 6 or under Age 7 to 12 Age 13 to 15 Age 16 to 18 Age 19 to 22 Household head’s employment status Not employed Self-employment Regular employment Household income (in 10,000JPY) Housing wealth (in 10,000JPY) Financial wealth (in 10,000JPY)
331.9 51.5 0.818 13.3 3.6
268.0 13.1 0.386 2.3 1.4
15.0 20 0 9 1
7234.3 80 1 24 12
0.124 0.300 0.134 0.133 0.172
0.410 0.648 0.377 0.381 0.451
0 0 0 0 0
0.170 0.148 0.524 721.0 2490.2 1088.0
0.376 0.355 0.499 477.3 2530.2 1912.7
0 0 0 0 0 0
3 5 3 3 4 1 1 1 8703.9 89730.8 27118.6
holdings. In the following analysis, six waves of the KHPS (2004–2009) are utilised to examine the housing and financial wealth effects on the household consumption.6 The KHPS provides detailed information about households’ housing and financial wealth holdings. As for housing wealth, the wealth measure is defined as net home equity (i.e. self-reported house value minus mortgage outstanding). This measure of house value is constructed from a question about subjective assessment of the value of current residence (‘How much do you think this lot/house would sell for on today’s market?’). Mortgage loan outstanding is also self-reported, and respondents are asked to provide the total outstanding at the end of the last year. As for financial wealth the precise form of question is ‘What amount of savings and deposits and securities does your household possess?’ Household’s non-housing debt outstanding is asked in a similar manner. As our focus here is to identify the housing wealth effects, we limit our sample to homeowners and descriptive statistics of the variables in the regression analysis are presented in Table 4.4.
Empirical results Our benchmark result of equation 4.2 is presented in the first column of Table 4.5. Coefficients for housing and financial wealth (bH and bF) are estimated to be 0.0069 and 0.0190, respectively, and are highly significant. Marginal propensities to consume (MPCs) are estimated to be 0.0027 for housing wealth and
Jones_c04.indd 76
2/8/2012 8:25:19 PM
Unemployment Risk, Homeownership and Housing Wealth
77
Table 4.5 Estimates of consumption and wealth effects.
Model dependent variable
[1]
[2]
[3]
log(total consumption)
log(non-housing consumption)
log(total consumption)
Coef.
Coef.
Coef.
(SE)
(SE)
(SE)
Age −0.0165** Age2 0.0002** Married 0.1354** Years of education 0.0211** Family size 0.0614** Number of children Age 6 or under −0.1230** Age 7 to 12 −0.0488** Age 13 to 15 0.0169 Age 16 to 18 0.0488** Age 19 to 22 0.1169** Household head’s employment status Not employed −0.0386* Self-employment 0.0089 Regular employment 0.0391** Household income (in 1 mil. JPY) 0.0357** Housing wealth (in 10 mil. JPY) 0.0069* Financial wealth (in 10 mil. JPY) 0.0190** Constant 4.9812**
(0.0178) −0.0372* (0.0175) 0.0065 (0.0150) 0.0431** (0.0023) 0.0359** (0.0028) 0.0076** (0.0030) 0.0188** (0.0876) 4.9020**
(0.0174) −0.0381* (0.0172) 0.0123 (0.0148) 0.0370* (0.0023) 0.0367** (0.0028) 0.0070** (0.0030) 0.0113** (0.0875) 4.9514**
(0.0179) (0.0178) (0.0152) (0.0023) (0.0027) (0.0032) (0.0887)
Test of b H = b F (F-stat/p-val)
(0.0050)
(0.0095)
(0.3452)
Marginal propensity to consume Household income (p-value) Housing wealth (p-value) Financial wealth (p-value) Elasticity Household income (p-value) Housing wealth (p-value) Financial wealth (p-value)
7.88**
(0.0034) −0.0157** (0.0034) −0.0160** (0.0034) (0.0000) 0.0002** (0.0000) 0.0002** (0.0000) (0.0171) 0.1326** (0.0167) 0.1381** (0.0173) (0.0022) 0.0226** (0.0022) 0.0222** (0.0023) (0.0044) 0.0641** (0.0043) 0.0608** (0.0044) (0.0135) −0.1362** (0.0133) −0.1195** (0.0137) (0.0080) −0.0541** (0.0079) −0.0470** (0.0081) (0.0114) 0.0171 (0.0112) 0.0168 (0.0116) (0.0124) 0.0442** (0.0117) 0.0468** (0.0125) (0.0108) 0.1229** (0.0107) 0.1146** (0.0110)
6.73**
0.1423 ** (0.0000) 0.0027 * (0.0145) 0.0076 ** (0.0000)
0.1372 ** (0.0000) 0.0029 ** (0.0075) 0.0072 ** (0.0000)
0.1461 ** (0.0000) 0.0028 ** (0.0080) 0.0045 ** (0.0005)
0.2577 ** (0.0000) 0.0171 * (0.0145) 0.0207 ** (0.0000)
0.2592 ** (0.0000) 0.0188 ** (0.0075) 0.0205 ** (0.0000)
0.2643 ** (0.0000) 0.0125 ** (0.0080) 0.0106 ** (0.0005)
Estimation Method
OLS
OLS
N
11,107
10,992
2
R
0.89
0.2726
0.2901
OLS 10,920
0.2688
Notes: Housing wealth measure used in the estimation is: self-reported house value in models [1] and [2], and net home equity (i.e. self-reported value minus mortgage outstanding) in model [3]. A set of dummy variables for regions, city sizes, survey years and housing types are included but omitted from the results. All monetary variables are deflated by the consumer price index. **, * and + indicate that the estimated coefficient is significant at the 0.01, 0.05 and 0.10 levels, respectively. Robust standard errors are presented in parentheses.
0.0076 for financial wealth, respectively. These results suggest that the wealth effect is substantially smaller for housing than for financial holdings. In fact, the null hypothesis that the estimated coefficients, and therefore MPCs, are equal is rejected (‘Test of bH = bF’ in Table 4.5). These results are
Jones_c04.indd 77
2/8/2012 8:25:20 PM
78
Challenges of the Housing Economy
in marked contrast with previous studies for the USA, where the wealth effects are larger for housing than for financial holdings (Bostic et al., 2009). There are several reasons explaining the relatively small housing wealth effect in Japan. First, housing wealth is intrinsically less liquid than financial wealth. Especially in Japan where there is relatively ‘thin’ second-hand housing market associated with high transactions costs, households are likely to feel that it is difficult to sell their house and move to another residence in order to release housing wealth. In fact, the five-year mobility rate in Japan is 24.1% (1998–2003), which is substantially smaller than the 50.4% in the USA (1995–2000). Furthermore, housing bequest motives are strong in Japan, and households may regard housing wealth as a non-marketable asset. The second reason is that home equity loans are extremely uncommon in Japan. Generally, the most popular way to release housing wealth is to sell the property or to refinance existing loans. Another option is to take out an equity loan. But the latter is uncommon in Japan, and it is somewhat difficult to liquidate one’s housing wealth in the short term. When we look at the elasticity estimates, however, we can draw a somewhat different conclusion with respect to the magnitude of wealth effects. Elasticity estimates presented at the bottom panel of Table 4.5 show that elasticity of total consumption is 0.0171 with respect to housing wealth, and is 0.0207 with respect to financial wealth. Compared with MPC estimates, elasticity estimates of housing and financial wealth are much more similar in magnitude, and this is simply due to the relatively large share of housing wealth within a household’s portfolio. This implies that the housing price movements have an economically significant effect on consumption, which is similar in magnitude to the financial asset price fluctuations. That is, 1 per cent changes in housing and financial asset prices will have a roughly equivalent impact on total consumption. In addition, the estimated income and financial wealth elasticities are comparable to previous estimates for the USA For example, Bostic et al. (2009) report that estimated financial wealth elasticities range from 0.018 to 0.024, which are quite similar to our benchmark estimate of 0.0207. Similarly, in their analysis, income elasticities are estimated to be between 0.162 and 0.236, which are somewhat smaller but still comparable to our estimate of 0.2577. These give us confidence that our key variables are capturing what we think they should. As for other variables, we also have results similar to previous studies. Among others, the age of household head has a U-shaped effect on total consumption, reaching its lowest impact at around the age of 50 (Campbell and Cocco, 2007). Family size has a statistically significant and positive impact on total consumption of 0.06, which is roughly equivalent to the US case (Bostic et al., 2009). Marriage and years of education are also both significantly positively associated with total consumption.
Jones_c04.indd 78
2/8/2012 8:25:20 PM
Unemployment Risk, Homeownership and Housing Wealth
79
Our main empirical findings are robust to a number of alternative specifications, including using alternative measures of consumption and wealth variables, excluding outliers, and controlling for sample selectivity. In model 2 of Table 4.5, we use non-housing consumption expenditure as an alternative dependent variable instead of total consumption. As discussed earlier, this is to cope with potential reverse causality running from the housingrelated expenditures to self-reported house values. The estimated results are virtually unchanged and we think that using total expenditure does not pose serious problems in our estimation. In model 3 we use alternative measures of housing and financial wealth variables. Net home equity (i.e. self-reported house value minus mortgage outstanding) is used in the estimation, and net financial wealth is defined in a similar way (i.e. financial asset minus non-housing debt). Even in this case, we find that the estimated coefficients on housing and financial wealth measures are highly significant and positive. Furthermore, the estimated coefficient on net home equity is almost identical to our benchmark case. However, the estimated coefficient on net financial wealth is now much smaller than our benchmark case. Consequently although the estimated coefficient is still larger for financial wealth than for housing wealth the null hypothesis of identical estimated coefficients cannot be rejected in this case. Finally, in Table 4.6, we examine an asymmetry of consumption response between gains and losses in housing wealth (equation 4.3). The result clearly shows that there are asymmetric consumption responses. Estimated coefficient on net housing wealth is substantially larger for households with losses in housing wealth than those with gains. Engelhardt (1996), using the US Panel Study of Income Dynamics, presents the same results about the asymmetric consumption responses. On the other hand, the result that we observe, a significant consumption response to decreases in housing wealth, is in sharp contrast with the evidence presented by Case et al. (2005) and Case and Quigley (2008). For example, Case et al. (2005), based on the aggregate data and using realised capital gains/losses as explanatory variables, report that the estimated effect of increases in housing wealth upon consumption is large and highly significant, while that of decreases in wealth is uniformly small and insignificant. There are several explanations for this discrepancy. First, different measures of capital gains/losses can explain the discrepancy. Sellers’ loss aversion leading to ‘sticky prices’ is often observed during a housing downturn (Genesove and Mayer, 2001). In this case, changes in transaction prices can be more modest than that in consumers’ perceived price levels. Therefore, our results based on perceived (self-reported) wealth changes would show much clearer consumption responses to capital losses. The fact that the US mortgage market is dominated by fixed-rate, self-amortising mortgages gives another explanation. As Case and Quigley (2008) argue, increases in
Jones_c04.indd 79
2/8/2012 8:25:20 PM
80
Challenges of the Housing Economy
Table 4.6
Estimated asymmetrical wealth effects on consumption. [7] log(total consumption)
Model dependent variable Age Age2 Married Years of education Family size Number of children Age 6 or under Age 7 to 12 Age 13 to 15 Age 16 to 18 Age 19 to 22 Household head’s employment status Not employed Self-employment Regular employment Household income (in 1 mil. JPY) Gain × net housing wealth Loss × net housing wealth Net financial wealth (in 10 mil. JPY) Constant Test of b G = b L (F-stat/p-val) Test of b G = b F (F-stat/p-val) Test of b L = b F (F-stat/p-val) Marginal propensity to consume Household income (p-value) Gain × housing wealth (p-value) Loss × housing wealth (p-value) Financial wealth (p-value) Elasticity Household income (p-value) Gain × housing wealth (p-value) Loss × housing wealth (p-value) Financial wealth (p-value) Estimation method N R2
Coef.
(SE)
−0.0124** 0.0001* 0.1418** 0.0189** 0.0572**
(0.0044) (0.0000) (0.0217) (0.0027) (0.0056)
−0.1290** −0.0522** 0.0157 0.0549** 0.1184**
(0.0166) (0.0103) (0.0139) (0.0156) (0.0136)
−0.0245 0.0313 0.0368+ 0.0353** 0.0061+ 0.0161** 0.0155** 4.9789**
(0.0224) (0.0219) (0.0192) (0.0027) (0.0035) (0.0040) (0.0033) (0.1145)
4.9900* 3.3300+ 0.0100
(0.0256) (0.0681) (0.9265)
0.1426** 0.0025+ 0.0065** 0.0063**
(0.0000) (0.0841) (0.0001) (0.0000)
0.2598** 0.0056+ 0.0110** 0.0153**
(0.0000) (0.0841) (0.0001) (0.0000) OLS 7,151 0.2691
Notes: A set of dummy variables for regions, city sizes, survey years, and housing types are included but omitted from the results. All monetary variables are deflated by the consumer price index. **, * and + indicate that the estimated coefficient is significant at the 0.01, 0.05 and 0.10 levels, respectively. Robust standard errors are presented in parentheses.
mortgage interest rates during housing downturn discourage existing homeowners with cheap fixed-rate mortgages from selling their houses, which further strengthens the downward stickiness of prices. In Japan, and also the UK, adjustable-rate mortgages are more prevalent. In addition, the
Jones_c04.indd 80
2/8/2012 8:25:20 PM
Unemployment Risk, Homeownership and Housing Wealth
81
fact that their results are based on aggregate data, while ours are based on individual-level data, might explain the discrepancy. As a result, a significant consumption response to decreases in housing wealth has particularly important implications for the overall Japanese economy. As shown in Figures 4.4 and 4.5 land and housing prices in Japan have continuously declined throughout the 1990s and 2000s. Our empirical results suggest that these declines are likely to be one contributing factor in the decline in household consumption, resulting in a shortage of demand. Furthermore recent housing market trends indicate that housing prices declined sharply in 2009 due to the current financial crisis emanating from the US subprime loan problem. This could further result in sluggish consumption.
Conclusion The global financial problem triggered by the US subprime crisis was initially not serious for the Japanese economy but ultimately hit it hard. This has had a number of important implications for households, the housing market and the overall economy. In this chapter we provide insights from Japan’s past experience in order to understand the potential impact of the current financial crisis. The consequences of the housing boom (bubble) and subsequent burst in the late 1980s and early 1990s have been seen as a long-term dampening of the Japanese economy. The chapter addresses two distinct impacts of the housing boom and bust on the households with different housing tenure status. One is the unemployment effects on the timing of owning a house by young tenants. The other is the housing wealth effects on consumption spending focusing on homeowners. With these two empirical analyses, we show that households of different housing tenure types were differentially affected by the labour and housing market movements during and after the bubble period. In particular, tenants were influenced more by labour market conditions, while owners were affected rather by housing asset price movements. Specifically, with regard to the unemployment effects on homeownership, the following results are obtained by estimating the duration model of homeownership and by conducting a simulation exercise under several scenarios, based on retrospective household panel data. Our empirical results indicate that typical young first-time buyers suffered from rising housing prices in the bubble era and from higher unemployment rate in the post-bubble period. First, the unemployment risk that a household expects to face plays an important role in home acquisition. During the bubble period, due to the extraordinarily high level of housing prices, the homeownership rate among young households (age < 40) dropped significantly from 42.0% in 1983 to 36.9% in 1988. After the bubble period, although housing prices and mortgage
Jones_c04.indd 81
2/8/2012 8:25:20 PM
82
Challenges of the Housing Economy
rates were historically low the homeownership rate still fell slightly (28.8% in 1993, to 25.8% in 1998, and 26.6% in 2003). This suggests that, after the bubble burst, although most prices (including CPI) came down and the government implemented various policies aimed at promoting homeownership, access to homeownership was still difficult for young renters due to increased income uncertainties associated with weak labour market demand and a high rate of unemployment. The adverse effects of the rising unemployment rate on home acquisition was particularly evident for younger renters. As for the housing wealth effects, the analysis finds that consumption spending by homeowners is strongly influenced by the housing asset price movements after the bubble period. Even though financing consumption using a house as collateral is not common in Japan the negative housing wealth effect seriously impacted on household consumption, which in turn led to weak consumption demand and a prolonged recession. Furthermore it is worth noting that there is an asymmetry of housing wealth effects on consumption between gains and losses. The negative housing wealth effect is larger than that of the positive one in an absolute value, which also impedes the economic recovery seriously. This housing wealth effect relates to owner households, especially the elderly whose rate of homeownership is higher than 80 per cent in Japan. The elderly cannot tap their housing wealth in the phase of declining housing prices. Although the Japanese government provides several economic policies intending to increase owner occupation it is difficult to bolster the rate of homeownership. For instance, policies aimed at younger prospective owners through the relaxation of borrowing/credit constraints or a tax credit for home acquisition or mortgage tax reduction or mitigation of the gift tax have not been effective so far because the number of housing starts are now historically small. In addition, according to our empirical results, the most effective way to encourage younger tenants to own their houses is remove the risks stemming from the labour market. A joint policy is needed for boosting ownership. On the other hand, providing a well-functioning rental housing sector is another requirement by younger tenants who may not ever own a house. This policy may take a longer time until it has effect. Furthermore providing a safety net sufficient for low-income households and unemployed tenants is an urgent issue in Japan. It is also important to understand how deflation can adversely affect tenants, prospective owners and owners. Under a deflated economy where private rent remains at a low level the private rented housing market might suffer a shortage of supply even if rental housing were of adequate quality. As for owners, continuously declining housing prices impair their housing assets which are the largest wealth of households, and owners cannot tap their wealth effectively after retirement. Therefore, we consider that curbing deflation is also an important candidate for housing policy.
Jones_c04.indd 82
2/8/2012 8:25:20 PM
Unemployment Risk, Homeownership and Housing Wealth
83
Appendices: Details of statistical analysis A.
Unemployment risk and homeownership
The spell duration until a home is purchased is modelled using a survival analysis approach. Our retrospective panel data includes the following two types of households: (1) households owning their houses at the time of the survey (i.e. those who purchased their own houses at a certain time in the past – referred to as the owner sample) and (2) households living in rental housing at the time of the survey (referred to as the renter sample). The spell duration for the owner sample is defined as the husband’s age at the time of purchase minus his age at the time of completion of his full-time education. For the renter sample, the spell duration is defined as the husband’s current age minus his age at the time of completion of his fulltime education. As a result, the owner sample includes the period from the time of graduation from school (the beginning of the spell duration) to the time of home purchase (exit time), while the renter sample includes the period from the time of graduation to the year when the current survey was conducted (2005). Since we can only observe the current and previous housing tenure status, we limit our sample to households whose husbands were less than 40 years of age at the time of their home purchases (for the owner sample) or at the time of the survey (for the renter sample). This ensures that the owners are still living in their first house and that the renters are not previous homeowners. As a result, the data contain a sample of prospective first-time buyers who are married couples. Details of key variables in our analysis are discussed as follows. After eliminating samples with missing or inconsistent information, we have 8,388 usable observations including 550 households.
Dependent event of home purchase The dependent event – the household’s home purchase – is defined as a dummy variable that takes the value of 1 if a household purchases its own house, or 0 if it continues to live in rental housing. A house to be purchased can be either a detached house or a condominium. Houses acquired through inheritance from parents or other relatives are excluded from the event.
Unemployment risk In addition to the housing purchase, our data also provide the complete history of individual employment status starting from the age of 15 for both the respondent and his/her spouse. Hence we can identify from the data
Jones_c04.indd 83
2/8/2012 8:25:20 PM
84
Challenges of the Housing Economy
Table 4.A1
Logit estimates of unemployment. Unemployment in year t + 1 (Ui,t+1)
Outcome: Baseline hazard (length of spell, spline variable) 1 to 5 years 6 to 10 years 11 to 15 years 16 to 20 years 21 years and above Demographic characteristics Married Number of children Household characteristics Household wealth (in 10,000JPY) Job-related characteristic Unemployed in year t (1 = yes) Ever been unemployed (1 = yes) Housing market characteristics Relative price of owner-occupied housing Mortgage interest rate log(Number of housing starts/population) Labour market characteristics Male unemployment rate Female unemployment rate Log-likelihood Number of observations
Coef.
(SE)
0.2428** 0.1774** 0.1970** 0.2273** 0.2414**
(0.0948) (0.0569) (0.0421) (0.0350) (0.0296)
0.6984** 0.0290
(0.2482) (0.0644)
0.0033**
(0.0005)
4.6765** 0.6039**
(0.3185) (0.1805)
−0.1810** 0.4358** −0.1501
(0.0268) (0.1117) (0.2952)
2.4013** −3.5058**
(0.5282) (0.6356)
−1,103.235 8,388
Notes: Family characteristics are also controlled but omitted from the results. A set of dummy variables for birth cohort (with a 10-year interval) and regions are also included to control for cohort and region-specific effects. **, * and + indicate that the estimated coefficient is significant at the 0.01, 0.05 and 0.10 levels, respectively. Robust standard errors are presented in parentheses.
whether or not a respondent/spouse was unemployed at each time period, t, in the past. Using this information, we estimate the probability of a male household head (husband) being unemployed at each point of time during his life course, conditional on the observed employment status, past unemployment history and a set of explanatory variables. The estimation results for the incidence of unemployment are presented in Table 4.A1. The results are as follows. Current unemployment as well as a past unemployment experience are significantly associated with a higher probability of being unemployed in the subsequent years. In addition, the incidence of unemployment among male household heads is positively associated with the unemployment rate of male workers, while it is negatively associated with the unemployment rate of female workers. Since in our dataset the household head for a married couple is assumed to be male, this result implies that there exists substitution between male and female
Jones_c04.indd 84
2/8/2012 8:25:20 PM
Unemployment Risk, Homeownership and Housing Wealth
85
workers. Using the estimation result, we construct the individual-specific unemployment probability (i.e., Pr(Ui,t+1 = 1) in equation 4.1) as a measure for unemployment risk.
Wealth accumulation As Jones (1995), Haurin et al. (1995) and Moriizumi (2003) noted, a household’s liquefiable wealth, which given the existence of certain down payment restrictions plays an important role in determining the housing purchase, is likely to be an endogenous variable. To cope with this endogeneity problem, an instrumental variable (IV) approach has been commonly used in previous empirical studies. We first estimate a wealth equation using the relevant set of exogenous variables and then use the predicted value of household wealth (WEALTH) in our model of home purchase (equation 4.1). Using a cross-section of the first wave, WEALTH is estimated by regressing the amount of total liquefiable assets (financial assets deflated by the CPI) on the husband’s age and education and a set of household characteristics. Since a sizable portion of the households in our sample do not hold any liquefiable wealth, the wealth equation is estimated by a Tobit regression of the sample living in rental housing in 2004. The results of the Tobit estimation are as follows: WEALTH = − 660.82 + 12.959 ⋅ AGE + 207.75 ⋅ HIGH (294.61)
(164.75)
(7.006)
+ 140.96 ⋅ JCOL + 527.94 ⋅ COL + 271.00 ⋅ OTHER (204.72)
(167.00)
[4.A1]
(213.68)
− 88.321 ⋅ EVERUE + 0.121 ⋅ Po , (81.946)
(0.063)
where the standard errors are in parentheses. The definitions of the explanatory variables are as follows. AGE is the age of the husband. HIGH, JCOL, COL and OTHER are dummy variables for the level of the husband’s education, indicating that the respondent has graduated from high school, junior college, 4-year college or higher and other educational institutes (such as vocational school etc.), respectively. Junior high school is omitted as a reference category. EVERUE is a dummy variable that takes the value of 1 if the husband has ever experienced unemployment. PO is the unit price of owneroccupied housing (prefecture-level).
Labour and housing market conditions As discussed in the previous section, our particular interest, in this paper, is to identify the impact of labour and housing market conditions on the household’s residential choice. As a primary factor representing the state
Jones_c04.indd 85
2/8/2012 8:25:20 PM
Jones_c04.indd 86
2/8/2012 8:25:22 PM
Test of b H = b F (F-stat/p-val)
Age Age2 Married Years of education Family size Number of children Age 6 or under Age 7 to 12 Age 13 to 15 Age 16 to 18 Age 19 to 22 Household head’s employment status Not employed Self-employment Regular employment Household income (in 1 mil. JPY) Net housing wealth (in 10 mil. JPY) Net financial wealth (in 10 mil. JPY) Constant (0.0123) (0.0083) (0.0126) (0.0125) (0.0107) (0.0161) (0.0157) (0.0134) (0.0010) (0.0017) (0.0019) (0.0780)
−0.1159** −0.0400** 0.0180 0.0510** 0.1033** −0.0280+ −0.0084 0.0006 0.0457** 0.0055** 0.0099** 5.0401** (0.1026)
(0.0030) (0.0000) (0.0144) (0.0020) (0.0038)
−0.0172** 0.0002** 0.1314** 0.0189** 0.0565**
2.67
(SE)
Coef.
log(total consumption)
[4]
[5]
[6]
Coef.
(0.0022)
(0.0182) (0.0181) (0.0158) (0.0027) (0.0030) (0.0044) (0.0891)
−0.0450* 0.0138 0.0387* 0.0385** 0.0074* 0.0252** 4.9684** 9.40**
(0.0140) (0.0082) (0.0118) (0.0126) (0.0111)
−0.1122** −0.0452** 0.0181 0.0499** 0.1166**
(0.0034) (0.0000) (0.0174) (0.0023) (0.0046)
(SE)
0.81
−0.0332+ 0.0145 0.0464** 0.0376** 0.0067* 0.0109** 4.9474**
−0.1178** −0.0464** 0.0103 0.0458** 0.1115**
−0.0173** 0.0002** 0.1270** 0.0243** 0.0595**
Coef.
(0.3690)
(0.0184) (0.0183) (0.0157) (0.0024) (0.0027) (0.0034) (0.0931)
(0.0156) (0.0085) (0.0121) (0.0130) (0.0112)
(0.0036) (0.0000) (0.0181) (0.0024) (0.0045)
(SE)
log(total consumption) log(total consumption)
−0.0150** 0.0001** 0.1317** 0.0189** 0.0600**
Estimation of consumption and housing wealth effects.
Model dependent variable
Table 4.A2
Jones_c04.indd 87
2/8/2012 8:25:22 PM
Quantile regression
Estimation method 10,504 0.2694
OLS
0.2720** 0.0113* 0.0203**
0.1512** 0.0029* 0.0099** (0.0000) (0.0125) (0.0000)
(0.0000) (0.0125) (0.0000)
9,877 0.2795
OLS
0.2709** 0.0122* 0.0105**
0.1500** 0.0027** 0.0044** (0.0000) (0.0141) (0.0012)
(0.0000) (0.0141) (0.0012)
Notes: Model [5] is estimated with a selected sample of households with their housing and financial wealth holdings within the range of mean ± 3S.D. Model [6] is estimated with non-mover households. A set of dummy variables for regions, city sizes, survey years and housing types are included but omitted from the results. All monetary variables are deflated by the consumer price index. **, * and + indicate that the estimated coefficient is significant at the 0.01, 0.05 and 0.10 levels, respectively. Robust standard errors are presented in parentheses.
10,920 0.1739
(0.0000) (0.0015) (0.0000)
0.3288** 0.0097** 0.0093**
N R2
(0.0000) (0.0015) (0.0000)
0.1818** 0.0022* 0.0040**
Marginal propensity to consume Household income (p-value) Housing wealth (p-value) Financial wealth (p-value) Elasticity Household income (p-value) Housing wealth (p-value) Financial wealth (p-value)
88
Challenges of the Housing Economy
of the labour market, we use gender-specific unemployment rates that are taken from the Labour Force Survey (Statistics Bureau). For measures of the housing market condition, we use the following variables. The mortgage interest rate is taken from the Financial and Economic Statistics Monthly (Bank of Japan). In the analysis, we use the average mortgage interest rate of city banks deflated by the CPI. The relative price of owneroccupied housing (RELPRC) is defined as the unit price of owner-occupied housing (PO) divided by the regional housing rent index (PR). The unit price of owner-occupied housing, PO, is calculated by the formula PO = PHλ PL1 − λ , where PH is the construction cost of housing unit per square metre of floor space, PL is the land price per square metre, and l is set at 0.5. The data on the construction cost are obtained from the Annual Survey of Statistics on Construction (Ministry of Land, Infrastructure, Transport and Tourism, MLIT). The data on land price are obtained from the Survey on Land Prices (MLIT). The data for both PH and PL are collected at the prefecture level. The regional index of housing rent (PR) is also a prefecture-level variable taken from the National Survey of Prices (Statistics Bureau of Japan).
B.
Housing wealth and consumption
In addition to a variety of statistical models presented in Table 4.5, we further conduct several robustness checks to our benchmark results. These results are presented in Table 4.A2. In models 4 and 5, we check whether outliers have any impact on our benchmark results. Previous studies suggest that the empirical results are highly sensitive to the choice of treatment of outliers. To cope with this issue, we estimate equation (4.2) by quartile regression which is considered to be much more robust to the outliers (model 4). More directly, we also test whether omitting outliers (i.e. those households with wealth measures out of mean ± 3SD range) from the estimation would change our benchmark results (model 5). These results show that our main conclusions are unchanged, with estimated coefficients on net housing and financial wealth remaining highly significant. Since mobility is a major source of equity release, housing and financial wealth measures should have discontinuous changes at the time of residential mobility. Disney et al. (2002) find clear differences in saving responses to house prices shocks between movers and non-movers among elderly households in the UK. To see whether such discontinuous changes drive our main results, in model 6 we estimate equation 4.2 with only non-movers. Again we find that housing wealth effect is of essentially the same magnitude as in our benchmark result.
Jones_c04.indd 88
2/8/2012 8:25:22 PM
Unemployment Risk, Homeownership and Housing Wealth
89
Notes 1. The baseline hazard of homeownership (5.1%) is calculated in the following way. First, we calculate the estimated hazard for each individual based on the observed unemployment probability. We then average this estimated hazard of each individual to obtain the baseline value. The estimated hazard based on the unemployment probability increased by 10% from the observed value (4.5%) and is obtained in a similar manner. 2. Since we cannot observe the past income stream from the data, we are forced to omit the measured income variable from our estimation. However, as Jones (1995) pointed out, accumulated wealth, rather than the permanent income level, is the major determinant of the housing acquisition decision. 3. By definition, homeownership probability is calculated as (1 – survival probability), which in turn depends on the predicted hazard obtained from equation (4.1). The survival probability at year t (i.e., the probability that a given household does not purchase their own house until year t) is given as S(t) = (1 − h(t) ) × S(t − 1) where h(t) is directly obtained from our estimation results presented in Table 4.3. 4. In our simulation a typical household head is defined as a college-educated male who is 22 years old in 1985, married throughout the period, has two children and lives in Tokyo. Although our simulation focuses on this specific case, we find that using other definitions of the typical household does not change the results fundamentally. 5. Our empirical results show that the male unemployment rate is significantly and positively associated with unemployment risk (Table 4.A1), and thereby a higher unemployment rate reduces homeownership probability. See Appendix A for details. 6. In this section we exploit the standard panel structure of the KHPS, whereas earlier we used it as a retrospective panel.
Jones_c04.indd 89
2/8/2012 8:25:23 PM
5 The Changing Nature of Household Demand and Housing Market Trends in China Edward C Y Yiu and Sherry Y S Xu
China, with its continuous and momentous economic growth in recent decades, became the second largest economy in 2010 with gross domestic product (GDP) of US$5.88 trillion in comparison with the USA at US$14.66 trillion. The economic (GDP) growth rate of China has consistently increased at about 10% per annum for about three decades. As the monetary system of China is relatively closed and isolated, the Chinese economy is not much affected by international financial crises. For example, Figure 5.1 shows that the economic growth rate slowed down’ between 1994 and 1999 and so led the Asian Financial Crisis in 1997. There was also a moderating of the economic growth rate in 2009 after the international financial crisis but it was still substantially positive at 8%. Rather than international influences the economic growth of China is strongly correlated with the domestic monetary policy as reflected in the inflation rate in Figure 5.1. China is a similar size (9.6 sq million km) in terms of area to the USA but the population size is four times bigger (1.3 billion in China and 0.3 billion in the USA, in July 2009). This chapter focuses on the most representative cities of the five regions in China: Beijing in the north, Shanghai in the east, Guangzhou in the south, Chongqing in the middle and Urumqi in the west to study the housing markets of China. There are more than 70 cities in China, a study of the five relatively prosperous cities among them is clearly Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c05.indd 90
2/7/2012 10:20:54 AM
The Changing Nature of Household Demand and Housing Market Trends in China
91
40%
35%
GDP per capita growth rate, China Inflation (GDP deflator annual %), China
30%
25%
20%
15%
10%
5%
0% 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
−5%
Figure 5.1 Economic growth and inflation of China, 1987–2010. Sources: China Statistical Yearbook 2010, and Mundi.com http://www.indexmundi.com/facts/china/gdp-deflator
a limitation in spatial coverage but it helps reduce the ‘noise’. Furthermore, the availability of market data in China is highly limited. The earliest official housing price indices of each city can only be traced back to 2009 from the China National Bureau of Statistics. Earlier data series of some cities may be available from other non-official sources but their trustworthiness is questionable. The chapter is structured as follows. The next section gives a brief overview of the development of the housing market and the evolution of housing policy in China. It also introduces the dramatic scale of recent house price inflation in the cities and explains government monetary policy toward dampening it down. The rest of the chapter assesses the potential reasons for the rise in house prices over the past decade including demographic factors, rising incomes and the role of capital markets and monetary policy.
Introduction to the housing market There were virtually no private housing markets in China before 1987 (after 1966), and housing was basically allocated within a public welfare system (Zhang, 1997, Lou et al., 2006), which resulted in building dilapidation and a heavy maintenance burden on the government. Fundamental change has
Jones_c05.indd 91
2/7/2012 10:20:54 AM
92
Challenges of the Housing Economy
Top tier: Private commodity housing market – provides housing at the market price or rent to the high income families.
Middle tier: Economic affordable housing – to be purchased by the mid- and low-earning families.
Lowest tier: Low-rent housing – provides rental housing to the lowestincome families.
Figure 5.2
Three-tier housing policy of China.
occurred as a result of the urban land reform since April 1987. Clause 4, Article 10 of the Constitution was amended in April 1988 to legalise the transfer of land use rights. The Land Management Law was also enacted in 1988 and the Provisional Regulations on the Conveyance, Granting and Transferring of the State Land’s Use Rights in Cities and Towns was introduced in 1991. Li (2002) describes the new institutional arrangement as the leasehold transferability of land use rights. In other words, land is still owned by the state, which is in line with socialist doctrine, but the land use rights are transferrable and tradable freely in the market. The land use rights expire after a designated time limit; for example, residential land use rights are normally for 70 years. This resembles the land grant system of leasehold interests in Hong Kong, and in fact the first trial of this land use rights transfer was in Shenzhen, which is the cross-border city with Hong Kong. However, as the state becomes one of the primary sources of land supply this creates a conflict of interest because constraining supply pushes up land prices and increases public income. The reform from a housing allocation system to a private housing market has generated a great demand for private housing, because the state no longer provides housing for its citizens. It has resulted in house price escalation and housing unaffordability problems. Since 1994, a three-tier housing policy has been introduced to provide affordable housing and low rent housing to people with relatively low incomes and the poorest as set out in Figure 5.2.
Jones_c05.indd 92
2/7/2012 10:20:54 AM
The Changing Nature of Household Demand and Housing Market Trends in China
93
This policy has evolved over time. The Department of Construction and Ministry of Finance promulgated the Urban Affordable Housing Construction Policing Method in 1994. The Notice of Further Deepening Cities Housing System Reform and Speeding Up Housing Construction was introduced in 1998. In 2004, the Administrative Measures for Economically Affordable Houses was promulgated to replace the 1994 policy. In the same year the Measures for the Administration of Urban Subsidised Housing for Families with the Lowest Income regulation replaced the 1999 Method of Cities Inexpensive Rental Housing Supervision. The result is that there are three segments or layers to the housing system with the government supporting the bottom two. Provincial governments solicit funds and allocate land to build low-rent housing units and economic affordable housing units directly. Low-rent housing is normally less than 50 square metres in size, and is owned by the government but rented to the lowest-income families. Economic affordable housing, however, is bigger in size, and is sold to the middle-income families at a price well below market prices. The government can also set a cap on the selling price of housing when the land use rights are granted to the developer to ensure affordability to housing to middle-income families. However, the provision of affordable housing and low-rent housing is lagging far behind the actual needs, because land sales have become one of the major incomes of local governments, and the free provision of sites for affordable housing affects their income. Moreover, most of the construction costs and the land cost are borne by local governments. The National People’s Congress reported that provincial governments failed to fulfil their construction goals for affordable housing in 2009. The central government has urged provincial governments to step up efforts to ensure the supply of affordable housing. The Ministry of Housing and Urban-Rural Development has signed documents in May 2010 with all the 31 provincial, municipal and autonomous region governments to guarantee the construction of this much-needed housing. Subsidised housing in China is still unlikely to meet the demand for housing, at least in the short to the medium run, which renders the private housing market sector almost the only supply of housing in China. Although the top-tier market is a private housing market the government is keen on implementing monetary measures to supervise the liquidity of the market. For example, the China Banking Regulatory Commission has required all bankers to strictly follow the down payment rules set out by the State Council.1 For example, the minimum down payment for first home buyers and the second home buyers of housing purchase are 30% and 60%, respectively. The mortgage rate for a second home buyer is also set at 1.1 times the benchmark interest rate. The state government has also
Jones_c05.indd 93
2/7/2012 10:20:54 AM
94
Challenges of the Housing Economy
3000
7.00% Shanghai second-hand housing price index (2003 = 1000, left axis)
6.00%
Monthly growth rate (%, right axis)
2500 5.00% 4.00%
2000
3.00% 1500 2.00% 1.00%
1000
0.00% 500 –1.00% –2.00%
0 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1
2003
2004
2005
2006
2007
2008
2009
2010 2011
Figure 5.3 Second-hand housing price index and monthly growth rate, Shanghai (2003 = 1000). Source: ehomeday, China, at http://www.ehomeday.com/cyfw/
been increasing the central bank benchmark interest rate, as well as the reserve requirement ratio of banks, in order to curb inflation and housing price increases. Housing prices in China have increased continuously by several fold in the past decade. For example, the housing price index of Shanghai has increased by 250% in the past 7 years, as shown in Figure 5.3. There are two recent peaks are in 2005 and 2007 when the monthly growth rate reached 4–6%. What are the causes of the housing price growth in China? There are many different hypotheses, such as the population growth hypothesis, the income growth hypothesis, and the monetary policy hypothesis. This chapter aims to assess the empirical evidence on these different alternative hypotheses.
Population growth hypothesis There are now more than 1300 million people and about 370 million households in China. The average household size is 2.91, and the per capita building space in urban areas is 27.1 sq m. The population has risen from just 580 million in 1953, through the decades to 690, 1010 and 1130 million in 1964, 1982 and 1990, respectively. The population growth rate before 1979 was more than 20% per decade on average. After implementing the
Jones_c05.indd 94
2/7/2012 10:20:55 AM
The Changing Nature of Household Demand and Housing Market Trends in China
95
US China
90–94 80–84
Age group
70–74 60–64 50–54 40–44 30–34 20–24 10–14 0–4 0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
% of population
Figure 5.4 Population pyramids of China and the USA, 2008. Sources: China National Bureau of Statistics, http://www.stats.gov.cn/tjsj/ndsj/2009/ indexeh.htm US Census, http://www.census.gov/popest/national/asrh/NC-EST2009/ NC-EST2009–01.xls
one-child (maximum) policy in 1978 the population growth rate has consistently decreased to about 15%, 10% and 5% on average in the decades to 1990, 2000 and 2010 respectively. Nevertheless the huge baby booms of the 1960s and the 1970s led to a sharp increase in housing demand from newly married couples in the past decade. There is therefore a demand-side hypothesis that the baby-boomers in China can explain the substantial housing price increases. This is supported by Figure 5.4 that contrasts the population distribution between China and the USA. The US population distribution (the top bars) shows a much more uniform distribution of population size (about 7%) for all age groups below 55 years old, but the population distribution in China (the bottom bars) is overwhelmingly dominant in the two age groups, 35–39 and 40–44. The age cohorts account for about 19% of the population, i.e. about 200 million people, and thus generate substantial new demand for housing units. However, population growth itself is unlikely to be a cause of the recent housing price escalation in China, because the housing supply increase in the past decade has been so enormous. For example, the floor space of newly built residential buildings in urban areas increases by an average of 600 million square metres per annum. It far exceeds the average (theoretical) demand of new residential floor space (about 350 million square metres per annum), as shown in Figure 5.5. The cumulative surplus in the past decade reaches 4000 million square metres. This oversupply phenomenon is also confirmed by Hou (2009) that, ‘the floor space of newly completed housing
Jones_c05.indd 95
2/7/2012 10:20:55 AM
96
Challenges of the Housing Economy
80.00 Cumulative actual new floor space supplied (100m m2)
New floor space (100m m2)
70.00
Cumulative theoretical new floor space demanded (100m m2) – estimated by population growth
60.00 50.00 40.00 30.00 20.00 10.00
08 20
07 20
06 20
05 20
04 20
03 20
02 20
01 20
00 20
99 19
98 19
97 19
19
96
0.00
Year
Figure 5.5 New floor space supply and demand in China, 1996–2008. Notes: The supply quantities are actual market supply reported in the source below. The demand quantities are estimated by multiplying the increase in the number of population in the age group of 15–64 to the average per capita building space in urban area in each year. It assumes that all the population growth is in urban areas, so this overestimates the demand. Source: China National Bureau of Statistics, Table 9–35 Floor Space of Newly Built Residential Buildings and Housing Conditions of Urban and Rural Residents, http://www.stats.gov.cn/tjsj/ndsj/2009/indexeh.htm
has exceeded that of housing sold each year from 1998 to 2006’, and Zhang (2001) indicates that this market imbalance existed even at the beginning of 1994, noting that, ‘the number of vacant new housing flats has risen dramatically’.
Income growth hypothesis An alternative demand-side hypothesis is that the real income increases of households as the economy has grown are the cause of the housing price boom. Support for this argument is found in Figure 5.6 that shows the strong growth of the nationwide average per capita disposable income (67% growth from 2004 to 2008, see the second line) in the past few years. It is even stronger than the nationwide average housing price growth (34% growth from 2004 to 2008, see the third line). Chow et al. (2008)
Jones_c05.indd 96
2/7/2012 10:20:55 AM
The Changing Nature of Household Demand and Housing Market Trends in China
97
250 GDP index of China Per capita disposable income index Housing price index Housing rental index
Indices (2004 = 100)
200
150
100
50
0 2004
2005
2006
2007
2008
Year
Figure 5.6 Nationwide average disposable income index, average housing price and rental indices, China 2004–2008 (2004=100). Notes: GDP of China in 2008 is US$4,401,600 million; per capita disposable income of urban households in 2008 is RMB 15, 780.8. Source: China National Bureau of Statistics, Table 9–2 Per Capita Annual Income; Table 8–17 Price Indices for Real Estate; http://www.stats.gov.cn/tjsj/ndsj/2009/indexeh.htm; GDP data from http://www.chinability.com/GDP.htm
also confirm the positive relationship between disposable income and housing price in an empirical study. However, it must be noted that the growth of household incomes has seriously lagged behind the growth of GDP (96% from 2004 to 2008, see the first line). In fact, the ratio of per capita income to GDP has been declining continuously, from 67% in the mid 1980s to 56% in 2006 (Aziz and Cui, 2007). This income disparity has significantly deteriorated, but there are also substantial regional differences in the ratio of housing price to household income. The disposable income from wages plus interest earnings from saving in some cities is unable to meet the rising standard of living. It is untenable to impute rising housing prices to the income growth in these cities. A further contradiction to this income growth hypothesis is the level of housing affordability as measured by the price-to-income ratio. In some cities, such as Shanghai and Beijing, it has recently been reported to be 20 or even higher2 which is remarkably high in comparison with the international norm, which is less than 5. Even if the housing quality improvement, such as bigger housing size, is taken into account (the per capita floor space of residential building in urban areas increased from 14 sq m in 1991 to 27 sq m in 20063), the price-to-income ratio in the largest cities is still much higher than the international norm.
Jones_c05.indd 97
2/7/2012 10:20:55 AM
98
Challenges of the Housing Economy
40 35 30 Y-o-Y % change
25 20 15 10 5 0 –5
3 4
5 6
7 8 2009
9 10 11 12 1
2
3
4
5
6
7 8 9 10 11 12 1 2 2010
–10 Year/Month National total
Shanghai
Chongqing
Beijing
Guangzhou
Urumqi
Xining
Figure 5.7 Annual housing price change, year on year, of six Chinese cities, March 2009 to February 2011. Notes: Data before March 2009 are not available. The sales price index data are year-toyear housing price change; level price information is not available. The cities considered in sales price indices are different from the cities considered in household income statistics, so direct matching is not possible. The chosen six cities in both databases are one of the closest matches, and are located in the six different regions. Source: Monthly sales price indices of buildings in 70 medium-large sized cities (90 sq m and below), National Bureau of Statistics, China; from http://www.stats.gov.cn/english/
Furthermore, housing rental growth is found to have lagged far behind income (7% growth from 2004 to 2008, see the fourth line), which indicates that the housing price growth is not driven by actual accommodation needs or excess demand reflected in the rent paid by tenants, but more likely by investment purposes. Following the logic of DiPasquale and Wheaton’s (1992) model as housing price is determined by its rental income and discount rate the cause of the recent housing price growth in China is more likely to be related to the monetary policy, including interest rate, inflation rate and a risk premium.
Regional differences Additional support for rejecting the income growth is given by the regional dimension. Regional differences in disposable income and housing price change are great. For example, Figure 5.7 shows the monthly housing price change of six major cities in China, 2009–2011. The maximum difference in
Jones_c05.indd 98
2/7/2012 10:20:55 AM
99
30000
25000
20000
15000
10000
5000
ia
t
N
in
gx
be Ti
ng ng ho C
ng ua G
qi
do
gh an Sh
ra ve la na N
at
io
ng
ai
0
ge
Annual per capita disposable income of urban households (RMB)
The Changing Nature of Household Demand and Housing Market Trends in China
Figure 5.8 Average per capita disposable income of households of six Chinese cities, 2008. Note: Shanghai and Chongqing are central administrative cities, Guangdong is a province; Tibet and Ningxia are autonomous regions. They are comparables because they are of a similar administrative level in China. Sources: Data from Table 9–15 Per Capita Annual Income of Urban Households by Region (2008), National Bureau of Statistics, http://www.stats.gov.cn/tjsj/ndsj/2009/ indexeh.htm
growth rate can reach 30%. The average of the maximum difference in growth rate is 16%. Figure 5.8 also shows the per capita annual disposable income of urban households of the national average and the six cities, which shows that the difference is almost 2.5 times. Unfortunately, the official data series of housing price index of each city are only available from 2009, and it limits this study to just a two-year period empirical test. Taking the average in China can therefore be very misleading where the range of economic performance in different provinces is extremely wide. For example, the eastern region produced 55.7% of national GDP in 2007, while the north-eastern, central and western regions only produced 8.5%, 18.7% and 17.1%, respectively. Similarly the eastern region accounted for 89.7% of foreign trade, whereas the north-eastern, central and western regions accounted for only 3.9%, 3.1% and 3.3%, respectively (Table 5.1). It is clear that a nationwide average does not reflect the actual situation of the individual regions, but the central monetary policy has to be applicable across the board, because of the same currency and the free flow of money within the country.
Jones_c05.indd 99
2/7/2012 10:20:56 AM
100
Challenges of the Housing Economy
Table 5.1 Percentages of national resources and economic output between the four regions of China, 2006.
Land Population GDP Foreign trade
Eastern
Northeastern
Central
Western
9.5 36.3 55.7 89.7
8.2 8.4 8.5 3.9
10.7 27.3 18.7 3.1
71.5 28.0 17.1 3.3
Source: China Statistics Yearbook 2007.
Monetary policy hypothesis The monetary policy hypothesis contends that the housing price increase in China is investment driven, where the investment opportunity comes from the combination of huge differences in regional economic performance and a centralised monetary policy. Hu et al. (2006) find – based on national aggregated empirical data – that macroeconomic fundamentals, such as GDP and the real interest rate, impose significant positive and negative effects on housing price respectively, but admit that ‘the housing markets vary a lot across different cities and provinces in China’. An aggregated analysis is therefore highly limited. In China, the investment preference for real estate, especially for housing, is overwhelmingly clear. Besides the Chinese culture of homeownership, there are at least two reasons for the investment preference: (1) great investment demand but lack of investment vehicles; and (2) a mismatch of credit supply to economic growth in different cities of China. These are now each considered in turn.
Great investment demand but lack of investment vehicles With the continuous economic growth of about 10% per annum in the past decades, the savings of households in China have greatly increased. Kuijs (2005) reports a steady 25% household saving rate (as per cent of disposable income) which is much larger than other OECD countries. For example, Qin and Ren (2008) found that, after adjusting all the other differences, the household saving rates in China was 17.5% higher than that in the USA. Shiller (2006) posits that the underdeveloped financial system and social security programmes in China, among other factors, explain why higher saving is necessary in China. Superficially, unlike in the USA, the saving interest rate has been relatively much higher than the inflation rate in recent years, and it does not seem to
Jones_c05.indd 100
2/7/2012 10:20:56 AM
The Changing Nature of Household Demand and Housing Market Trends in China
101
10 8
Inflation, China Three-month deposit rate
6 4 2 0 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 2008 2009 2010 2011 –2 –4
Figure 5.9 Inflation and interest rates, China, January 2008 to February 2011. Notes: The inflation rate is the CPI year-to-year growth rate. The interest rate is the three-month deposit rate per annum. Sources: Inflation data in China: http://www.tradingeconomics.com/Economics/InflationCPI.aspx?Symbol=CNY Interest rate data in China: http://www.tradingeconomics.com/economics/libor-rate. aspx?symbol=cny
have a negative interest rate effect driving a housing bubble. However, interestingly, Aziz and Cui (2007) point out that ‘the interest rate on household deposits has been capped by the government … China’s banks have enjoyed higher interest rate margins’. For example, the interest rate offered for three-month deposit in Dec 2008 was just 1.71%, in comparison with the 5.31% central bank overnight rate. Thus, as contended by Aziz and Cui (2007), ‘the rise in corporate profits did not translate into higher household income in China’. Figure 5.9 shows the inflation rates and interest rates of China from 2008 to 2011, revealing a negative real interest rate regime, except in the period from November 2008 to December 2009. The investment demand is plausibly believed to be driven by this negative real interest rate phenomenon. In fact, with the rapid declining of the ratio of per capita income to GDP (Aziz and Cui, 2007) the citizens in China envisage that their disposable income from wages plus interest earnings from saving will be unable to meet the rising standard of living they demand. Many households are therefore turning to investment opportunities. Investment demand in China is huge but the available investment vehicles are severely limited. The stock market and the bonds market are relatively small from an international
Jones_c05.indd 101
2/7/2012 10:20:56 AM
102
Challenges of the Housing Economy
8 National total Beijing
Y-o-Y % change of consumer price index
6
Shanghai Guangdong Chongqing
4
Tibet Ningxia
2
0 1
2
3
4
5
6
7
8
9
10 11 12
1
2
3
4
5
2009
6
7
8
9
10 11 12
1
2
2010
–2
–4
–6
Year / Month
Figure 5.10 Inflation rates of Chinese cities, January 2009 to February 2011. Sources: Inflation data in China are from National Bureau of Statistics, China: http:// www.stats.gov.cn/english/
perspective. It is also reported that household disposable income generated from investments (including profit, interest rate, etc.) is less than 8%, which is one of the lowest in the world. Worse still, relevant laws and governance on investment markets in China are major constraints. Housing, a traditional investment commodity, is therefore – in the light of the housing reform – one of the best investment choices in China.
Mismatch of credit supply and economic growth With the large variations in economic growth among regions the differences in inflation across China is huge. Figure 5.10 shows the discrepancies in the actual reported inflation rates in the seven cities of China of different regions. The maximum range and the average range from Jan. 2009 to Feb. 2011 are 5% and 3.3% respectively. For example, when Guangdong was in a deep deflation of negative 4% in mid 2009, the inflation rate in Tibet was positive and growing to the peak of 4% at the end of 2009, while Beijing was suffering slight deflation. Figure 5.11a shows the scatter plot of the housing price year to year percentage changes versus the inflation rate of these six cities in China.
Jones_c05.indd 102
2/7/2012 10:20:56 AM
The Changing Nature of Household Demand and Housing Market Trends in China
103
40
30
HPC, Beijing HPC, Shanghai HPC, Guangzhou HPC, Chongqing HPC, Urumqi HPC, Xining
20
10
0
–10 –6
–4
–2
0
2
4
6
8
Y-o-Y % change of inflation
Figure 5.11a Scatter plot of year-on-year percentage housing price changes versus inflation of six Chinese cities, March 2009 to February 2011.
Y-o-Y % change of housing price
25
20
15
10 y = 2.5587x + 6.6752 R2 = 0.6848
5
0 –3
–2
–1
0 –5
1
2
3
4
5
6
Inflation (%)
Figure 5.11b Scatter plot of national year on year housing price percentage changes versus inflation of China, March 2009 to December 2010.
Figure 5.11b plots that of the nationwide data. It shows a very strong positive relationship between the inflation with the housing price change in every city, but to different extents. The relationship between the nationwide averages shows a much larger change in housing price than that of inflation, the magnification effect is
Jones_c05.indd 103
2/7/2012 10:20:56 AM
104
Challenges of the Housing Economy
Table 5.2 Panel regression of real interest rate effect on housing price changes in Chinese cities, 2009–2011. Dependent variable: housing price change (year-to-year %), HPC Method: panel least squares Sample: March 2009 to February 2011 Period included: 24 Cross-sections included: 6 Total panel (balanced) observations: 144 Variable Constant, C Real interest rate, RI
Coefficient
Std. error
t-statistic
Prob.
8.028
0.551
14.569
0.000
−1.059
0.219
−4.838
0.000
Effects specification Cross-section fixed (dummy variables) Weighted statistics R-squared 0.319 Adjusted R-squared 0.289 SE of regression 6.612 F-statistic 10.675 Prob (F-statistic) 0.000
Mean dependent var SD dependent var Sum squared resid. Durbin–Watson stat.
8.016 7.841 5990.1 0.159
almost three times. It casts a question mark over the monetary policy hypothesis but this ignores the great difference in inflation between cities. As noted earlier, central monetary policy has to be uniform, because moneyflows between cities within China are free and at no risk, so the inflationhedging need among cities can vary substantially. For example, the three-month deposit rate was fixed at 1.71% from January 2009 onwards, resulting in a widely different real interest rate across cities, although there are some administrative measures and a bank loan quota system that are implemented on a regional basis. The inflation-hedging hypothesis can be tested by taking into account these location effects with a panel regression. Table 5.2 is a simple panel regression analysis showing the negative and significant effect of the real interest rate on the housing price changes in different cities, confirming the hypothesis of the investment driven housing price model. The panel is composed of data from six cities/regions/provinces, from March 2009 to February 2011, i.e. a six-cross-section times 24-month panel dataset. A fixed cross-section effect is adopted to allow different location effects. The real interest rate, RI, is found to exert a negative (−1.06) and statistically significant effect at the 1% level on housing price change. In other words, 1% change in the real interest rate imposes a 1.06% change of the housing price in the opposite direction, after taking into account the fixed effect of each city. It is a one-to-one inflation-hedging effect. The explanatory power of just one variable is relatively high at about 30%.
Jones_c05.indd 104
2/7/2012 10:20:57 AM
The Changing Nature of Household Demand and Housing Market Trends in China
105
The result of the regression analysis is limited by at least two factors: (1) only one independent variable and (2) the very short time period of data, due to the lack of available data. In view of the very short data series, lag effects cannot be considered, but it is commonly found in other countries that monthly price changes are most susceptible to interest rate changes. Other localised factors have been controlled by the fixed effect dummy, but other nationwide factors, such as down-payment regulation changes, cannot be controlled in the specifications. The empirical results (with the accepted limitations) imply that the recent housing price increase in China is caused by the negative real interest rate, which is the result of monetary policy. With the supply of RMB 4000 billion of new credit into the market and the currency peg to the US dollar after the 2008 crisis (until late 2010) the nominal interest rate dropped sharply and the inflation rate grew rapidly, especially in some more prosperous cities/ provinces/regions. The real interest rate in China has become more and more negative since March 2010. The latest nationwide inflation rate in March 2011 (the time of writing) had climbed to 5.4%, i.e. 7.2% increase from the bottom −1.8% in July 2009, but there is little change in the deposit rate. In view of the very limited investment vehicles in China the housing market has become almost a pure investment or even a speculative commodity market for inflation hedging. Local individuals and companies are the dominant investors of the housing market in China, because foreign investors are strictly controlled. However, land prices are bid-up not only by the private sector but also by the state-owned enterprises.4 They are found to be even more aggressive in their land investment, for as Wu et al. (2010) find, ‘they paid 27% more than other bidders for an otherwise equivalent land parcel’.
Conclusions The housing market and house prices in China are generally perceived to be driven by fundamental socio-economic-demographic reasons, such as a substantial increase in the disposable income (income growth hypothesis) in recent decades, the growing-up of the baby boomer (population growth hypothesis) since the 1960s, etc. However, this chapter puts forward evidence to refute these explanations, and argues that there is an investment-driven housing market boom – the monetary policy hypothesis. When other investment vehicles are limited, property investment becomes one of the best choices once transaction costs and risk are taken into consideration. It explains why housing price growth is several times stronger than the housing rental growth in recent years in China. It also sheds light on the incredibly high housing affordability
Jones_c05.indd 105
2/7/2012 10:20:57 AM
106
Challenges of the Housing Economy
index (price-to-income ratio) in China. Lastly, it tells why the oversupply of housing did not stop the housing price surge in recent years. Unfortunately, the hypothesis cannot yet be fully tested empirically and robustly because data in China, especially housing transaction price data, are not readily available. The global financial crisis in 2008 further confirms this monetary policy hypothesis. China is one of the countries with strong economic growth and a continuing rise in housing price after the crisis. The impact of the credit crunch on the Chinese economy and its housing markets was very limited and short-term. Following the injection in 2008 of RMB 4000 billion liquidity by the Chinese government into the economy it rebounded and began to overheat. Inflation and the housing bubble have become the biggest policy challenges in China since late 2010. Realising that the cause of the inflation and the housing bubble is liquidity-driven the Chinese government raised the required reserve ratio for banks more than 10 times, raised benchmark interest rates slightly and allowed the currency exchange rate to increase faster, from RMB 1 = US$ 0.146 in July 2010 to RMB 1 = US$ 0.154 in May 2011. It also increased the housing down payment percentage for second home purchases as well as charging property (ownership) tax in some cities. The regional economic performance in China is extremely unbalanced due to infrastructure and policy differences. As the China Economic Review (2011) reports, ‘the [nationwide] property index’s accuracy had been widely questioned in the past’, China’s National Bureau of Statistics stopped publishing its nationwide index of property prices in February 2011, and released separate data for the 70 cities that comprised the index. Colliers (2011) shows a parallel trend between the housing and office price increases in Beijing, Shanghai, Guangzhou, Shenzhen and Chengdu, which further strengthens the above monetary policy hypothesis. Central monetary policy, such as the interest rate and currency exchange rate, is uniformly applicable to the whole country. The real interest rate in individual regions can then be widely different with inherent economic problems and implications for the housing market. The liquidity-driven economy is also likely to be very fragile, especially one with long-term negative real interest rates. The over-concentration of bank loans in property markets can also be disastrous. A recent report by the Economist Intelligence Unit (2011) points out that, ‘China’s property sector is the main driver of the economic cycle. If residential investment slows in 2011, so will the Chinese GDP growth rate.’ As the monetary policy serves other macroeconomic purposes, such as curbing inflation and enhancing employment etc., the long-term trends of the Chinese housing market can be expected to be volatile and strongly linked with inflation if the financial system and social security policy remain unchanged.
Jones_c05.indd 106
2/7/2012 10:20:57 AM
The Changing Nature of Household Demand and Housing Market Trends in China
107
Notes 1. Notice of the State Council on Matters Relating to Further Regulating the Property Market, and Notice of the People’s Bank of China and the China Banking Regulatory Commission on Relevant Matters Relating to Improvement of Differentiated Housing Credit Policies, and Notice of the CBRC General Office on Improving Housing Financial Services and Strengthening Risk Management, 2011. 2. Property Investment Index 2010, http://www.numbeo.com/property-investment/rankings. jsp. Accessed August 2010. 3. Data from Table 9–35 Floor Space of Newly Built Residential Buildings and Housing Conditions of Urban and Rural Residents, National Bureau of Statistics, http://www.stats. gov.cn/tjsj/ndsj/2009/indexeh.htm 4. State-owned enterprises are corporations that are owned by the state, but are managed by the state-owned Assets Supervision and Administration Commission of the State Council.
Jones_c05.indd 107
2/7/2012 10:20:57 AM
6 Structural Sustainability of Homeownership in Australia Judith Yates
Unlike a number of western economies, Australia’s housing system appears to have survived the global financial crisis of 2007–2009 relatively unscathed. After a short-lived downturn (discussed further below), dwelling prices returned to the double-digit inflation experienced a few years earlier. A recent IMF paper shows Australia as one of just three of eighteen OECD countries where house price change was positive from 2007 to 2009 (Loungani, 2010), and over the past two decades, Australia’s growth in real house prices has exceeded that of many OECD countries (Tumbarello and Wang, 2010). These trends are detailed and updated in chapter 1. Although house price growth was widespread in most OECD countries from the mid 1990s to the mid 2000s cyclical peak, Australia’s higher than average growth rate over two decades arises from the fact that, unlike other OECD countries, it has not faced a major downturn in real house prices (André, 2010). Girouard et al. (2006) and Claessens et al. (2008) show that, in fact, Australia has not faced a major downturn in real house prices in the past 40 years or so. One result is that homeownership in Australia has not been under pressure from cyclical downturns. This chapter examines the possibility that sustained increases in real house prices, rather than volatility in real house prices, might create structural pressures on Australia’s homeownership rate over the next
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c06.indd 108
2/7/2012 10:20:42 AM
Structural Sustainability of Homeownership in Australia
109
30 to 40 years. It highlights the need for a definition of intergenerational or structural sustainability that goes beyond concerns with sustainability of homeownership only during cyclical downturns. Structural sustainability is defined as that which ensures that ‘the housing needs of the present generation can be met without compromising the ability of future generations to meet their own needs’ Yates et al. (2008, p1). This definition is borrowed from the 1987 report of the United Nation’s World Commission on Environmental Development (WCED) commonly known as the Brundtland report, after its chair. This report defined sustainable economic development as ‘development seeking to meet the needs of the present generation without compromising the ability of future generations to meet their own needs.’ (WCED, 1987, p49). On this definition, structural sustainability implies that future generations have access to homeownership under the same housing cost conditions in relation to their incomes as past and current generations. In other words, it requires access to homeownership to be intergenerationally equitable. This concern with intergenerational equity or structural sustainability of homeownership was motivated by the omission of any consideration of housing in the first Intergenerational Report (IGR) for Australia, released in 2002 (Australian Government, 2002) and in subsequent reports released in 2007 and 2010 (Australian Government, 2007, 2010). These intergenerational reports were designed, ‘to assess the long-term sustainability of current government policies over the forty years following the release of the report, including by taking account of the financial implications of demographic change’ (Australian Government, 2002: p iii) and to provide, ‘a comprehensive analysis of the challenges that Australia will face over the next forty years’ (Australian Government, 2010: p ii). In particular, the reports were designed to identify emerging issues associated with maintaining living standards of an ageing population. Omission of housing from these reports and, in particular, omission of an analysis of what might happen to homeownership over the next 40 years, is critical because of the role that homeownership plays in protecting the living standards of households as they age. The following section offers a brief overview of Australia’s housing system as it has developed in the post World War II era, of the policies that have contributed to current outcomes and of the impact that housing has on living standards of older and lower-income households. The next section provides projections which suggest that Australia’s current homeownership rate is not sustainable over the next 40 years. It also draws out and dissects the assumptions that have been made in reaching this conclusion. The final section briefly draws conclusions for future generations.
Jones_c06.indd 109
2/7/2012 10:20:42 AM
110
Challenges of the Housing Economy
Australia’s housing system Aggregate homeownership trends The system of housing provision in Australia is dominated by market provision of housing and, in particular, by private homeownership which grew rapidly in the immediate period following World War II and reached its current level of around 70 per cent by 1961. At the last census, around 70 per cent of private dwellings were occupied by their owners, around 25 per cent were occupied by households paying rent to a private landlord or living rent free. The remaining 5 per cent of dwellings were occupied by households paying rent to a public or not-for-profit landlord. As shown in Table 6.1, the homeownership rate has remained relatively unchanged since reaching its current level in 1961. The post-war growth of homeownership occurred during the long boom of the 1950s and 1960s when economic and demographic factors combined to generate a period of unprecedented economic growth and at a time when there was a severe post-war housing shortage. The willingness of households to borrow to finance their access to homeownership was underpinned by low interest rates, secure and rising incomes generated by strong economic growth and associated low unemployment. The ability to use borrowed funds to gain access to homeownership was underpinned by house prices that, for most households, were affordable in relation to incomes (as discussed below).
Table 6.1
Housing tenure in Australia, 1947–2006. Rented %
1947 1954 1961 1966 1971 1976 1981 1986 1991 1996 2001 2006
Owner-occupied %
Private
Social
53 63 72 71 69 68 70 70 72 69 70 70
40 31 24 22 22 21 21 21 22 22 23 24
4 4 4 5 6 5 5 5 6 6 5 5
Source: from 1947 to 1961, Bourassa et al. (1995: 85); from 1961, ABS census, special tabulations (tenure not stated allocated pro rata).
Jones_c06.indd 110
2/7/2012 10:20:42 AM
Structural Sustainability of Homeownership in Australia
111
The rapid expansion of homeownership in the 1950s was assisted by government providing mortgage finance at a subsidised rate of interest and by the sale of public housing to sitting tenants on extremely favourable terms. Since then, it has been supported by both explicit and implicit policies including direct subsidies in the form of generous grants to first-time home purchasers and even more generous favourable tax exemptions to established owners. Bourassa et al. (1995) provide an overview of the relevant policies that have supported homeownership in Australia in the period leading up to the most recent house price boom. Those implemented since the mid 1990s are covered in a government report on first-home ownership (Productivity Commission, 2004).
Market pressures Since the mid 1980s, however, this foundation of Australia’s housing system has been challenged, as a number of structural factors, both global and local, placed it under considerable pressure. Global trends have fuelled housing demand. Real household incomes have increased; disinflation has meant that nominal interest rates have fallen and borrowing capacity has increased; deregulation and financial innovation has meant that finance is more readily available; taxation systems generally have encouraged investment in housing. André (2010), Ellis (2006) and Girouard et al. (2006) provide excellent analyses of a number of these fundamentals as they relate to house price growth in the OECD countries since the 1970s. In Australia these demand pressures have been added to by high population growth (primarily as a result of high migration) and even higher household growth as a result of social and demographic change arising from the ageing of the population, a decline in, or deferring of, partnering, falling fertility and an increase in household dissolution. They have also been added to by the contribution that increases in real house prices have made to total household wealth, which has facilitated borrowing for both owner-occupied and investment housing by those who already owned their own homes (see for example La Cava and Simon, 2003, 2005). A number of studies have verified the importance of these factors and assessed the relative magnitude of their effects for Australia. Earlier studies highlight the role of income and demographic factors (for example, Bourassa and Hendershott, 1995; Abelson et al., 2005) while later studies focus on interest rate and wealth effects (for example, Otto, 2007 and Fry et al., 2010). Supply responses in Australia generally have been sluggish to these economic and demographic drivers of demand. The first two reports from the National Housing Supply Council (NHSC) estimate that currently there is a shortage of housing in Australia and project ongoing shortages over the next 20 years on the basis of comparing past trends in stock additions to
Jones_c06.indd 111
2/7/2012 10:20:42 AM
112
Challenges of the Housing Economy
demographic projections of underlying demand (NHSC 2009, 2010). They also provide considerable information on some of the long list of impediments to efficient supply. Among those listed as providing obstacles in the short to medium term are: labour shortages; shortages in, and/or high prices for, building materials; strategic and statutory planning processes; taxes and charges including developer contributions for infrastructure; lack of knowledge of, or confidence in, consumer preferences; the cost of, and access to, finance; and the presence of market power to maintain high prices (NHSC, 2010: p127). In addition to the conventional planning or regulatory reasons generally given elsewhere for supply constraints (for example, Barker, 2004 or Glaeser et al., 2005a) additional constraints arise in Australia from the concentrated nature of its urban settlement pattern, with 40 per cent of the population living in its two largest cities (Sydney and Melbourne), and almost two-thirds living in one of its six state capitals. These constraints are exacerbated by a concentration of employment opportunities in the centres of these cities, by metropolitan strategies built on increasing urban densities, by a lack of investment in basic transport infrastructure and by significant environmental constraints such as a lack of water. Ellis (2010) and Productivity Commission (2004) provide analyses of local factors that affect Australian housing markets. Although relatively little empirical work has been undertaken in Australia on supply elasticities, the lack of responsiveness of supply to demand pressures has increasingly been seen in official circles as a problem, most of which has focused on infrastructure issues and the cost of land at the fringe of the major cities. Richards (2009) provides one example from a Reserve Bank of Australia perspective. Kennedy (2010) offers a similar view from a Treasury perspective. In the longer run, however, it is the pressure on land prices imposed by increasing urbanisation that is likely to provide the main constraint on the responsiveness of supply. In their pioneering study of supply elasticity in the Sydney housing market from 1991 to 2006, Gitelman and Otto (2010) estimate both that the aggregate housing supply curve for Sydney is relatively inelastic and that supply elasticities vary within the city, increasing with distance from the centre (as population and housing density of local government areas fall). They compare their results with those reported by Green et al. (2005) for US cities and report that these indicate that Sydney mirrors a group of only six US cities that have supply elasticities less than unity (Gitelman and Otto, 2010). The study also finds strong evidence that the aggregate elasticity of supply declined between 1991–1996 and 2001–2006 and that the supply of multiunit dwellings is more elastic than that of separate houses. Estimates of changes in land supply gradients for both Sydney and Melbourne suggest that similar results would apply for Melbourne (Productivity Commission, 2004).
Jones_c06.indd 112
2/7/2012 10:20:42 AM
Structural Sustainability of Homeownership in Australia
113
350
Index 1975 = 100
300 250 200 150 100 50 0 1960
1965
1970
1975
1980
Real house prices Real house price trend
1985
1990
1995
2000
2005
2010
Real average earnings
Figure 6.1 Real house prices and earnings in Australia, 1975–2010. Source: Productivity Commission (2004), updated with ABS House Price Indexes, Cat No. 6416.0. CPI adjusted. Earnings data based on full-time adult average weekly ordinary time earnings from 1981 (RBA Statistics, Table G06); prior to 1981 derived from annual data on male average weekly earnings (RBA Occasional Paper No, 8A Australian Economic Statistics 194–50 to 1982/83.)
Demand trends, along with supply constraints, have contributed to the upward pressures on real house prices that have been observed in Australia for the past 50 years or so. As Figure 6.1 shows, real house prices began to diverge from real earnings as far back as the early 1970s and long before the 1990s increases in house prices induced by increases in the availability of mortgage finance, regarded as one of the major factors contributing to the global increase in real housing prices from the mid 1990s (André, 2010; Tsatsaronis and Zhu, 2004). In the 40 years to 2010, real house prices grew at an average rate of around 3 per cent per year. This is broadly in line with average growth in real GDP per capita (of 1.9 per cent per annum) plus population growth (of 1.4 per cent per annum) over the period (Australian Government, 2010) but well in excess of the 1 per cent increase in real average weekly earnings over the same period. In the 25 years since housing finance became more readily available as a result of financial deregulation, real house price growth increased to an average of 4 per cent. Between 2000 and 2010, real house prices grew at over 6 per cent per year and to a degree that the gap between prices and incomes has never been so wide. An increase in investor demand arising from 1999 changes to the tax system has also been identified by the then Governor of the Reserve Bank of Australia as a key factor contributing to the boom at the start of this period (Macfarlane, 2003).
Jones_c06.indd 113
2/7/2012 10:20:42 AM
114
Challenges of the Housing Economy
(%) 500 400 300 200 100 0 1960 –100
1970
1980
1990
2000
2010
–200 Deposit gap/annual average earnings * 100
Figure 6.2 Deposit gap measure of homeownership accessibility in Australia, 1975–2010. Source: House price data based on HIA/CBA Housing Report median first home prices from 1986; prior to 1986, based on data reported in Abelson and Chung (2005); interest rates based on standard variable mortgage rates (RBA Table F05); earnings data as in Figure 6.1. Borrowing capacity based on a 25-year loan and repayments set at 30 per cent of earnings.
Accessibility outcomes One critical outcome of this persistent increase in real house prices is the impact on access to homeownership for first-home buyers. Access to housing for first-home buyers depends on the relationship between house prices and borrowing capacity as determined by income, the cost of housing finance and lenders’ requirements. This gap between house prices and what households can borrow given their income and contemporary interest rates – the deposit gap – potentially provides a significant wealth constraint on those households’ ability to become homeowners. The gap can be compared consistently over time by taking the difference between the median house price bought by first-home owners in Australia and the maximum loan a first-home buyer on average earnings could borrow given a 25-year loan term, current bank mortgage interest rates and the standard prudential criteria applied to first-home buyers (repayments set at 30 per cent of income and a 20 per cent deposit requirement). Based on these calculations over the past 40 years there has been an increasing deposit gap faced by moderate-income households buying their first home, and this is shown by the index in Figure 6.2. Over time, this borrowing capacity constraint has been relaxed for some households. Housing loan calculators that provide indicative estimates of the maximum loan that can be borrowed generally require potential
Jones_c06.indd 114
2/7/2012 10:20:43 AM
Structural Sustainability of Homeownership in Australia
115
borrowers to specify their income, household size and existing loan commitments (Reserve Bank of Australia, 2005). In 2007, just prior to the global financial crisis, higher-income earners (on double average weekly earnings) and lower-income households with no children were permitted repayment ratios of up to 40 or 50 per cent. The 30 per cent ratio used in Figure 6.2 applied for a household on the equivalent of average weekly earnings, with a single income earner and two children. Borrowers have conventionally been required to provide a 10–20 per cent deposit (relaxed temporarily by the emergence of non-bank lenders prior to the onset of the credit crunch). In the past decade, first-home buyers have received deposit assistance of at least $7000 (increased to up to $35,000 in the stimulus package that followed the world banking crisis. They have also received concessional treatment in relation to transactions costs which, for non-first-home buyers, vary from 2 to 4 per cent for a median-priced property. Loans with lower deposits are required to pay mortgage insurance. In the 1950s and 1960s, when homeownership was affordable for moderateincome households, average weekly earnings were sufficient to support a loan that was more than adequate to purchase the Australia-wide medianpriced home. There was no deposit gap. By the 1970s, however, a significant deposit gap began to emerge as inflation and associated increases in nominal interest rates created front loading problems for home purchasers. Figure 6.2 also highlights the sensitivity of this deposit gap constraint on borrowing to cyclical changes in house prices and interest rates. The shortrun reduction in this constraint in 2008/09 reflects the combined impact of the temporary downturn in house prices associated with the onset of the international banking difficulties and the dramatic 400 basis point drop in interest rates that represented one of Australia’s policy responses to it. Figure 6.3 charts nominal and real interest rates over the period covered by this study. By the 1980s, the impact of high inflation and high interest rates meant that would-be first-home buyers with average earnings had to save the equivalent of one year’s income before the loan they could afford would be enough to purchase a median-priced home. The decline in inflation and the associated decline in nominal interest rates through the 1990s brought a considerable improvement in affordability for aspiring homeowners, but the house price boom from 2000 raised affordability problems to an all-time high. Households on average weekly earnings would have to save three or four times their annual income to meet the current deposit gap associated with median-priced dwellings. They would need to save even more once state-based transaction costs associated with home purchase were taken into account. The steady increase in interest rates from 2002 until late in 2008 represents an early response by the central bank to the increase in house prices that began in 1999 and was at risk of continuing. This response took some
Jones_c06.indd 115
2/7/2012 10:20:43 AM
116
Challenges of the Housing Economy
(%) 20 15 10 5 0 1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
–5
Nominal interest rate
Real interest rate
Figure 6.3 Real and nominal mortgage interest rates in Australia: 1960–2010. Source: Real interest rates defined as nominal rate less actual (CPI) inflation for the past year; data from RBA (Table F05 and Table G02).
of the heat out of the housing market prior to the onset of the global financial crisis. It meant that Australia’s housing market was not ‘overvalued’ as it might otherwise have been in 2008, and provides one of the explanations of why the subsequent downturn was not as severe as in most countries. Aspiring purchasers who do gain access do so by a variety of means. Increasingly, access to homeownership has depended on having two earners in the household with consequent repercussions on childbearing and family life. For moderate-income households one option is to purchase a dwelling that is cheaper than the Australia-wide median assumed in the affordability measure reported here, with associated implications regarding dwelling size and location. Richards (2008) shows that, in 2005/06, only 33 per cent of transacted dwellings Australia-wide would have been accessible to the median-income household in the potential first-home buyer 25–39 year old age group, compared with a long-run average of around 45 per cent. Only 3 per cent were affordable for low-income households in 2007/08 (COAG Reform Council, 2010). Increasing dwelling price gradients in each of the large cities (reflecting increased transport and time costs in gaining access to the centre) have meant that affordable dwellings are often poorly located. Both Hulse et al. (2010) and Wood et al. (2008), for example, show that the low levels of dwellings affordable for first-home buyers in Melbourne increasingly were located in fringe areas starved of public transport services and in areas where employment opportunities were relatively weak.
Jones_c06.indd 116
2/7/2012 10:20:43 AM
Structural Sustainability of Homeownership in Australia
117
For households on lower incomes who are unable to meet the deposit gap, none of the above strategies enables them to access homeownership. Single-income households in particular can find it especially difficult to overcome the homeownership threshold. Difficulties in buying have been accentuated by social changes such as later partnering among the young, and by more flexible household relationships (particularly when transaction costs make it difficult to make the transition from a two-income home-purchasing household to two one-income tenant households). They have also been accentuated by an increasingly flexible labour market with a greater reliance on fixed-term contracts, part-time work and a casualised workforce, all of which make it increasingly difficult for lower-income workers to commit to mortgage repayments over a long period of time. Finally, the reduction in inflation which brought about the general decline in interest rates in the past two decades has meant that initial high repayment burdens are sustained for a much longer period of time. Mortgages are no longer reduced in real terms by the combination of fixed mortgage repayments (as occurs with the conventional Credit Foncier Loan employed in Australia) and increasing nominal wages (as generally occurs rapidly with high inflation).
Housing system outcomes Declining affordability, in conjunction with these economic and social changes, has had a discernible impact on homeownership rates among younger households. Since the beginning of the 1980s, when the deposit gap increased from one to more than two years’ income, the proportion of owner/ purchaser households in the 30–34 year old age group and in the 35–39 year old age group has declined systematically, as can be seen in Figure 6.4. For younger 25–29 year old households, this systematic decline began in the mid 1970s. Hulse et al. (2010) provide a more detailed analysis of the changing characteristics of young lower-income households from 1980 to the present. However, Baum and Wulff (2003) show that there is considerable evidence that young Australian households still aspire to become homeowners despite the increasing difficulties that they face in meeting their aspirations. Declining homeownership rates among those approaching mid life raises important questions as to the life-long housing attainment of these cohorts. While some may be delaying homeownership, it seems unlikely that many will buy for the first time after reaching 40 years of age. These households will be moving from the younger to the middle to the older generation over the 40 or so years ahead envisaged in the intergenerational reports noted above. Their inability or choice not to become owners during their prime household formation years raises significant doubts about whether Australia’s homeownership rate will be sustained at its 50-year level of 70 per cent over the next 40 years.
Jones_c06.indd 117
2/7/2012 10:20:43 AM
118
Challenges of the Housing Economy
(%) 75
50
25
0 1961
1966 <25 years
1971
1976
1981
25–29 years
1986
1991
30–34 years
1996
2001
2006
All households
Figure 6.4 Homeownership rates by age in Australia, 1961–2006. Source: Special request tabulations from Census data. Age refers to the reference person in the household. As age breakdowns are not available for 1966, they have been interpolated from data for 1961 and 1971.
This has critical implications for the intergenerational sustainability of the Australian housing system. New generations of households will not be able to access homeownership on the same conditions as past generations and, indeed, may not be able to access homeownership at all. This raises questions about the capacity of homeownership to continue to act as a form of insurance that protects the aged against poverty in retirement because of the lower levels of housing costs associated with outright homeownership, as argued by Yates and Bradbury (2010). Yates and Bradbury (2010) also highlight the extreme polarisation of wealth by tenure and associated intergenerational inequities to which increasing real house prices have contributed. By increasing the capacity of those with housing wealth to obtain housing finance in order to add to their housing wealth, any such wealth inequalities are likely to add to the non-sustainability of Australia’s homeownership rate. Reduced and/or delayed access of even moderate-income younger households to homeownership imposes increased pressure on the rental sector. Given the small size of the social rental sector, this pressure falls heavily on the private rental market. A number of studies in Australia have pointed to the high burden that housing outlays place on lower-income households in the private rental sector. Of particular concern is the incidence of high housing costs in relation to income for the (still) relatively small number of older households who currently rent privately (see, for example, Yates and Gabriel, 2006).
Jones_c06.indd 118
2/7/2012 10:20:43 AM
Structural Sustainability of Homeownership in Australia
119
Future projections of homeownership sustainability Yates et al. (2008) have shown that, if age-specific homeownership rates for the young remain at their current levels and if the incremental increase as they age over the next 20 years (and then the next 20 years again) is no greater than it has been over the past 20 years, then Australia’s aggregate homeownership rate in 40 years’ time will have declined from 70 per cent to 66 per cent. This will occur despite the ageing of the population which, in the past, has contributed to a stable aggregate homeownership rate. If affordability constraints mean that young households currently excluded from homeownership are not able to ‘catch up’ with an increased rate of access to homeownership as they age, then homeownership rates among the aged in 40 years’ time will be even lower and the aggregate homeownership rate will fall even further. As a result, there will be an increase in after-housing poverty (that is, a decrease in their ability to meet their non-housing needs) among older households who are forced to rely on the private rental market to meet their housing needs. Future households, both young and old, will not be able to meet their housing needs under the same housing cost conditions in relation to their incomes as past and present generations. This pessimistic projection that, under current conditions, Australia’s homeownership rate is structurally unsustainable is based on a number of key assumptions. The first is that real house prices will continue to increase faster than household incomes with the result that, given the current cost of housing finance and current lending standards, deposit gaps for future generations of first-home buyers will continue to be significantly above those faced by earlier generations. This outcome will arise because of the second key assumption, namely that ongoing demand pressures will be met by an increase in housing supply that is less than perfectly responsive. A third assumption is that the structural changes that reduced the cost and increased the availability of housing finance that took place in the mid to late 1990s have been fully worked through. This implies that, over the next 40 years or so, mortgage interest rates will be maintained at an average real value of around 5–6 per cent over the cycle. It also implies that the innovations in housing finance that led to an easing in lending standards are more likely to become less rather than more liberal. Two final assumptions are associated specifically with tenure. They are that economic factors will continue to favour homeownership for those who can obtain access. A final assumption is that the private rental market will expand to house the growing proportion of the population unable to access homeownership. Each of these is expanded on below.
Jones_c06.indd 119
2/7/2012 10:20:43 AM
120
Challenges of the Housing Economy
House price fundamentals and real house price growth During the long house price boom that began in the mid 1990s much was written in both Australia and elsewhere questioning the sustainability of real house prices at what was proving to be above average growth rates and at rates which exceeded the growth in income. The question of whether housing prices in Australia have been above their fundamental values for much of the past decade, or the extent to which they have been above their fundamental values, is one about which there is little agreement. Varying assessments have been made. There appears to be some agreement that, at the peak of the 2003/04 boom, there was some degree of overvaluation (Fry et al., 2010; Bodman and Crosby, 2004) but less agreement as to whether post-2008 house prices are overvalued. As indicated above, this agreement about values in 2003/04 was sufficient for the Reserve Bank of Australia to express concern about speculative behaviour in housing markets and to increase interest rates in an attempt to discourage this. Bloxham et al. (2010) provide supporting evidence. A 2010 OECD reports suggested, as at 2008, a ‘certain degree of overvaluation could be present in Australia’ (André, 2010, p 18). On the other hand, in a series of reports between 2008 and 2009, the IMF changed their assessment of Australia having a house price gap (between actual house prices and the level that could be explained by fundamentals) in excess of 20 per cent (IMF, 2008: Box 3.1) to there being no evidence of a significant overvaluation of house prices in Australia (IMF, 2009). Consistent with the international literature on this issue, a common assumption of this literature for Australia has been that house price growth is sustainable if it is in line with fundamental factors of demand (driven by demographic factors and economic factors such as interest rates, expected inflation, expected house price inflation and taxes) and supply (driven by construction costs and land supply) and that, while there may or may not be some correction from their peak values at 2009/10, house prices will continue to increase from their trend 2000 levels – see, for example, Otto (2007) or Abelson et al. (2005). Counter views about sustainability tend to assume that a growing gap between incomes and house prices reflects a housing bubble and is therefore unsustainable. Himmelberg et al. (2005) provide an overview of some misperceptions about the sustainability of real house price growth based on simple measures such as trends in real house prices or price-to-income measures. In their view, these measures ignore many of the fundamentals described above. In particular, they fail to reflect accurately the state of housing costs: First, the price of a house is not the same as the annual cost of owning, so it does not necessarily follow from rising prices of houses that ownership is becoming more expensive. Second, high price growth is not evidence
Jones_c06.indd 120
2/7/2012 10:20:44 AM
Structural Sustainability of Homeownership in Australia
121
per se that housing is overvalued. In some local housing markets, house price growth can exceed the national average rate of appreciation for very long periods of time. Third, differences in expected appreciation rates and taxes can lead to considerable variability in the price to rent ratio across markets. Finally, the sensitivity of house prices to changes in fundamentals is higher at times when real, long term interest rates are already low and in cities where expected price growth is high, so accelerating house price growth and outsized price increases in certain markets are not intrinsically signs of a bubble. (Himmelberg et al., 2005, p68) Girouard et al. (2006) provide an explanation of how the unprecedented growth in global house prices from the mid 1990s could be explained largely by changes in fundamentals and, in particular, by reductions in interest rates associated with lower inflation and the development of new and innovative financial products. They suggest that, whether the higher house price levels that resulted from this higher rate of growth would be sustainable is likely to depend on whether longer-term interest rates remain at, or close to, the low levels that applied by the mid 2000s. In broad terms, increases in real house prices can occur if there are increases in demand, brought about by an increase in household incomes, growth in the number of households or a reduction in the user cost of housing – caused, in turn, by a reduction in (risk free) interest rates, an increase in taxes or, somewhat incestuously, an increase in expected house price inflation. The Intergenerational Reports in Australia that motivated this chapter’s concern with structural sustainability of homeownership contain demographic and economic projections that suggest that demand will continue to increase. In the most recent of these (IGR3), population growth is projected to increase at an average annual rate of 1.2 per cent over the next 40 years, only slightly lower than the 1.4 per cent average annual rate of growth in the previous 40 years (Australian Government, 2010). A number of reports have converted these population projections into household projections. The National Housing Supply Council, for example, using the same 2006-based population projections as used in IGR3, projects household growth of 1.6 per cent per year for the next 20 years (NHSC, 2010, Table A2.2). The Productivity Commission (2005) in its report on the ‘Economic Implications of an Ageing Australia’, using the slightly lower 2001-based population projections in IGR1 and IGR2, projects household growth of around 1.1 per cent over the 40 years to 2046. All projections highlight the ageing of the population over the period with the numbers of households in each 10-year age group increasing but with the proportion of households with a reference person aged 65 years or more increasing from 21 per cent in 2006 to 35 per cent in 2046 (with differences arising because of differences in assumptions about migration). All, however, still project absolute growth in the number of first-home entry-age households. This growth in the projected number of households will add to pressures on housing demand.
Jones_c06.indd 121
2/7/2012 10:20:44 AM
122
Challenges of the Housing Economy
Economic projections likewise keep pressure on demand. Compared with a growth rate of 1.9 per cent per year over the past 40 years, real GDP per capita is projected to increase at a slightly lower rate of 1.5 per cent per year over the next 40 years as a result of changes in the share of the working-age population and participation rates (Australian Government, 2010: p 14). Real wages are projected to increase at this same rate of 1.5 per cent. The combined effect of these projections is a projected growth in real GDP of 2.7 per cent per year over the next 40 years, marginally below the annual average of 3.3 per cent over the past 40 years. The house price modelling reported in Otto (2007) and Abelson et al. (2005) suggest, therefore, that the underlying growth in real house prices over the next 40 years will grow at an average annual rate of a little less than 2.7 per cent if lending and supply conditions are unchanged.
Supply Sustained demand pressures add to house price pressures only if the supply of housing is less than perfectly elastic. Indeed, Shiller (2007a, 2009) has argued against permanently increasing house prices on the grounds that, ultimately, land supply is not constrained, suggesting that ‘we do not really have a land shortage. Every major country of the world has abundant land in the form of farms and forests, much of which can be converted someday into urban land’ (Shiller, 2009, p1) or ‘[n]ew urban areas can be built on land that is now cheap. Cities can economise on land by raising the population density and building high-rises.’ (Shiller, 2007a, p98) Such solutions may be relevant in many countries although, ultimately, they are likely to result in growing income-based spatial sorting into what Gyourko et al. (2006) call ‘superstar cities’ as high-income households bid up the price of scarce land. Unlike many countries, Australia, as the driest continent on earth, does not have an abundant potential supply of urban land. A lack of water in much of the country is one reason why it is one of the most urbanised countries in the world, despite its large land mass. Over the next 40–50 years, the proportion of the population living in one of its (state) capital cities is projected to increase from its present value of 64 per cent to 67 per cent (Australian Bureau of Statistics, 2010). This will add to already existing pressures on land prices in these areas and will be reflected in the increasing land price gradients indicated in the previous section.
Lending conditions The homeownership projections for the next 40 years described above were based on the assumption that the borrowing capacity for future generations would be no better than those of the post-deregulation generations (which implies that they will be worse than those who became homeowners in the
Jones_c06.indd 122
2/7/2012 10:20:44 AM
Structural Sustainability of Homeownership in Australia
123
1960s and 1970s). This, in turn, was based on an assumption that real mortgage rates would average out at about 5–6 per cent over the cycle and nominal rates would be 2.5 per cent higher (consistent with the IGR projection of an inflation rate of 2.5 per cent). This, in turn, is consistent with current monetary policy which, since inflation targeting was introduced as a policy objective in the early 1990s, has successfully maintained inflation in the target range of 2–3 per cent over the business cycle. Interest rates are important for two reasons: real interest rates are critical for determining the user cost of owning a home (and, coincidentally, the equilibrium rental value of housing) and nominal interest rates are critical for determining the constraints on access to homeownership defined by borrowing capacity. Neither real nor nominal interest rates were covered in the IGRs, but the above assumptions presumed that real interest rates are determined by global economic factors and that risk-adjusted real rates of return are equal over all assets in the long run. In his overview of the theoretical and empirical literature on the impact of population ageing on financial markets Poterba (2001) suggests that the mean equity return over the past 100 years (in the USA and the UK) has been around 5–6 per cent in real terms. This places an upper bound on the risk-free real interest rate. With a risk premium of 2 per cent (Flavin and Yamashita, 2002), a realistic upper bound for the risk free real interest rate over the next 40 years would be 3.5 per cent. Dimson et al. (2003) report marginally higher 100-year estimates with a risk-free bill rate of 3.8 per cent for the USA and 5.5 per cent for the UK. They suggest plausible forward-looking equity premiums of 3 per cent per annum. The assumption of a 2 per cent risk premium is conservative compared with the 6 per cent equity risk premium reported by the Commonwealth Department of Finance and Administration (2006) derived from the same Dimson et al. (2003) source and used as an illustration in its discussion of discount rates for equity investment in Australia. While the demand for housing might be influenced by its user costs (covered below), access to housing finance is determined by the deposit constraints imposed on borrowers and on the amount of credit lenders are prepared to provide for a given level of income. The deposit gaps, illustrated in Figure 6.2, were based on conventional criteria set on the assumption that borrowers on incomes equal to average weekly earnings could borrow to the point where repayments did not exceed 30 per cent of income. On the same criteria, and on the assumptions about house prices and interest rates, deposit gaps would remain at the same multiple of income as illustrated in Figure 6.2 (that is, at the equivalent of either one or four years of annual average earnings depending on whether house prices are assumed to return to their 2001 trend levels or to continue to grow from their current 2010 levels). Whether price levels return to a trend based in 2001 or 2010, however, deposit requirement will become increasingly onerous in terms of
Jones_c06.indd 123
2/7/2012 10:20:44 AM
124
Challenges of the Housing Economy
average earnings because of the presumption that economic and demographic trends will ensure that house prices rise faster than real incomes. In the years preceding the global financial crisis, competition among lenders resulted in an easing of lending standards compared to those assumed here with, for example, borrowers being able to borrow with 100 per cent loans. The response to the international banking difficulties and the increase in prudential regulation that is proposed under Basel III (see glossary) both suggest that lending standards are more likely to be tighter rather than more relaxed in the future. This suggests that households with higher incomes and greater wealth are those who are more likely to be able to access housing finance and who, therefore, are more likely to be those who can satisfy and increase their demand for housing, either as owner occupiers or as investors in rental housing. Their continued willingness to do so, given the assumptions made about interest rates and inflation, depends on there being no change in other factors that determine housing demand.
User costs and returns to housing The effective cost of housing services and the effective returns available from investment in housing are key economic drivers that can affect future housing demand. These, in turn, are affected by the way in which housing is treated by the tax system. The Australian taxation system provides considerable incentives to invest in property either through equity finance for owner occupiers or through debt finance as an investor. For owner occupiers, equity investment is encouraged because the income from housing, either in the form of imputed income or capital gains, is tax free. For investors, debt-financed investment is encouraged by tax deductibility of interest costs against current income and deferral and discounting of any tax liability on capital gains. Since the introduction of capital gains taxes in Australia in 1986, owner-occupied housing has been exempt. Changes to the capital gains tax regime in 1999 (from a tax on real (realised) gains for investors to a tax on nominal (realised) gains), in conjunction with house price inflation and full interest deductibility, further increased the tax incentive for rental investment in Australia, particularly for geared investment. The assumption made in deriving the homeownership projections provided above is that these tax advantages will not be removed in the next 40 years. In releasing the report of a large-scale tax review undertaken in Australia during 2009 and 2010, the Treasurer stated that some tax changes would never be implemented. Taxing the family home was one of those explicitly mentioned. Yates (2010) shows the disproportionate benefit that these tax incentives provide to older high-income and wealthy homeowners. Such incentives encourage older households to retain or even increase
Jones_c06.indd 124
2/7/2012 10:20:44 AM
Structural Sustainability of Homeownership in Australia
125
their consumption of housing rather than to decrease it as household size and housing needs are reduced. This incentive is currently reinforced by a social security system which provides an income- and asset-tested aged pension but which exempts owner-occupied housing from the asset test. There are, however, some indications that income poor but asset rich homeowners who require care at home or in a residential facility as they age will be required to contribute to the costs of such care by releasing equity from their owner-occupied homes (Productivity Commission, 2011). A similar marginal reduction in incentives is also likely to apply to the current taxation treatment of investors in rental housing as the result of a recent review of the Australian taxation system. This provides details of the current system of taxation and the abnormal returns that can be obtained currently by geared property investment (Treasury, 2008). However, while the panel that undertook this review recommended that these investor benefits be reduced over time they did not propose that they be removed entirely (Treasury, 2009).
Rental supply The projection that homeownership rates in Australia are not sustainable at their current levels, given current economic and policy settings, raises the obvious question of how will those who do not achieve homeownership be housed. The final key assumption that underpinned the homeownership projections reported at the start of this section is that the private rental market will expand to meet the housing needs of the growing population. It is unlikely, however, that it will do so in a way that ensures that housing costs remain as affordable for lower-income renters in the future as they have been for renters in the past. Rental property in Australia is provided primarily by individual small-scale investors who obtain a return on their investment both from rental yield and capital gain. The fact that the private rental market has expanded as the population has increased provides evidence that there are investors willing to invest in residential property. As indicated above there are considerable tax incentives that encourage such investment. These incentives, however, are greater for high-income investors investing in property with high expected capital gains. Wood et al. (2010) use survey evidence to estimate that the internal rates of return obtained from individual investment in rental property in Australia are ‘very high’. But they also show that they are biased in such a way that ‘low tax bracket investors will only invest in relatively low value rental housing that attracts rents that are high relative to property values’ (Wood et al., 2010, p 230). Wood et al. (2010) highlight the importance of capital gains in the returns obtained by investors in the Australian residential rental market. Their
Jones_c06.indd 125
2/7/2012 10:20:44 AM
126
Challenges of the Housing Economy
analysis suggests that if the rate of real house price growth slows, as might arise with the slower projected growth in real GDP, then higher returns will be sought through (more heavily taxed) rental yields. This suggests that the factors that have kept real rent inflation at or close to zero over the past few decades may no longer be present, with the result that rents will rise relative to dwelling values. This view is supported by the period after the 2004 slowdown in dwelling price inflation when there were distinct signs of rapid increases in real rents. Wood et al. (2010)’s identification of tax clientele effects that arise from the asymmetric treatment of rents and capital gains suggests that these will push up rents relative to property values in the low segments of the rental market, making rental housing less affordable for lower-income households. There is also mounting evidence that the supply of rental housing affordable for lower-income households in Australia has declined significantly over the past 25 years or so (National Housing Supply Council (2009, 2010) or Yates and Wulff (2000, 2005) ). Wood et al. (2010) provide an explanation of why this can occur. Any pressure on rents at the low end of the market and associated loss of low-rent supply has several potential impacts on lowerincome households excluded from homeownership. The most obvious is that those who are able to gain access to rental housing will face higher housing costs relative to their incomes than has been the case over the past 40 years. For those who cannot afford higher rents in relation to income, there may be an increase in overcrowding as they move in with friends. Alternatively, there may be a reduction in household formation with adult children remaining in the family home. For those who do not have family and friend options there will be an increase in homelessness.
Conclusions This chapter has been concerned with the structural sustainability defined as that which ensures ‘the housing needs of the present generation can be met without compromising the ability of future generations to meet their own needs’. It has focused specifically on the structural sustainability of homeownership in Australia and, by implication, on the structural sustainability of Australia’s housing system given the significant role that homeownership has played in meeting the housing needs of the current generation. It has been shown that underlying real house price trends in Australia over the past 40 years (since the aggregate homeownership rate reached its current level), in conjunction with economic and social trends, have created access constraints for first-home buyers in Australia. One outcome of these trends is that homeownership rates among young first-home buyers have
Jones_c06.indd 126
2/7/2012 10:20:44 AM
Structural Sustainability of Homeownership in Australia
127
declined. A positive implication of this in terms of the impact of the global financial crisis on Australia’s housing market was that there were fewer highly geared marginal home purchasers to be affected by the relatively short-term fall in house prices that did occur. This chapter has suggested that, on current institutional and policy settings, long-term demand and supply projections over the next 40 years indicate that this upward pressure on real house prices is likely to be maintained, with the result that homeownership trends already observed are unlikely to be reversed. Future generations will no longer have access to homeownership on the same terms as past generations and the current generation is unlikely to be able to ‘catch up’ to match the age-specific homeownership rates of past generations. This means that they will not have access to the same tax advantaged form of asset accumulation as previous generations and will not have the same protection from housing costs in their old age as is currently the case. At the same time, real rents in the private rented are likely to rise. Such a set of outcomes will put increased pressure on the government to relieve the housing costs of an ageing and growing population.
Jones_c06.indd 127
2/7/2012 10:20:44 AM
7 Impacts on Wealth and Debt of Changes in the Danish Financial Framework Over a Housing Cycle Jens Lunde
Denmark has been experiencing the trough of possibly the most extraordinary housing market cycle in the nation’s economic history. Unlike possibly most of the housing market downturns across the world the collapse in the Danish housing prices from 2006 onwards was primarily caused by the national housing market itself before the subprime crisis. The housing market turnaround brought to an end a long housing market cycle with the previous peak in 1986 and trough in 1993 followed by a strong and unprecedented upturn. The housing market collapse has been at the epicentre of financial and economic crises in Denmark. The first signal of a weakening banking sector was seen after the downturn in the housing market had started. In January 2008 the small Bank Trelleborg had to be forcibly sold to the larger Sydbank. In July 2008 Roskilde Bank also found itself in difficulties and had to close (Danmarks Nationalbank, 2008). This marked the start of the Danish banking crisis. In the second round, the global financial crisis following the Lehman Brothers bankruptcy in September 2008 intensified the problems. The appearance of a widespread banking collapse in Denmark was stopped by government, parliament and central bank intervention in the second half of 2008 and the beginning of 2009. The outcomes of these interventions, as
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c07.indd 128
2/7/2012 10:20:26 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
129
in other countries, have been a number of takeovers and mergers. A general feature of the failed banks is a high exposure to lending on commercial properties and the financing of highly leveraged property companies, but there have also been consequences for lending in the housing market (OECD, 2009). In 2011, Denmark, like many other advanced economies, is facing serious risks to recovery from the financial, economic and housing crises. The underlying issue, as Reinhart and Rogoff (2009) note in ‘This Time Is Different’, can be expressed thus: ‘Unfortunately, a highly leveraged economy can unwittingly be sitting with its back at the edge of a financial cliff for many years before chance and circumstance provoke a crisis of confidence that pushes it off.’ (2009, p 1). Overall, the international economic policy reactions to the crises were to lower interest rates, accept huge state deficits and in some countries to make specific housing and mortgage subsidies (Scanlon et al., 2008, 2011). This chapter sheds light on potential question marks over the long-term success of these strategies for the housing market by reference to the role of mortgage finance. This chapter focuses on the role of housing finance in a country which has probably the highest level of personal debt in the world. The changes in a range of financial housing market indicators are chronicled together with parallel distributional implications for different age groups over the full housing cycle from 1987 to 2007. The research is important as an assessment of the financial soundness of owner occupiers is relevant to analyse not only the long-term risks to the housing market but also for the accompanying macroeconomic financial stability. In the aftermath of the crises some analyses of the immediate consequences were published, for example concerning the end of the housing bust (Case and Quigley, 2010) and the mortgage market (Scanlon et al., 2011). This study takes a longer horizon comparing decades rather than years (see also Reinhart and Reinhart, 2010). The methodological approach can be seen as an extension of the ‘so-called signal approach’ (Reinhart and Rogoff, 2009). The method is similar to Shiller (2007b) and Reinhart and Rogoff’s (2009) as we look for some ‘markers’ or ‘signals’ in the changes in owner occupiers’ wealth, debt and interest expenditure positions, which might have appeared before the ends of the boom in 1986, the bust in 1993 and of the recent boom in 2006–07. In this way the mechanisms of a housing crisis are identified by observing the different financial and housing stability ratios for the Danish owner occupiers over a full housing market cycle. It begins by examining the Danish house price cycles over the past three decades drawing on the key indicator of real housing prices. Real house prices can also be seen as a financial stability indicator, ‘For banking crises, real housing prices are nearly at the top of reliable indicators on early warnings of crises.’ (Reinhart and Rogoff, 2009, p 279). This chapter then examines
Jones_c07.indd 129
2/7/2012 10:20:26 AM
130
Challenges of the Housing Economy
the changing structure of owner occupation since the mid 1980s focusing on how the generational composition of the sector has adapted over this time. The next section explains the mortgage finance system in Denmark and how it has become more flexible especially since the mid 1990s. It also sets out the scale of the nation’s debt relative to other countries. The chapter then considers the financial stability of owner occupiers by reference to the underlying changes to the debt/wealth and financial commitments over the housing market cycle. Finally conclusions draw together the different strands and identify the key policy issues.
Dynamics of recent Danish housing market cycles The development in the Danish real house prices from 1975 to 2010 contained two full housing market cycles as Figure 7.1 indicates. The real house price peaks in 1979Q2 and in 1986Q1 were nearly of the same height, while the third and most recent peak in 2007Q3 really looked like a mountain climbing as that peak was 72 per cent above the ‘all-time-high peak’ in 1986Q1. In between, troughs in real house prices appeared in 1982Q4 and 1993Q2. The paths down to these troughs contained real house price reductions of 34 per cent and 36 per cent respectively. These contractions produced more than 10,000 foreclosures annually.
Index 1975q1 = 100
200
150
100
50
2010Q1
2005Q1
2000Q1
1995Q1
1990Q1
1985Q1
1980Q1
1975Q1
0
Year
Figure 7.1 Real house price index for one-family houses in Denmark, quarterly data, 1975Q1–2010Q4. Source: Monadata from Danmarks Nationalbank, quarterly house price index, deflated with the implicit consumer price deflater. After 2005 data from Statistics Denmark Figure by cand. polit Henrik Pedersen and Jens Lunde.
Jones_c07.indd 130
2/7/2012 10:20:26 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
131
From the peak in Danish housing prices of 1986 the following housing market downturn lasted seven years through to a trough in 1993 when nominal and real prices for houses and flats turned around. The number of transactions of houses and flats were at the minimum in 1993Q1 (without seasonal adjustments). The private building of houses and flats was at an all-time-low in 1993. The private building of new houses and flats increased after 1993 even though prices for existing houses and flats were much lower than for new-build units. A strong improvement in the affordability of owner-occupied housing is likely to have brought the end of the housing downturn in 1993. By then, real house prices had fallen 36 per cent since the first quarter of 1986. Interest rates also fell, improving the affordability of buying a house or flat after the Edinburgh Agreement of the European Council in 1992 allowed Denmark to sign the Maastricht Treaty, clearing the way for it to participate in the next stage of European economic and monetary union. A new government was elected in 1993 and initiated a ‘quick start’ of the Danish economy. As part of a stimulation package, several restrictions on owner occupiers’ access to raising mortgage loans were modified or removed; for example, loan terms were generally increased from 20 to 30 years. Moreover, the low interest rates rendered remortgaging profitable. All these changes worked to reduce debt service burdens significantly which boosted house prices. It was the beginning of a long rise in house prices that continued over 11 years until 2007. The first sign of the latest downturn in the Danish housing market was almost two years before prices began to fall with a drop-off in transactions. The sales of owner-occupied one-family houses in Denmark peaked in 2005Q4 (see Figure 7.2) and of flats two quarters earlier. At the peak of the housing market, sales of houses were around 20 per cent above a ‘normal’ level but with the downturn the number of transactions had roughly halved by the end of 2008 and the first half of 2009. Subsequently sales did recover but at the end of 2010 were still far below the ‘normal’. A year later Danish residential construction activity turned down too. According to construction statistics, the housing permits peaked in March 2006, while the housing starts peaked in June 2006 and the completions in September 2006. Since then the level of new residential building has been decimated as seen in Figure 7.3. House prices peaked in the third quarter of 2007 and the subsequent decline was unprecedented; over the seven quarters to 2009Q2 real house prices fell by 21 per cent, twice as fast as the housing crises from 1987 to 1993. However, in the next six quarters from 2009Q3 to 2010Q4 real house prices have stabilised. Despite this stabilisation there is still the possibility that the house price level in Denmark may be above its ‘long run equilibrium level’. At the end of 2010, the OECD Economic Outlook estimated that the Danish
Jones_c07.indd 131
2/7/2012 10:20:27 AM
132
Challenges of the Housing Economy
16000 14000 12000 10000 8000 6000 4000 2000 2010.3
2010.1
2009.3
2009.1
2008.1
2008.3
2007.3
2007.1
2006.3
2006.1
2005.3
2005.1
2004.1
2004.3
2003.3
2003.1
2002.3
2002.1
2001.3
2000.3
2001.1
2000.1
0
Figure 7.2 Number of one-family house transactions in Denmark, quarterly, 2000–2010:4 (seasonally adjusted*) * Not seasonally adjusted from 2009Q3 on. Source: Statistics Denmark. 40000 35000 30000 25000 20000 15000 10000
Building permits Started
5000
Completed
0 20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
Figure 7.3 Residential construction in Denmark, 2000–2010: annual number of building permits, starts and completions. Source: Statistics Denmark.
house-price-to-rent and price-to-income ratios are 25 per cent above their long-run averages (OECD, 2011a). The Danish Ministry of Economic and Business Affairs published in 2010 a new version of their housing price model that estimates that changes in ‘fundamentals’ explained three-quarters of the real house price rise at 88 per cent from 1995Q1 to 2009Q3
Jones_c07.indd 132
2/7/2012 10:20:27 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
133
(Økonomi- og Erhvervsministeriet, 2010). The unexplained quarter of the housing price rise could give expression to the idea that a ‘housing price bubble’ existed, i.e. based purely on expectations, also suggesting that prices still need to adjust down further. There are further reasons why real house prices may fall. The long-run effects of the expansionary macroeconomic policy measures in Denmark to address the recession, as in many other nations, will be ultimately followed by a budget tightening and, at best, restraints on household incomes. In addition, it seems unrealistic that the central bank can maintain a low interest rate level permanently. Together these factors suggest the prospects for real house prices to fall again in the foreseeable future with uncertainties about the implications for attitudes towards owner occupation.
Changing structure of owner occupation Owner occupation levels change slowly, and this section examines underlying forces and how they are influenced by the housing market cycle. From 1987 to 2007, the owner-occupation level remained broadly stable in Denmark, slightly below 50 per cent, when self-employed are not included (see Figure 7.4), and a little above 50 per cent with self-employed included
80 70 60 50 40 30 20 10
<30 years
40–49 years
60–69 years
30–39 years
50–59 years
>70 years
2007
2006
2005
2004
2003
2002
2000
2001
1999
1998
1996
1997
1995
1993
1994
1992
1991
1990
1988
1989
1987
0
All
Figure 7.4 Average owner occupation ratio and by age in Denmark, 1987–2007 (excluding the self-employed).
Jones_c07.indd 133
2/7/2012 10:20:27 AM
134
Challenges of the Housing Economy
300000 250000 <30 years
200000
30–39 years 150000
40–49 years 50–59 years
100000
60–69 years 50000
>70 years
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
Figure 7.5
Number of owner occupiers by age in Denmark, 1987–2007.
(Lunde, 2005a). However, the absolute level falls after 2005 and the overall pattern, hides significant generational differences. There is an increase in the owner-occupation rate after 1993 for households, whose oldest member is in their sixties. The origin dates back to the end of the 1960s and the beginning of the 1970s, when the tenure was especially financially attractive (see Lunde, 2005a). With the ageing of the population, the broadly stable rate of owner occupation hides a lower proportion of households at the beginning of the family life cycle becoming homeowners than previously. If the age structure of the Danish population had remained stable, the overall owner-occupation level would have been 2 per cent lower (Lunde, 2005a). In particular, the percentage of owner occupiers below 30 years of age has fallen significantly while there is also a modest fall for owner occupiers aged 30–39 years (see Figure 7.4). In absolute terms the number of Danes under 30 years of age who are owner occupiers has fallen by a third over the 20 years, 1987–2007 (see Figure 7.5). Many factors have influenced these changes but weakening affordability is not one of the causes as the net expenditure-to-income ratios have fallen as shown below. A partial housing explanation is that private cooperative housing has become more common among younger persons and families. For older owner occupiers, their growing longevity means automatically that this increases the average age of owner occupiers. A positive effect of this change in the composition of the owner-occupied sector towards older households and therefore less indebted owners is to reduce the average debt level and riskiness among owner occupiers. A key influence on the housing market has been the restructuring of mortgage finance over this period and the chapter now considers this issue.
Jones_c07.indd 134
2/7/2012 10:20:27 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
135
Changing mortgage finance and indebtedness Denmark has a relatively efficient mortgage system as borrowers have access to a range of mortgage products, and mortgage-backed bonds are sold as investment products especially for the portfolios of pension saving institutions. Mortgage banks collect together loans of the same interest rate type, amortisation form, term and other characteristics and pool them as mortgage bonds with the same characteristics. As new mortgages are issued they are bundled together to support the issue of new bonds of an equivalent value. The spread between debtor’s and investor’s interest rates is normally rather low (approximately 0.5–1 per cent). The mortgage system is more than 200 years old and has a good reputation; see Mercer Oliver Wyman (2003), Panell (2003) and Frankel et al. (2004). Over the housing price upturn years Denmark’s mortgage system became more flexible with new products and less restrictions, in common with other countries (Green and Wachter, 2007: Scanlon et al., 2008). This system was rocked by the global financial crisis emanating from the USA. It is useful to compare this system with the US mortgage-backed securities that are based on pools of mortgages with different characteristics and quality. Moreover these pools could be the base for several rather different securities, carrying different types and sizes of combined risks. Repayment options also mark a fundamental difference between the two mortgage systems. US mortgages can only be prepaid at price 100, while Danish long-term fixed interest rate mortgages can be prepaid at price 100 or with bonds, bought in the market at a lower price as relevant at rising interest rates. When the subprime crisis exposed the real risk on US CDOs (see glossary) and other mortgage bond securities, this insecurity spread to other bond markets with liquidity problems in the financial turmoil in the second half of 2008. At the top of the global financial crisis, the spread between Danish mortgage and state bonds reached as much as 160 basis points against a normal level at 20–30 basis points (Danmarks Nationalbank, 2009a). In Denmark a new mortgage must respect a loan-to-value (LTV) maximum of 80 per cent, fixed by law. Through a credit valuation, the applying borrower has to qualify to be able to pay the debt service at a fixed interest rate mortgage with a 30-year term. A mortgage bank has to make its own valuation of the market price for the borrower’s property, to determine the LTV for the offered mortgage amount. However, through the housing price upturn buyers and established owners had easy access to raise a bank loan with security above this 80 per cent mortgage limit. The majority of the first-time buyers have therefore been able to finance the total or close to the total buying price with loans, inclusive of 30-year mortgages. Therefore, nearly all Danish owner occupiers below 60 years of age have mortgages and
Jones_c07.indd 135
2/7/2012 10:20:28 AM
136
Challenges of the Housing Economy
Table 7.1 Percentage of owner occupiers with mortgage and with net liabilities and the median owner’s mortgage to housing wealth and net liability to housing wealth ratios by age in Denmark, 2007. Percentage of owner occupiers with
Age (years) <30 30–39 40–49 50–59 60–69 >70 Total
Size of ratio for the median owner occupier
Mortgage
Net liabilities
Percentage mortgage to housing wealth*
>90 >90 >90 >80 >60 >30 >70
>90 >90 >90 >80 >60 >30 >70
74 67 56 42 22 0 43
Percentage net liabilities to housing wealth* 93 77 62 42 14 −7** 42
* Housing wealth has been measured by the publicly assessed property value. ** Negative net liabilities mean that the owner has a net saving beside the housing wealth.
other net liabilities (financial liabilities net of financial assets) and this practice is reflected in high debt ratios for younger owner occupiers, see Table 7.1. The percentage of owners with mortgages and net liabilities as well as the decreasing size of the debt-to-wealth ratios follow an obvious life cycle. The affordability of owner occupation has been improved by the introduction of adjustable interest rate and interest-only mortgages. Before 1996, the fixed interest rate repayment mortgage was the dominant loan type in the Danish mortgage system. In that year adjustable/variable rate mortgages began to be offered and especially after year 2000 they captured a large part of the market. At the end of 2010 more than two-thirds of all outstanding Danish mortgages were subject to adjustable rates as can be seen in Table 7.2. Interest-only mortgages now cover more than half of all outstanding mortgages as Figure 7.6 indicates. Their popularity grew rapidly after their introduction by legislation in 2003 and their share of all mortgages has increased even after house prices turned down. Danish interest-only loans mortgages contain no repayments for the first ten years. They are available with fixed interest rates but the majority incorporate some form of variable rate, subject to annual adjustment. In this regard, Denmark has followed the international mortgage market trends as noted in Scanlon et al. (2008). In fact, in comparison with other countries Denmark is one of the countries with the most widespread use of mortgage financing. Recently, the percentage of owner occupiers with a mortgage ranged from an estimated 25 per cent in Spain to 87 per cent in the Netherlands (Scanlon et al., 2008). The highest residential mortgage debts to GDP ratios for 2009 for the European countries were: the Netherlands
Jones_c07.indd 136
2/7/2012 10:20:28 AM
Jones_c07.indd 137
2/7/2012 10:20:28 AM
37.0 28.0 28.9
January 2003
Source: Danmarks Nationalbank. MFI statistics.
Agriculture Owner-occupied dwellings All outstanding mortgage loans
Percentage 43.9 36.2 36.9
January 2004 51.1 46.0 45.3
January 2005 58.3 49.7 49.2
January 2006 58.7 46.8 47.4
January 2007 71.8 45.3 52.2
January 2008
78.1 45.9 55.3
January 2009
83.4 59.1 65.1
January 2010
Table 7.2 Adjustable interest rate mortgages as percentage of all outstanding mortgages in Denmark, 2003–2011.
86.8 65.0 69.9
January 2011
138
Challenges of the Housing Economy
60 50 40 30 20 10
2007:Q1 2007:Q2 2007:Q3 2007:Q4 2008:Q1 2008:Q2 2008:Q3 2008:Q4 2009:Q1 2009:Q2 2009:Q3 2009:Q4 2010:Q1 2010:Q2 2010:Q3 2010:Q4
2005:Q4 2006:Q1 2006:Q2 2006:Q3 2006:Q4
2003:Q4 2004:Q1 2004:Q2 2004:Q3 2004:Q4 2005:Q1 2005:Q2 2005:Q3
0
Figure 7.6 Interest-only loans as a percentage of all outstanding mortgages. 2003Q4–2010Q4. Source: Danmarks Nationalbank. MFI statistics.
105.6 per cent, Denmark 103.8 per cent, Ireland 90.3 per cent and the UK 87.6 per cent (EMF, 2010). Iceland had an even higher ratio at 118.6 per cent in 2007, but newer data are not available. At the turn of the century, Denmark had a ratio at 68.6 per cent, far above the ratios for the other countries in the Hypostat statistic. This indicates that Denmark, compared to other nations, traditionally has had a relatively high mortgage debt level over the years and in fact its owner occupiers and households are among the most indebted. In an OECD analysis of 15 countries, Denmark had the highest ratios of household debt to disposable income (see Table 7.3) and household debt to GDP, and the largest liabilities as a percentage of net wealth. John Muellbauer with knowledge of the debt data presented below found that ‘Denmark [is] the world champion for high levels of household debt’ (Muellbauer, 2007, p 271). In 2010, an OECD study found that the ‘combination of rapidly rising house prices and easy access to credit has led to a dramatic increase in the level of household debt in most OECD countries over recent years’ (André, 2010, p 22). Among 16 OECD countries, the average household debt as a percentage of disposable income had risen even more from about 120 per cent in 2000 to 167 per cent in 2007 and further to 171 per cent in 2009. The international league table is still headed by Denmark with a ratio at 318 per cent in 2007, rising to 357 per cent in 2009. These statistics demonstrate the advantage of the efficient Danish mortgage system which has delivered accessible housing finance with high affordability of owner occupation. However, the disadvantage is connected
Jones_c07.indd 138
2/7/2012 10:20:28 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
139
Table 7.3 Household debt as a percentage of disposable income in selected OECD countries. 1995, 2000 and 2009. Country
1995
2000
2009
Italy Germany France Finland United States Japan Spain Canada Sweden United Kingdom Australia Czech Republic Norway Ireland Netherlands Denmark
38 97 66 72 92 130 59 103 88 110 89 – 131 – 104 188
53 114 77 70 101 135 86 113 106 117 122 186 142 98 164 239
76 99 106 118 128 132 141 148 163 171 187 199 209 222 284 357
Source: Author’s calculations based on OECD (2011).
to the high debt level, which suggests a high risk level for these owners and borrowers as well as for lenders. There are a number of elements to this potential risk in terms of the balance of debt to housing wealth, and capital debt and mortgage repayments relative to incomes. In other words, risks to owner occupiers partly depend on the size of their borrowing, i.e. on their leverage or solvency ratio. For variable-rate mortgages there is the possibility of serious changes in the size of their payments through rises in interest rate, and this payment risk can be translated to a credit risk and even foreclosure. From a bank’s perspective current credit assessments as part of the mortgage application focus entirely on ensuring that a borrower must be able to pay the debt at a fixed interest rate. Normally, borrowers are allowed to choose other mortgage loan types afterwards. In this process, banks very often do not include the risk of falling house prices as they do not demand that the borrower applicant has a comfortable equity in advance. In the financial stability report for the first half year of 2009, the Danish central bank focused on owner occupiers’ take-up of adjustable rate and interestonly mortgages and found that the most leveraged owner occupiers used these loan types to the highest degree (Danmarks Nationalbank, 2009b, 2011). The next section looks in more detail at the reality of these issues and to what they have changed over the housing market cycle and with the liberalisation of mortgage finance.
Jones_c07.indd 139
2/7/2012 10:20:28 AM
140
Challenges of the Housing Economy
Financial stability of owner-occupied households The chapter now examines the underlying changes to the debt/wealth and financial commitments of Danish owner occupiers over the 20 years, 1987–2007, a period that encompasses – as noted above – both a long market downturn to 1993 and the subsequent rise to the pinnacle of house prices at the middle of 2007. The housing market also experienced a dramatic and rapid transformation in mortgage finance as chronicled earlier. The analysis draws on primary individual data from income tax returns unique to Denmark and assesses the following ratios: ● ● ● ●
Housing wealth to income Net liability to income Net liability to housing wealth Net interest expenditure to income
To measure these ratios it is necessary to know a household’s net debt and the market value of their property. The data used in the study are taken from income taxation statistics1 as a random sample of about 1 in 30 – approximately 40,000 – of all owner-occupier households (exclusive of the selfemployed) within each specific year. The numbers and the high data quality ensure the reliability of the results. For a more detailed presentation of the statistical sources, see Lunde (2005b and 2009). Income tax statistics for Denmark include data on each family’s income, interest income and interest expenditures, housing wealth, financial assets and financial liabilities. The Danish taxpayers know from experience that the tax authorities’ information is rather precise, as the income tax system is relatively efficient. For financial items, the tax authorities receive exact data directly from the financial institutions. A few concepts have to be defined here. The owner occupier’s housing wealth includes the total value of the family’s houses, flats and/or summer houses, owned solely for the purpose of meeting the family’s own housing needs. The owner occupier’s net liabilities are equal to their liabilities (mostly mortgages and bank loans) minus their financial assets, all calculated in market values. The family’s gross income includes all household members’ incomes. Disposable incomes do not exist in tax statistics. Finally, the household’s net interest expenditures equal their interest expenditures minus interest incomes. Ideally, the housing wealth should be measured by actual market values, but the valuation must rely on the publicly assessed property values as only a minor fraction of the properties are traded annually. In accordance with Danish law, these valuations are set biannually by use of multiple regression analysis, where the values of a number of characteristics are determined
Jones_c07.indd 140
2/7/2012 10:20:28 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
141
through the sale price statistics to construct the publicly assessed property values. In the past, the appraised values were adjusted on 1 January each year. From 2003 on, assessments have been done on 1 October but in alternate years only. Over the decades, the market prices for houses sold around the time of valuation are, on average, about 10 per cent above the publicly assessed property values (Lunde, 2005b). In the second half of 2003, average market prices for sold houses were 9 per cent above the publicly assessed property values of 1 October 2003. In 2005 this average underestimation was 18 per cent, but for 2007 the deviation was ‘normal’ again at 8 per cent. The average 10 per cent spread between the market prices for sold houses and their publicly assessed property values can be compared with the seller’s transaction costs, typically 7–8 per cent of the sale price. On average the publicly assessed property value is closer to the seller’s proceeds than the market price at the time of valuation. The Danes’ most important assets beside the owner-occupied dwellings are the long-term pensions saving (in pension funds, life insurance companies and other institutional pension saving institutions). It can be criticised that these assets are not included in the tax statistics’ wealth accumulation. In practice, this pension saving is very illiquid so its omission here can be defended. Moreover, the pension saving is lowest among younger owners, who, as the most indebted, have the highest risk of entering negative equity.
Housing wealth to income ratios The curves for the temporal trends in housing wealth to (gross) income ratios from 1987 to 2007 given in Figure 7.7 look much like the real housing price curve over the cycle. With positive growth rates for incomes, the comparison must show that housing wealth to income ratios decreased more than the real housing prices through the downturn to 1994 and increased less than the real housing price through the housing upturn to 2007. There is a caveat to the interpretation of these ratios as households might have increased their housing wealth by buying a second home or through modernisation, renovation and extension to the house.2 As a consequence, economic growth influences the number, size and quality of the family’s property and their incomes, i.e. both the numerator and the denominator in the housing wealth to income ratio. Moreover, the presented housing wealth to income ratios are calculated for individual households and grouped in deciles, as they are rather informative for detecting distributional changes. In Figure 7.7, housing wealth to income ratios for all owner occupiers (except the self-employed) are given for the second decile, the median (the half with the lowest ratios) and for the eighth decile (the 20 per cent with the highest ratio will lie above this ratio). These housing wealth to income ratios fall through the downturn to 1994 at broadly the same rate as real house
Jones_c07.indd 141
2/7/2012 10:20:28 AM
142
Challenges of the Housing Economy
700 2nd decile 600
Median 8th decile
500 400 300 200 100
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
Figure 7.7 Housing wealth to income percentages for all owner occupiers by decile in Denmark (exclusive of self-employed), 1987–2007.
prices. However, with the long upturn through to 2007 the differences in wealth to income ratios magnify with wealth generation rising dramatically faster than incomes for the owners with the highest wealth to income ratios. Figure 7.8 examines only the housing wealth to income ratios for the owners between 30 and 39 years of age over this 20 year period. This age group is chosen for further analysis because many families enter owneroccupation for the first time at that age as also seen in the jump in the number of owner occupiers from those under 30 (Figure 7.5). Other owners in that age might have moved up the housing ladder from a smaller to a larger and more expensive dwelling. Figure 7.8 shows that the fall in house prices from 1987 to 1993 is distributionally neutral in terms of wealth within this age group. Over the subsequent long upturn from 1993Q2 to 2007Q1, the real house prices rose 169 per cent, while the housing wealth to income ratio for median owners in this age group only increased by 126 per cent. The owner occupiers with the highest wealth to income ratios in this age group again benefit most in terms of wealth generation with the rise in house prices from 1994 to 2007: the ratio for the second decile increased 105 per cent, the ratio for the median owner increased 123 per cent and the ratio for the eighth decile increased 147 per cent. One explanation of the distributional widening is that the house prices in the capital region where the highest income groups tend to live increased much more than in the rest of the country, especially through the years 2003 to 2007. This is confirmed in that the extra increase in the highest deciles – for example the eighth decile in Figure 7.6 – took off from 2003. The more wealthy owners therefore became
Jones_c07.indd 142
2/7/2012 10:20:28 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
143
500 450 400 350
2nd decile Median 8th decile
300 250 200 150 100 50 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
Figure 7.8 Housing wealth to income percentages for owner occupiers in Denmark by decile (exclusive of self-employed) aged 30–39, 1987–2007.
richer, particularly through these bubble years, but since 2006 this change has been reversed as the more expensive houses and flats in the capital region have had the highest price falls.
Net liability to income ratios Net liability to income ratios give an important measure of the indebtedness of owner occupiers as a household’s income is the principal source for repaying debt, except at a sale of the home. The ratio is defined here by the family’s financial liabilities minus its financial assets and by the family members’ gross incomes. These financial liabilities include mortgages, other loans, secured against the value of their home as well as unsecured debt. Danish owner occupiers’ widespread use of mortgage and other debt financing is outlined in Table 7.1. In 2007 more than 60 per cent of all Danish owner occupiers had net liabilities which were higher than their gross income, although more than 20 per cent had net savings beside the wealth embedded in their housing and/or second home. For owners aged 30–39 nearly all had debt, almost 90 per cent had net liabilities above their gross income and 10 per cent had net liabilities more than four times their income. For the median owner occupier below 30 years of age the ratio of net liability to income was as high as 260 per cent in 2007. The ratios decrease with age, following an obvious life cycle, and only among the oldest owners, i.e. above 70, does the median owner have net savings that supplement housing wealth. Net liability to income ratios also vary over the housing market cycle as Figure 7.9 demonstrates from 1987 to 2007. These ratios to a degree fall
Jones_c07.indd 143
2/7/2012 10:20:29 AM
144
Challenges of the Housing Economy
300 250 200
Below 30 years
150
30–39 years 40–49 years
100
50–59 years
50
–50
70 years and above 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
60–69 years
–100
Figure 7.9 Net liabilities to income percentages for the median owner occupiers (exclusive self-employed) in the different age groups in Denmark, 1987–2007.
Table 7.4
Increase in owner occupiers’ debt in Denmark, 2004–2010.
Year
Banks’ lending to households for housing a term above five years, end of period DKK billion
Banks’ lending to households for housing a term above five years, end of period percentage increase
Mortgage banks’ lending to owner occupied dwellings, end of period DKK billion
2004 2005 2006 2007 2008 2009 2010
104.9 142.2 169.5 197.0 223.0 238.4 253.4
58.9 35.6 19.2 16.2 13.2 6.9 6.3
880.7 1002.0 1116.5 1216.1 1277.5 1335.4 1377.0
Mortgage banks’ lending to owner occupied dwellings, end of period percentage increase 6.6 13.8 11.4 8.9 5.1 4.5 3.1
Source: Danmarks Nationalbank.
during a downturn as occurred for the two youngest age groups from 1988 to their minimum level in 1994. It is interesting to note that this relationship lags as even though the house prices turned down at the beginning of 1986 borrowing – and hence the ratio – continued to increase until 1988. This reaction has been repeated during the recent housing crisis as outstanding mortgages among owner occupiers grew until the end of 2010 (see Table 7.4). However, for the ages above 50, their net liability to income ratios did increase over the whole cycle and thereby even through the downturn, possibly due to reduced inflation and changes in access to credit. In parallel with the upturn in housing prices the net liability to income ratios increased strongly for all age groups from 1994 to 2007 but remarkably for all the age
Jones_c07.indd 144
2/7/2012 10:20:29 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
145
groups (below age 60) the increase in the net liability to income ratios were lower than in the housing wealth to income ratios. The housing wealth to income ratios and the net liability to income ratios continued in a steep rise over the last years (2005–2007) in the upturn of the cycle. The peak of the cycle is arguably the highest credit risk for mortgage banks as lenders are in cruise control, financing houses and flats anticipating still higher market prices through the upturn in accordance with prevailing rules and financial practice. This issue also applies to equity withdrawal with loans backed by estimated house values. By this process, any market price could be financed as long as the borrower is deemed able to pay the future debt services, even if the market price had moved far away from a ‘long-run market price’. If dramatic decreases in housing prices occur in coming years these cruise control prices create high credit risks for the banks. While banks continued to lend mortgages to owner occupiers after the housing prices fell from 2006/07, the rate of increase in the outstanding mortgage was falling. The combination of higher mortgage debt and lower housing prices has one definite consequence: the solvency among the owner occupiers has worsened on average. But this conclusion is based on aggregate data, and solvency has not necessarily worsened among less solvent owners.
Net liability to housing wealth ratios The extent of an owner occupier’s wealth can be assessed by reference to the household’s net liability to housing wealth ratio. When the ratio is above 100 per cent the household is carrying negative equity, i.e. the market value of their properties and other assets is below the (market) value of the family’s debt. As a mortgage may only be one of a household’s loans, the definition in this study includes all financial liabilities (minus financial assets), which is a more accurate measure than simply looking at negative housing equity. This was confirmed by Hellebrandt and Kawar: ‘In addition, the importance of negative equity for a given household will depend on whether they have other assets, like financial investments, or other debts, like personal loans. It is the overall financial position of the households that matters.’ (2009, p 111). The impact of the housing cycle 1987 to 2007 on the net liability to housing wealth ratios for different age groups is presented in Figure 7.10. These ratios and thereby the owner occupiers’ capital structures have been very stable over the many years from 1988 to around 2003–2005, even through the housing price downturn to 1994. This is incompatible with what might be described as conventional wisdom as Hellebrandt and Kawar declare: ‘when house prices fall, the number of households in negative equity tends to rise.’ (2009, p 110). However, owner occupiers above 50 years of age became more indebted over the period. This movement has been more
Jones_c07.indd 145
2/7/2012 10:20:29 AM
146
Challenges of the Housing Economy
140 120 Below 30 years
100
30–39 years
80
40–49 years
60
50–59 years
40
60–69 years
20
–20
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
70 years and above
–40 140 120 Below 30 years
100
30–39 years
80
40–49 years
60
50–59 years
40
60–69 years
20
–20
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
70 years and above
–40
Figure 7.10 Net liability to housing wealth percentages for the median owner occupiers (exclusive self-employed) in the different age groups in Denmark, 1987–2007.
pronounced among the owners with most debt (not shown). There are a range of possible explanations including mortgage equity withdrawal, outright ownership does not seem to be a personal target any longer, loan terms have been longer, a huge remortgaging activity (with lengthening of the terms) has been ongoing and the low inflation over many years has weakened the fall over time in the real values of borrowers’ debt. Still, in general, the net liability to housing wealth ratios for older owner occupiers are relative low, and few are in negative equity. Among younger owner occupiers negative equity is widespread but relatively stable inside the single age groups. Around half of the owners below age 30, 40 per cent of owners between 30 and 39, slightly more than 40 per cent between 40 and 49, and 10 per cent between 50 and 59 are in negative equity. As housing wealth is measured by the publicly assessed property values that encompass an underestimation of around 10 per cent, these
Jones_c07.indd 146
2/7/2012 10:20:29 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
147
owner occupiers’ solvency percentages are arguably worse than the reality. However, if sale transaction costs are taken into account, the figures are inconveniently realistic. The stable relationship in the owner occupiers’ capital structure over the cycle breaks down slightly as prices rise steeply after 2003 when the ratios begin to fall, except for those over 70 years. Around 2005 the improvement in the ratios is slightly more pronounced than expressed in the curves in Figure 7.10 because the underestimation of the market values through the publicly assessed property values was higher than normal. Unfortunately, some difficulties remain in order to give a precise timing for the improvements of the net liability to housing wealth ratios as the public assessments have only been done biannually since 2003. The explanation of this improvement of the owner occupiers’ net liability to housing wealth ratios is not straightforward. It is possible that established owner occupiers did not respond to the increase in their equity by withdrawing a sufficient equity to maintain their capital structure unchanged. As house prices have fallen since 2007, for high proportions of the owner occupiers these equity gains are likely to have disappeared. The consequences of negative equity can be burdensome for owner occupiers, other owners of properties and lenders if the borrowers became unable to pay their debt services. Therefore negative equity can have serious implications for financial stability when housing prices are falling. In the earlier housing and financial crisis in Denmark over the years 1987 to 1993 there were a large number of foreclosures – more than 10,000 annually in the worst years. Through the recent crisis the numbers of foreclosures have been much lower, at least until now. The most important reason is likely to be the lower debt services, partly through low interest rates and expenditures, as discussed below, and new loan types with low payments. The Danish Ministry of Economics and Business Affairs published a study, where the housing wealth figures were transformed from publicly assessed property values to market values. After this correction, as shown in Table 7.5, 11.5 per cent of the owner occupiers were in negative equity in 2007 instead of the 14.6 per cent based on publicly assessed property values. Such a correction, as noted earlier, does not include transaction costs of around 7–8 per cent of the selling price. Furthermore, as nominal house prices have fallen 15 per cent since the 2007Q3 peak to 2010Q4 and owners have increased their mortgage debt since 2007 (see Table 7.4), the solvency position of owner occupiers and negative equity must have worsened.
Net interest expenditure to income ratios Over the studied cycle, the isolated effect of real house price changes improved housing affordability from 1987 to 1993 and worsened it over the long upturn until 2007. Initial affordability is key to market decisions as
Jones_c07.indd 147
2/7/2012 10:20:29 AM
148
Challenges of the Housing Economy
Table 7.5 Percentage of owner occupiers with negative equity in Denmark: the property value estimated by market values and by publicly assessed property values, 2004–2007. Percentage of owner occupiers with negative equity, estimated by… Year
Market values
Publicly assessed property values
2004 2005 2006 2007
18.1 16.0 15.5 11.5
24.8 20.3 21.1 14.6
Note: Values of cars and pension saving wealth not included. Source: Økonomi- og erhvervsministeriet (2010).
buying, borrowing and lending decisions for the most part are based on the first year’s payments. With interest payments forming the largest part of the families’ debt services and housing expenditures combined, this section studies interest payments (set against any interest income) relative to income for all owner occupiers as a measure of the ability (affordability) to remain in the family’s house. The analysis is based on a household’s net interest expenditure to (gross) income ratios, where the net interest expenditures equal the interest expenditures minus interest incomes. These ratios are influenced by debt size, interest rates and loan types. After 1993, as prices rose, owners increased their borrowing amounts but this is balanced by lower interest rates. Existing owner occupiers, as well as other mortgagors, changed from fixed interest rate mortgages to adjustable rate mortgages at new borrowing and at remortgaging to benefit from the lower interest rate regime. Generally owner occupiers’ net interest expenditure to income ratios as a result were lowered substantially from 1987 to 2007. This finding contrasts with a broad international conclusion in an OECD paper on household debt: ‘The interest-service burdens have been relatively stable since peaking in the late 1980s and early 1990s (the exception is the Netherlands), with the general increase in indebtedness having been mostly offset by declines in borrowing costs.’ (Girouard et al., 2007, p 13). Comparison of Tables 7.6 and 7.7 reveals that between 1987 and 2007 net interest expenditure to income ratios were lowered significantly in all age groups in Denmark. Nevertheless, once again, as with the other financial stability measures considered above, a strong life cycle to affordability is observed. The younger households have the highest net interest expenditure to income ratios and these ratios fall significantly with age. Analysis by year reveals that net interest expenditure to income ratios reduced over the cycle until 2005 and by most for the younger age groups as presented for owner occupiers in all age groups (excluding the self-employed)
Jones_c07.indd 148
2/7/2012 10:20:30 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
149
Table 7.6 Net interest expenditure to income ratios for owner occupiers in Denmark (excluding the self-employed), by age and size of decile, 1987. Age <30 30–39 40–49 50–59 60–69 >70 All
1
2
3
4
5
6
7
8
9
10
8.0 8.9 4.2 −0.1 −8.4 −16.4 −2.9
12.9 12.7 7.9 3.2 −2.9 −8.6 1.3
16.0 15.5 10.5 5.7 −0.6 −4.9 6.0
18.5 17.7 13 7.8 1.2 −2.3 9.8
20.6 20.0 15.4 10.3 3.5 −0.8 13.5
22.9 22.2 17.9 13.1 6.3 0.1 16.8
25.6 24.5 20.8 16.3 10.0 2.5 20.3
28.9 27.7 24.4 20.6 15.1 7.7 24.2
34.6 32.7 30.4 27.5 24.0 17.9 30.3
>34.6 >32.7 >30.4 >27.5 >24.0 >17.9 >30.3
Table 7.7 Net interest expenditure to income ratios for owner occupiers in Denmark (excluding the self-employed), by age and size of decile, 2007. Age <30 30–39 40–49 50–59 60–69 >70 All
1
2
3
4
5
6
7
8
9
10
5.7 5.7 3.3 −0.1 −4.2 −11.3 −2.2
8.6 8.2 5.6 2.6 −1.0 −5.8 1.2
10.5 9.9 7.3 4.5 0.9 −2.8 4.4
12.2 11.3 8.7 6.1 3.3 −1.0 6.7
13.8 12.8 10.2 7.7 5.5 0.0 8.7
15.5 14.3 11.6 9.5 7.9 3.2 10.7
17.6 16.2 13.5 11.4 10.3 6.8 12.8
20.7 18.5 15.9 13.8 13.5 10.5 15.5
26.5 22.5 20.2 18.0 18.9 16.1 20.1
>26.5 >22.5 >20.2 >18.0 >18.9 >16.1 >20.1
as illustrated in Figure 7.11 for the 8th decile. The number of highly indebted older owner occupiers, however, has been growing and the differences in ratios across age groups have narrowed. From 2005 to 2008 the payment ratios increased as a result of higher interest rates on rising debt as confirmed by Figure 7.9. Over the cycle the reductions of the net interest expenditure to income ratios were strongest from 1987 to 1994 and moderate after 1994. These trends are shaped by tax reforms which came into effect from 1987, 1994 and 1999 respectively, all of which brought reductions of the tax rates for deducting private household interest expenditures against income tax. These will have increased interest payments net of tax and removed a good part of the ratios improvement before tax. A remarkable part of the reduction of the net interest expenditure to income ratios is due to the introduction of adjustable rate mortgages in 1996 and their quick expansion, especially after 2000, as demonstrated in Table 7.2. The attraction related to the lower adjustable rates available because short-term interest rates lay below long-term interest rates. In Denmark, the higher interest rates on 30-year fixed interest rate mortgages is also due to the loans enrolled call option at price 100, which adds up to 1 percentage point to the long-term interest rates. When the central banks and the markets
Jones_c07.indd 149
2/7/2012 10:20:30 AM
150
Challenges of the Housing Economy
35 30 25 20 15 10 5 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
0
<30 years
40–49 years
60–69 years
30–39 years
50–59 years
Above 70 year
Figure 7.11 Net interest expenditure to income percentages for the owner occupiers in the 8th decile in the different age groups in Denmark (excluding the self-employed), 1987–2007. 35 30 25 Interest payments. DKK billions. Ordinary repayments. DKK billions. Mortgage debt services. DKK billions.
20 15 10 5 2003Q3 2004Q1 2004Q3 2005Q1 2005Q3 2006Q1 2006Q3 2007Q1 2007Q3 2008Q1 2008Q3 2009Q1 2009Q3 2010Q1 2010Q3 2011Q1
0
Figure 7.12 Households’ quarterly interest payments and ordinary repayments to mortgage-credit institutes on outstanding mortgage lending in Denmark, 2003Q4–2010Q4. Source: Danmarks Nationalbank. MFI Statistics.
increased interest rates from 2005 onwards, the net interest expenditure to income ratios rose immediately. This response to short-term interest rates occurred because in 2005 half of mortgages (now two-thirds) were based on adjustable interest rates. Otherwise the reaction would have been slower. The impact of recent interest rate policy is clearly seen in households’ interest payments in the central banks aggregate data shown in Figure 7.12.
Jones_c07.indd 150
2/7/2012 10:20:30 AM
Impacts on Wealth and Debt of Changes in the Danish Financial Framework
151
In particular, the influence of the widespread use of interest-only mortgages after their introduction in 2003 can be followed as repayments of households’ mortgages are very low and stable, even though households’ debt has been still growing in the time period. The increases in the interest rates could in the end have been the factor, which brought about the downturn in housing prices in 2006/7. There is therefore a high risk for the future for borrowers with variable interest rates when the central banks move away from the recent low interest rate policy.
Conclusions In common with other major economies Denmark has experienced a severe housing market trough after perhaps the most extraordinary housing market cycle from 1987 to 2007 in the nation’s economic history. The late 1980s saw a reduction of real house prices of a third, and more than 10,000 foreclosures annually. From 1993, the long upturn took off, leading to a peak in the third quarter of 2007 and meant that real house prices rose by 169 per cent. From this peak real house prices fell faster than ever seen before, by 21 per cent, until the second quarter of 2009. Prices have broadly stabilised since then, at least over the following six quarters, but there is a concern that this is only a temporary soft landing. There are doubts about the fundamental relationships between market values and underlying economic determinants. There are therefore worries about future financial stability especially given the scale of indebtedness of Danish households. The review of financial housing indicators demonstrates a degree of cyclical variability with the housing market but with significant deviations in the noughties. The housing wealth to income ratio cycle has a lower amplitude than real house prices. The ratio decreased more than the real housing prices through the downturn to 1994 and increased less than the real housing price through the housing upturn to 2007. However, with the long upturn through to 2007 the differences in wealth to income ratios between households magnify, with wealth generation for higher income groups rising dramatically faster than incomes. Net liability to housing wealth ratios, equivalent to owner occupiers’ capital structures were very stable from 1988 to around 2003–2005, even through the housing price downturn to 1994. This stable relationship breaks down as prices rise steeply after 2003 when the ratios begin to fall, but with the downturn in house prices since 2007 owner occupiers’ equity gains are likely to have disappeared. Affordability defined by net interest expenditure to (gross) income increases from 1987 to 1993 and while it worsened over the long upturn until 2007 the increase in the borrowing amounts required as real house prices rose was assuaged by lower interest rates and a shift from fixed interest rate mortgages to variable
Jones_c07.indd 151
2/7/2012 10:20:30 AM
152
Challenges of the Housing Economy
rate mortgages. Since the house price peak in 2007 housing wealth has declined with undoubtedly a rise in negative equity, yet affordability has improved and the continuation of lower interest rates remains the key to housing market stability. The analysis of housing market trends and financial indicators also reveals some important long-term trends and policy issues linked to generational differences. The level of owner occupation has been stable for decades at just under a half of all households. However, in absolute and proportional terms the number of owner occupiers under 30 years of age has fallen significantly and there has also been a modest fall in the 30–39 years age group over the 20 years, 1987–2007. The recent recession will not have reversed this trend given the reduction in the number of transactions. The analysis also shows that young owner occupiers are the most heavily in debt and that they gained least from the long rise in house prices. They are the most vulnerable too to any future market instability. The housing market upturn has benefitted the elderly at the expense of the young, creating barriers to owner occupation and financial strains on many who did succeed in buying a home.
Notes 1. The data in the paper have been made available for this study by ‘Lovmodelsekretariatet’ of the Danish Ministry of Finance. I am very grateful for these data as well as for the important personal support, willingness and enthusiasm I have met from Martin Ulrik Jensen, Ministry of Economics and Business Affairs, who has provided me with the basic statistics. I have also received invaluable support from the former head of the secretariat, Peter Bach Mortensen, who contributed the basic statistics to earlier versions of the paper together with Sune Enevoldsen Pedersen. The views expressed here are those of the author. 2. By buying a flat for renting to the child or another tenant the family’s gross wealth is increased. Then the family is a landlord, counted among the self-employed and not enrolled in the data behind the study.
Jones_c07.indd 152
2/7/2012 10:20:30 AM
8 Market Stability, Housing Finance and Homeownership in Germany Peter Westerheide
The recent experience with bubbles in many housing markets, for example in the USA, Ireland and Spain, has directed public attention to the impact of housing finance systems on the stability of housing markets. While smoothly functioning mortgage markets with low equity requirements and high loanto-value ratios have long been advocated for the sake of high homeownership ratios, they are now increasingly blamed for their strong incentives towards unsustainable indebtedness of private households and for the great sensitivity of the housing market to monetary shocks. Conversely, housing finance systems with considerable down payment requirements gain recognition because countries with such systems have proved to be rather stable in the past. However, substantial down payment requirements are just one particular element of national housing systems that might contribute to stability only within a given institutional framework. As financial accelerator models (briefly discussed in a later section of this chapter) demonstrate, down payment constraints can even introduce an excess sensitivity of house prices to income shocks through multiplier effects. This potential destabilising effect would be mitigated if down payment requirements had a positive impact on the saving behaviour of private households, or if rental markets provided adequate alternatives to owner occupied housing.
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c08.indd 153
2/11/2012 2:42:39 AM
154
Challenges of the Housing Economy
This chapter focuses on the German housing market with its particular institutional characteristics, i.e. an exceptionally low ratio of homeownership among private households and a very broad rental market, offering long-term renting opportunities in all quality segments of the market. The following section of this chapter discusses the basic characteristics of the German housing market. The third section is devoted to a description of residential property price developments in Germany within an international comparison. In the next two sections the role of income development and housing investment is analysed, followed by a description of the German system of housing finance. In the sixth section the impact of the recent financial crisis on the housing market is discussed and the seventh section focuses on the relationship between saving behaviour and down payment in a theoretical and empirical perspective. The final section draws some conclusions.
Characteristics of the German housing market The German housing market has a number of peculiar characteristics which make it perhaps unique in international comparative terms. One distinct feature of the market is the unusual stability of housing prices, both with respect to the moderate house price increase and the low volatility of price growth. Further particularities are the extraordinarily low rate of homeownership and the dominating role of private landlords in the rental market. It is different from most Anglo-Saxon countries, and also divergent from most countries in continental Europe, in that the German rental market is not dominated by short-term leases or social housing. On the contrary, it offers a suitable alternative to owner occupied housing in nearly all segments of the housing market (for a comprehensive description see Voigtlaender, 2009). Only around 43 per cent of all German households are owner occupiers, with the remainder split between tenants and members of housing cooperatives. The latter is essentially a hybrid model, although as it is usual for each single member to have only a very small contribution to the capital stock of a cooperative it is near to renting. In western Germany the homeownership rate has – on the average – been fairly stable over the past two decades. Only in eastern Germany have homeownership rates increased markedly (see Table 8.1). This reflects the long-term process of catching up in the former German Democratic Republic after German reunification, starting from a very low level of homeownership at the end of the socialist area and reflecting the privatisation efforts of formerly state-owned properties as well as new construction of owner-occupied dwellings. However, the stability of the ownership ratio in western Germany hides some differences and dynamics beneath this aggregate surface. On the one
Jones_c08.indd 154
2/11/2012 2:42:39 AM
Market Stability, Housing Finance and Homeownership in Germany
155
Table 8.1 Development of German homeownership rates (percentage of all households living in owner occupied dwellings) 1993–2008. Year
Germany
West German federal states
East German federal states
1993 1998 2003 2008
39.0 40.3 43.0 43.2
45.0 43.6 45.6 45.7
19.0 25.9 31.7 32.5
Source: Federal Statistical Office, income and expenditure surveys, own calculations. East German federal states include East Berlin.
Saarland
59%
Rhineland-Palatinade
57%
Baden-Wuerttemberg
53%
Schleswig-Holstein
52%
Lower Saxony
50%
Bavaria
49%
Hessen
48%
Northrhine-Westphalia
43%
Thuringia
43%
Bremen
43%
Brandenburg
41%
Saxony-Anhalt
38%
Mecklenburg-Vorpommern
34%
Saxony Hamburg Berlin
32% 24% 16% 0% 10% 20% 30% 40% 50% 60% 70%
Figure 8.1 Homeownership rates in the individual German federal states (% of all households living in owner occupied dwellings). Source: Federal Statistical Office, income and expenditure survey 2008, own calculations.
hand, homeownership rates differ strongly between metropolitan areas, smaller cities and the countryside. In the metropolitan areas – illustrative examples are Berlin and Hamburg – homeownership rates are exceptionally low. On the other hand, the large and economically strong federal states in the German south such as Baden-Wuerttemberg and Rhineland-Palatine, but also the northern states which are dominated by large rural areas such as Lower Saxony and Schleswig-Holstein show rather high homeownership rates, as Figure 8.1 indicates.
Jones_c08.indd 155
2/11/2012 2:42:39 AM
156
Challenges of the Housing Economy
20% Change in homeownership rates (percentage points)
16.9% 15% 9.8% 10%
5% 1.9%
0.8% 0% –0.6% –2.6%
–5% Younger than 31
31–40
41–50
51–65
66–80
Older than 80
Figure 8.2 Change in homeownership rates in age groups in Germany, 1991–2007. Source: SOEP, own calculations. Tenure status refers to the main residence. Data refer to West Germany.
There are also remarkable cohort effects pointing to strong generational differences. If different cohorts of the same age are compared – for example, 65 to 80 year old heads of household in 2007 with the same age group in 1991 – a remarkable increase in the ownership rate of nearly 10 percentage points is found (see Figure 8.2). While homeownership rates among older households are increasing over time, they tend to decrease for young households. This finding has implications for the ongoing demographic change in Germany which is likely to induce a substantial increase in the homeownership rate due to the ageing population. A further distinctive characteristic of the German housing market is its broad diversification on the supply side of the rental market. In 2006 – the most recent data available – nearly 61 per cent of the rental housing stock was owned by small non-professional landlords. Only around 39 per cent of the rental housing stock was held by professional residential property companies. Among the latter, 10 per cent was supplied by cooperatives, 11 per cent by municipalities and other public institutions, and only 17 per cent by private property companies, as detailed in Figure 8.3. In the last years before the financial crisis, ownership structures changed to some extent due to a number of portfolio transactions, mostly from the public sector to the private sector, and often involving acquisitions by international financial investors. Taken together, public housing corporations have sold around 545,000 dwellings, while private companies have acquired around 615,000 units (see Table 8.2).
Jones_c08.indd 156
2/11/2012 2:42:39 AM
Market Stability, Housing Finance and Homeownership in Germany
Churches and others 1% Private
157
Public (municipal) 10%
professional corporations 17%
Public (others) 1%
Cooperatives 10% Individuals 61%
Figure 8.3 Structure of the rental sector (share in the number of dwellings). Source: Demary et al. 2009, own calculations.
Table 8.2 Housing transactions of different types of owners in Germany, 1999–2010. Purchases
Municipalities
Federal govt. / countries
Municipalities Federal. govt. / countries Private companies Other
151,000 0
27,000 148,000
192,000 384,000
14,000 0
−222,000
10,000 0
33,000 1000
885,000 92,000
10,000 12,000
−323,000 615,000
Total
161,000
209,000
1,553,000
36,000
−69,000
Sales
Private companies
Others
Net change
Source: BBSR housing transactions database. Only transactions with more than 800 apartments each recorded.
However, the broad picture of well-diversified ownership in the rental sector remains. This ownership structure is a core distinctive feature of the German housing market as the market is dominated by non-professional landlords, basically not-for-profit-oriented cooperatives and also basically not-for-profit-oriented public residential property companies. All these players usually have a long-term investment horizon. While private landlords often regard housing investments as part of their old age provision, cooperatives and public residential property companies commit themselves inter alia to the long-term provision of affordable housing for low- and medium-income households. Nevertheless, only 8 per cent of the housing stock is explicitly dedicated to social housing, compared to, for example, 17 per cent in the UK (OECD, 2010).
Jones_c08.indd 157
2/11/2012 2:42:40 AM
158
Challenges of the Housing Economy
German housing price trends: an international comparison In international comparative terms German residential house prices have been extraordinarily stable. As Figure 8.4 shows, among the large European countries, only Germany lacked any indication of a housing market bubble in the years before the recent financial crisis. In real terms housing prices decreased over several years as overall consumer prices rose stronger than rents and housing prices. Figure 8.4 also indicates that housing prices even slightly declined in nominal terms from 1995 to 2009. Low price growth usually also implies low volatility of price changes as Figure 8.5 demonstrates. However, there are some qualifications with respect to German house price trends because if we consider different housing price indices we can observe rather heterogeneous stories. The conventional housing price indices published by BulwienGesa (a private German market research institute with a long tradition in real estate market analysis) and which are widely used and also adopted by the German central bank in its housing market reporting, refer to two standardised types of dwellings (terraced houses of 100 sq m and apartments of 70 sq m in a medium-quality market segment) in 127 German cities and show a different pattern over time from the indices provided by GEWOS (a private research institute focusing on housing markets), which refer to average transaction prices and include a broader range of dwelling types and regions of Germany (see Figure 8.6).
500 450
US Germany France
400
Italy
350
UK
300
Spain
250
Ireland Netherlands
200 150 100 50 1995:1 1995:3 1996:1 1996:3 1997:1 1997:3 1998:1 1998:3 1999:1 1999:3 2000:1 2000:3 2001:1 2001:3 2002:1 2002:3 2003:1 2003:3 2004:1 2004:3 2005:1 2005:3 2006:1 2006:3 2007:1 2007:3 2008:1 2008:3 2009:1 2009:3
0
Figure 8.4 Selected international house price trends, 1995–2009. Source: OECD. 1995 = 100.
Jones_c08.indd 158
2/11/2012 2:42:40 AM
Market Stability, Housing Finance and Homeownership in Germany
159
A comparison of different hedonic price indices for the German residential market provides an even more scattered picture, as shown in Figure 8.7. These indices are available only for brief periods of time, but even for this short interval they show quite diverging price trends. As this example suggests, exact house price measurement in the regionally and qualitatively very heterogeneous German housing market is a challenge not yet fully accomplished, and it still involves considerable uncertainty over the true market trends. The situation is aggravated by the low turnover in the market. 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0%
y
an
m er
G
US
s
ly
nd
Ita
la er
th
ce
an
Fr
n
ai
Sp
UK
nd
la
Ire
Ne
Figure 8.5 International comparison of standard deviation of house price changes in 1995–2009. Source: OECD. Based on quarterly data. 108 106 104 102 GEWOS price index owner-occupied housing
100 98 96
BulwienGesa housing price index
94 92 90 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
88
Figure 8.6 Comparison of different German conventional residential price indices, 1997–2010. Source: Association of Private Building Saving Societies (Verband der Privaten Bausparkassen), Financial Thomson Datastream.
Jones_c08.indd 159
2/11/2012 2:42:40 AM
160
Challenges of the Housing Economy
110 105 100 95 90
Destatis used dwellings VDP owner occupied apartments HPX new home F+B detached houses F+B owner-occupied apartment
85
VDP owner-occupied houses HPX apartment HPX existing home F+B terraced houses
80 4
1
2
3
2006
4
1
2
3
2007
4
1
2
3
2008
4
1
2
3
4
2009
Figure 8.7 Comparison of different German hedonic residential price indices, 2005Q4–2009Q4. Source: Federal Statistical Office, Association of German Pfandbrief Banks (vdp), Hypoport (HPX), Forschung und Beratung für Wohnen und Umwelt GmbH (F+B). 1995 Q4 = 100.
Because German households tend to build or buy their home and hold it afterwards for long periods – often in fact for their lifetime – it is not feasible to calculate repeated sales indices in order to implicitly control for quality differences by comparing essentially the same objects over time. But even after taking into account the various outcomes of the different price indicators the general finding, of an unusually moderate price growth with very low volatility in Germany, can be accepted as a stylised fact.
Housing demand and housing investment There are many possible reasons for the smooth development of housing prices in Germany. Certainly, the slow growth of private households’ disposable income over a long period of time (see Kholodilin et al., 2009) had some impact, particularly in the years following the reunification boom in Germany. From 1995 to 2009 the gross disposable income of German private households grew on average only by 0.8 per cent annually in real terms. Figure 8.8 shows that in the same period the average annual growth of real disposable income amounted to 2.2 per cent in France, 2.4 per cent in the UK, 3 per cent in the USA and even 3.4 per cent in Spain. Among the bigger European countries, only Italian households were worse off, with an annual growth of their gross disposable income of just 0.5 per cent.
Jones_c08.indd 160
2/11/2012 2:42:40 AM
Market Stability, Housing Finance and Homeownership in Germany
170
161
Germany France
160
Netherlands 150
United States
140
Spain Italy
130
United Kingdom
120 110 100 90
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
80
Figure 8.8 International trends in household gross real disposable incomes. Source: AMECO, own calculations. 1995 = 100. 80 70 60 50 40 30 20
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
10
Figure 8.9 Completed dwellings in Germany per 10,000 inhabitants, 1991–2009. Source: Federal Statistical Office.
Adding to this, large housing capacities were built up in the first years after the reunification (see Figure 8.9). This housing boom was not limited to eastern Germany but pertained to the whole country due to very favourable tax conditions. The moderate price growth in Germany can therefore also be attributed to a positive supply shock in the first half of the 1990s. Data on housing provision show that the average living space per capita increased by 16 per cent between 1995 and 2009. In 2009, each member of a German
Jones_c08.indd 161
2/11/2012 2:42:41 AM
162
Challenges of the Housing Economy
household had 41.7 sq m of living space at its disposal on average. Since their peaks in the mid 1990s, housing completions continuously decreased down to historically low levels at the end of the noughties. Due to the slow increase of German house prices – which was even lower than the moderate income growth – housing affordability, measured by the ratio of average house price to average income, has improved substantially in the fifteen years. The price income ratio decreased by around 30 per cent between 1995 and 2009. With extremely low interest rates, unexpectedly positive labour market developments and also some fear of inflation, private housing investment demand has accelerated since the financial market crisis. Consequently, in 2010, the market turned to some extent and house prices started to rise. Whether this is the beginning of a new boom phase for the housing markets is still an open question. Housing market sentiment in Germany is increasingly influenced by demographic expectations, suggesting a weakening housing demand with a shrinking and ageing population. Although this is to some extent a superficial and too simplistic view, neglecting qualitative changes in housing demand and still growing household numbers, it is going to put some pressure on the market in the coming years. The market will also increasingly be influenced by inheritance of wealth in the coming years. However, the impact of these intergenerational transfers – some €236 billion annually, estimated by Empirica, a German research institute (Empirica, 2007) – on the housing market is hard to assess. While inheritances will enable some tenant households to overcome the financial barrier to homeownership – either by inheriting the financial means for a down payment or by inheriting a house right away – most of the beneficiaries of substantial transfers will already be homeowners themselves. The impact of inheritances on homeownership rates is therefore likely to be limited. However, there might be a substantial supply effect on the housing market through inherited second houses and apartments.
The contribution of the financing system Besides low income growth and oversupply it is also very reasonable to assume that the German housing finance system has contributed to the moderate growth of house prices. The behaviour of German banks as mortgage lenders, and private households as borrowers, can be characterised as long term oriented and quite risk averse. Therefore a leveraged house price boom – as has occurred in a number of other countries – has not been observed in Germany. One distinctive feature of the German housing finance system is the rather high down payment requirement for private house buyers. While equity ratios for housing corporations differ widely among company types and legal forms (see Kröencke et al., 2011; Kurzrock et al., 2011), the private
Jones_c08.indd 162
2/11/2012 2:42:41 AM
Market Stability, Housing Finance and Homeownership in Germany
163
home buyer usually has to pay down a minimum of 30 per cent of the house value. This is not a legal requirement but follows from the refinancing structure for private mortgages. First mortgages are partly refinanced by German covered bonds (Pfandbrief). A first mortgage refinanced by a Pfandbrief can be secured only by approximately 60 per cent of the collateral value, which is around 50 per cent of the transaction value of a dwelling. In order to finance another 20 per cent of the transaction value, a secondary mortgage is usually obtained. This often involves loans from building savings societies. According to recent results from the income and expenditure surveys of the German Federal Statistical Office, 40 per cent of all German households have a positive balance on a building savings account. A peculiar characteristic of the building savings society system is that it entitles the saver to obtain a secondary mortgage at a fixed interest rate. This provides strong incentives for the accumulation of housing equity in building savings accounts. Building saving therefore is a hybrid instrument, which involves credit and equity finance. An important reason for the widespread use of this financing vehicle is that it is subsidised by government savings incentives. According to Bundesbank statistics, private balances on building savings accounts amounted to €130 bn in December 2010, while loans from building savings societies for residential investment totaled €109 bn. This compares to a total outstanding volume of €1.102 bn of housing loans in Germany, which makes clear that building society saving and borrowing plays an important role in the German housing finance system. The mortgage market is dominated by annuity mortgages with long-term maturities of up to 30 years, and interest rates which are fixed usually for a period of 10 years. The current period of low interest rates induces a number of households to request even longer fixed interest periods. Although other mortgage products (with flexible interest rates and with interest-only periods) are in principle available, these play a rather marginal role in the German market. Subprime mortgages have never been introduced in the German market. A market for high LTV loans for households with high incomes but low assets has also never really developed. A major reason for this is the tax system: owner-occupied housing in Germany is not taxed as an investment good but as a consumption good. That implies that interest expenses cannot be deducted from the tax base. Therefore one important driver for credit financing of house purchases – which accelerated the growth of interest-only mortgages for example in the Netherlands – is missing in Germany. By contrast, landlords are allowed to deduct interest expenses and other expenses (e.g. for depreciation, repairs, etc.) from their taxable income. Leveraging therefore makes more sense for them. But even in this sector – which is dominated by small individual landlords (see earlier) – financing structures are usually conservative. Small private landlords tend to have similar conservative financing structures to owner occupiers. Municipal
Jones_c08.indd 163
2/11/2012 2:42:41 AM
164
Challenges of the Housing Economy
housing corporations and cooperatives also tend to be well capitalised. In particular, housing cooperatives follow a very conservative financing model. Only private professional housing corporations are sometimes highly leveraged, especially if they focus on development and opportunity investments (for detailed information see Kroencke et al., 2011).
The impact of the financial crisis on the German housing market Due to conservative financing standards, the German housing market was not severely impacted by the financial crisis in 2008/9. Private home buyers and small private landlords in particular, but also long-term core investors such as municipal housing corporations and cooperatives, faced no major difficulties on credit markets, as an analysis carried out by the Centre for European Economic Research on behalf of the German Federal Ministry of Transport, Building and Urban Development has confirmed (Kroencke et al., 2011). Credit growth had already stagnated before the crisis, reflecting a long-term downward trend in housing investment that was not caused by credit restrictions. Although insecurities about future job perspectives and wage rises depressed the housing market to some extent in 2008 and 2009, these were calmed considerably in the course of 2010. However, the crisis has had some impact on the structure of residential loan supply. In particular, large private banks, state-owned banks (‘Landesbanken’) and dedicated mortgage banks reduced their loan volumes. On the other hand, regional banks, credit unions and savings banks, have increased their exposure. This change in the structure of credit supply has reduced the availability of high-volume credit for large professional residential property companies, particularly for those that have relied to a major extent on short-term revolving credit. But these companies represent only a tiny share of all landlords in Germany. On the positive side, as a consequence of easy monetary policy, mortgage interest rates have decreased substantially since late 2008 and are currently set at a historic low. This has had an overall positive influence on the real estate market, especially for those investors who follow a conservative financing strategy and have high creditworthiness.
High down-payment constraints and stability: contradicting aims? The role of savings behaviour The German example seems to suggest that conservative financing standards and comparatively high down payment requirements are conducive to house price stability. High equity stakes in housing finance make households
Jones_c08.indd 164
2/11/2012 2:42:41 AM
Market Stability, Housing Finance and Homeownership in Germany
165
less vulnerable to income and interest rate shocks. On the other hand, the subprime crisis seems to have proved that easy credit supply leads to bubbles and induces risks of high volatility on housing markets. However, the impact of down payment requirements on the stability of housing markets is ambiguous in theory. This has been shown in a number of papers discussing the existence of financial accelerator mechanisms in housing markets (for an overview see Almeida et al., 2006). With high down payment constraints, a substantial share of private households tends to be credit constrained. It has been shown in the literature that this introduces risks for price stability in the housing markets as well. If private households face binding credit constraints and have no opportunity to resort to renting a house instead of buying one, this implies that first-time homeowners will tend to purchase cheaper properties than they would otherwise do. Rady and OrtaloMagné (1998, 2002, 2004, 2006) demonstrate that markets in this situation are excessively sensitive to changes in the demand of first-time homeowners. If the demand of prospective first-time house buyers grows – i.e. as a consequence of an exogenous income shock – prices in the respective housing market segment rise, which in turn leads to capital gains for constrained existing first-time homeowners. These capital gains lower the credit restrictions of the latter group, many of whom might then trade up to better houses as they profit from multiplier effects through leveraging their capital gains. If for example a maximum loan-to-value ratio of 80 per cent is possible – i.e. a minimum down payment of 20 per cent is required – a capital gain of one unit leads to an additional borrowing capacity of four units and additional overall purchasing power of existing homeowners of five units for a new home. Benito (2006) has summarised the stylised facts following from these housing market conditions: House price increases tend to be more than proportionate to income increase, and income shocks to first-time home purchasers (FTP) are particularly important. It results in a positive correlation between house price fluctuations and transaction volumes. House prices in different market segments are correlated, but as a consequence of multiplier effects, prices of typical starter homes/FTP properties (low/medium-standard apartments, terraced houses) are less volatile than typical prices of properties in higher market segments. Consequently, the composition of house buyers also varies with house price inflation. Lamont and Stein (1999) show correspondingly that a higher incidence of high loan-to-value (LTV) households increases housing market volatility (44 US cities). This is confirmed by Benito (2006) for the UK, based on 147 districts in the British Household Panel Survey from 1993 to 2002. However, this does not follow directly from the models of housing markets with down payment constraints cited above. On the contrary, markets with unconstrained buyers (i.e. without any down payment constraint) should show the lowest volatility because no multiplier effects due to exogenous income
Jones_c08.indd 165
2/11/2012 2:42:41 AM
166
Challenges of the Housing Economy
shocks can occur. Any increase in down payment constraints has two effects: it increases the number of constrained buyers (the precondition for the existence of multiplier effects) and it lowers LTV ratios (which determine the size of the multiplier effects) at the same time. The impact therefore of higher down payment constraints on the price stability of housing markets is ex ante unclear, as long as people do not save more in order to compensate for higher barriers to homeownership. Only unambiguous for the stability of the housing market is the positive impact of down payment constraints on saving ratios of prospective homeowners. From a theoretical perspective, the relationship between saving and down payment constraints can be established on the grounds of the user cost concept of tenure choice. Assuming lower user costs of owner occupying instead of renting the same quality and quantity of housing, any household would immediately switch to owner occupied housing, borrowing money to purchase a house and realising a higher path of consumption over the life cycle.1 Consumption smoothing becomes more complicated if down payment requirements – i.e. borrowing constraints – are involved. In this case perfect consumption smoothing is not possible, and the advantages of owner occupation in terms of user costs have to be weighed against the disadvantages of the distortion of the intertemporal allocation of savings and consumption. Only if planned savings are higher than required down payments does no distortion occur. Households otherwise have several options (not mutually exclusive): ●
●
●
●
saving more (i.e. consuming less in early periods of the life cycle, suffering from the disutility of saving), reducing housing consumption per unit of time in the owner occupation period (i.e. purchasing a smaller/cheaper house), thereby diminishing the cumulated utility advantage of owner-occupied housing in the future, shortening the owner occupation period by postponing the house purchase, also diminishing the cumulated utility advantage of housing in the future, resorting to renting instead of owning.
These possible reactions have different implications for overall savings, the owner occupation ratio, investment in owner occupied houses and housing market stability: ●
●
Jones_c08.indd 166
increasing savings for down payments has a negative impact on nonhousing consumption at least among younger households/renters, but is very likely conducive to higher housing market stability (as long as precautionary savings are not diminished in turn), choosing smaller/less expensive houses has a negative impact on housing investment and the house price level, while the impact on stability is unclear: constrained choices are inherently unstable,
2/11/2012 2:42:41 AM
Market Stability, Housing Finance and Homeownership in Germany
●
●
167
postponing house purchases has a negative impact on housing investment and prices but is likely to be conducive to higher stability of the housing sector, abstaining from purchasing houses at all lowers the owner occupation ratio (and most likely also overall housing consumption since housing demand of renters usually is lower than that of owner occupiers).
A household’s reactions to higher down payment constraints will be influenced inter alia by ●
●
●
●
the household’s rate of time preference for current non-housing consumption: the more the household prefers present consumption the less it will react to higher down payment constraints, cost advantages of owner occupied housing, determined by ° the price-rent differential, ° non-pecuniary advantages of owner-occupied housing (such as: autonomy), ° the availability of comparable housing quality in the rental market, the household’s income, which determines its natural rate of capital accumulation and therefore the likelihood that down payment constraints have a distorting effect or not,2 the household’s risk aversion (which is ambivalent with respect to tenure choice, because renting and owner occupation both bear (but they are different) risks.
Due to the complexity of this decision it seems impossible to forecast ex ante on theoretical grounds how private households will react to higher down payment constraints. Former empirical analyses on saving and down payment constraints have also produced contrasting results. Engelhardt (1994) found negative effects of higher down payments constraint on the likelihood of saving for a down payment (discouragement effect) and also a negative effect on the savings ratio of savers for homeownership, based on simulations and an empirical analysis for Canadian households. Sheiner (1995) detected, on the contrary, a positive effect on the savings ratio for renter households based on the US PSID data for 1984. While the former studies are based on micro data, Jappelli and Pagano (1994) report higher saving ratios in countries with higher down payment requirements based on macro data. Preliminary results of micro econometric analyses, based on current data of the 2007 German Socio Economic Panel, do not point to a positive reaction of wealth accumulation for the higher down payment requirements in the German market (Westerheide 2011). Instead, this analysis suggests that higher house prices have a negative impact on savings for a cross section of young tenant households in Germany. In quantitative terms, the negative
Jones_c08.indd 167
2/11/2012 2:42:41 AM
168
Challenges of the Housing Economy
impact seems to be substantial. Moreover, the probability of dedicated savings for a home purchase declines markedly. This behaviour in Germany can partly be explained by low user cost advantages of homeownership compared to renting, particularly because quality differences between both rental and ownership markets are minor – at least compared to countries where rental markets are dominated by social housing. The exceptional stability of the German market can therefore not be attributed to positive savings responses to high down payment constraints – at least not in a cross sectional analysis for a given level of rather high average constraints. It seems more likely that the interaction between the broad rental market as an alternative to owner occupation and down payment constraints is responsible, contributing to market stability because private households resort to renting instead of deciding in favour of the inherently unstable alternative of a constrained purchase of a dwelling. These results shed some light on the effects of higher down payment requirements on saving behaviour of renters. In general in a housing market environment with a broad rental market, demand for owner occupied housing is probably being reduced. This is not necessarily the case in markets under comparatively tight mortgage market conditions in countries with a less developed rental market. In this environment higher down payments are more likely to have a positive impact on the saving behaviour of tenants.
Conclusion and outlook The German housing market has proved to be very stable in the past two decades, lacking strong volatility or leverage-driven price surges. As argued above, one reason for that is the conservative financing system, allowing for only a comparatively low degree of leverage. Missing incentives for high leverages from the tax system are another reason, because owner occupiers’ payments of mortgage interest are not deductible from taxable income. As well as the financing structures, the important role of the rental market must not be neglected. As a consequence of a broad variety of housing opportunities on the rental market in all quality segments German private households are not forced to trade up the ladder – moving from first constrained choices up to preferred alternatives on the market for owner occupied houses. The moderating influence of the rental market is also visible in the analysis of the savings behaviour of German households, as they can decide in favour of opportunities on the rental markets if down payment constraints become prohibitive for them. Long-term perspectives for the German market are increasingly impacted by the demographic change towards an ageing population with smaller and
Jones_c08.indd 168
2/11/2012 2:42:41 AM
Market Stability, Housing Finance and Homeownership in Germany
169
older households. While the homeownership rate is likely to increase due to higher homeownership rates among the elderly, this does not imply that the rental market will be marginalised. On the contrary, a new range of demand segments is entering the rental market. One of these new customer groups is formed by ‘young professional tenants’, i.e. well-qualified people ‘living apart together’ as a couple, but occupying separate domiciles due to distant workplaces in different urban centres. Another growing segment of rental housing markets is housing with services for the elderly, that offer a comfortable housing environment for individuals with current or future assistance needs.
Notes 1. If differences in the risk of tenure status are neglected or risk neutrality is assumed. 2. Income, however, might be endogenous to down payment constraints as soon as households increase labour market participation in order to accumulate higher down payments (Bourassa and Hoesli 2008). Conversely, housing consumption is usually related to income, with some income brackets even having an income elasticity of more than one (housing as a superior/ luxury good).
Jones_c08.indd 169
2/11/2012 2:42:41 AM
9 The Responsiveness of New Supply to House Prices: A Perspective from the Spanish Housing Market Paloma Taltavull de La Paz
The credit crunch hit the Spanish housing market harder than other economies. The crisis occurred just as the surge in housing starts that began in the late 1990s had reached a new historical zenith. This chapter explains the impact of the credit crunch on the Spanish housing market and in particular examines the evidence about how new supply reacted to the external shock. In support of this task it compares and contrasts housing supply elasticities in different Spanish regions over time. The analysis maintains that the strong expansion on housing supply was possible due to the flexibility of the development sector showing a high elasticity of new housing supply. Such flexibility, it is argued, is also the reason for the strong fall in housing construction as well as the tiny reduction in prices following the financial crisis. Most literature supports the idea that the more dynamic the housing markets the larger elasticity of new supply. Reduced values of the elasticity are associated with housing markets having unaffordable prices with small building construction rates, scarce new supply and being inefficient in resource allocation due to the inability to react to market signals. It is generally accepted that housing markets show an elastic reaction to new supply in the long run but the price elasticity can be much stronger in shorter periods and in some regions. This suggests that the spatial dimension
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c09.indd 170
2/7/2012 1:55:50 PM
The Responsiveness of New Supply to House Prices
171
of new supply is relevant to understanding how strong the reactions of housebuilders are when prices moves up or down, and why there are differences between regions. When the new supply elasticity is high, house construction increases/decreases faster than prices rise/fall. It is important also to set housing supply elasticities within the wider macroeconomic context. In a period of economic contraction housing development reacts much faster than during an expansionary phase because in the latter the process needs time for the selection of land, planning permissions and finance arrangements. In the former, developers just stop new projects or delay investment in the current ones. The associated effects of housebuilding on the aggregate economy contributed to economic growth (added value and employment) but dramatically contracted it during the crisis period. It can be argued that dynamic housing markets are not a key element during an expansion period due to wider strong wealth creation but during a contraction of economic activity a rapid fall in housebuilding will severely amplify the downturn. This implies that any crisis associated with the housing market penalises especially those economic regions holding the most dynamic housing markets, i.e. those which have high supply elasticities. This chapter is organised as follows. The next section provides an overview of new housing supply trends in Spain since the 1960s. The following section examines the underlying processes in the housing market pre and post the credit crunch crisis. The next section sets out an econometric model to estimate how new supply housing elasticities have varied over time for Spain and its regions and presents its results. These results focus on the influence of the credit crunch. Finally, the conclusions review these results in the context of the housing market over the last two decades.
Housing supply trends Spain experienced the longest growth wave in housebuilding in recent history over the period 1994–2010, with starts reaching an all-time record in 2007 as Figure 9.1 shows. The next previous long cycle was as far back as 1960–1975 but it was less significant. This latest long upturn in housing construction activity reflected substantial urban expansion and promoted positive effects on the general economy both in terms of wealth and income. The strong new housing supply growth was the development sector’s reaction to demand impulses arising from a number of different fundamental sources. As in other countries, the availability of financial funds was reflected in the housing market by an increase in mortgages supporting housing purchases and homeownership (see below). The Spanish market also experienced a significant growth stemming from demographic change. There were three sources of population increase. First, the number of new
Jones_c09.indd 171
2/7/2012 1:55:51 PM
172
Challenges of the Housing Economy
70
Thousands of units
60
Starts
50 40 Vertical (value) axis major gridlines 30 20
Completed
10 Credit crunch shock 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
0
Figure 9.1 Housing starts and completions in Spain (thousands of units), 1960–2011. Source: DG de Arquitectura y Vivienda (MFOM) and M Vivienda.
households rose as a result of the young structure of Spanish population that reflected a follow-on wave of the baby boom that took place in Spain between the 1960s and 1970s, a decade later than in other European countries. Second, there was the arrival of significant numbers of immigrants for labour purposes (around four million people between 2001 and 2007). Third, Spain saw a rising flow of European citizens buying houses, reflecting a new way to live (retiree families, but also part-time workers, and very concentrated in a few regions). The last flow began in the mid 1990s but increased dramatically from 2001 as Figure 9.2 demonstrates. It also shows how the market was enlarged through a change in mobility patterns over the past decade, that led to more households moving within Spanish regions. Most of the internal movements concentrated in those areas with rising economic activity first before spreading out into other areas from 2003. Migration flows coincided with exceptionally positive economic conditions, namely job creation, low interest rates, low relative housing prices and the intensive flow of liquidity noted above in both Spanish and European financial institutions. It enabled the potential flow of demanders to become homeowners, including immigrants, but also brought rising prices. The credit crunch and the subsequent loss of jobs reduced the arrival of immigrants from 2008 but the level of mobility of population was maintained. The latest information available (for 2009) shows that the number of new foreign residents in Spain dropped to the 2003 level, but that they were still arriving. There was a very positive supply response to this rising demand due to changes in the regulation of the land market and in the development industry
Jones_c09.indd 172
2/7/2012 1:55:51 PM
The Responsiveness of New Supply to House Prices
173
Number of persons per month moving from one house to another in different city
250,000
Within Spanish regions From outside Spain Credit crunch shock 200,000
150,000
100,000
50,000
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1988
1989
0
Figure 9.2 Population mobility within Spain and from abroad 1988–2009. Source: INE, EVR.
which in the past had acted as traditional constraints of new developments. After the recession of 1991–1994, most municipalities prepared new land plans in order to avoid the effect that land scarcity had had on the housing market in the 1980s: strongly rising housing prices close to 29 per cent during the previous expansion period (1985–1990). Land authorities’ concern was to promote a more flexible regulatory framework to create land in periods of pressures; the policy began first in one region and then was adopted in the rest of the country.1 In addition as the demand pressures arrived in the different regions most large developers, as well as smaller ones, moved into these areas to develop new land and were able to attract immigrant construction workers, so that there was no shortage of labour in the housing supply industry. The importance of new construction for the housing market is indicated by the fact that it accounted for a growing proportion of total transactions in the first half of the last decade (see Figure 9.3). The vast increase in new construction expanded the market size to around 900,000 transactions during 2006. The new house market share of total transactions was stable at around 34 per cent during 2004–5 but jumped to 59 per cent in 2008 as the credit crunch impacted strongly and negatively on the sales of existing houses. New sales maintained this majority share of total houses sold until the end of 2010. A strong market contraction was experienced after 2008 when the liquidity flows from the financial markets stopped and Spanish banks and saving
Jones_c09.indd 173
2/7/2012 1:55:51 PM
174
Challenges of the Housing Economy
300,000
250,000 Total New houses Existing houses
200,000
150,000
100,000
50,000
11 20
10 20
09 20
08 20
07 20
06 20
05 20
20
04
0
Figure 9.3 Total, new and second-hand quarterly housing transactions in Spain, 2004–2010. Source: Spanish Ministry of Housing.
banks, first, contracted finance to developers and then to households. The structure of the Spanish financial system is based on the existence of two different institutions,2 domestic banks (with 30–35 per cent of mortgage market share) and savings banks (non-profit institutions which expend their benefits in social projects and with a strong presence on mortgage markets). Both have developed a dense network of offices across Spanish territory creating a strong competence with sufficient flow of financial services in all sectors and regions. The aftermath of the credit crunch was a strong reduction in housing starts to a historical minimum (see Figure 9.1), an increase in unemployment in the building sector and a consequent fall in the economic cycle. Housing transactions also fell to half of the level at the top of the cycle but still maintained an average of 450,000 deals yearly (see Figure 9.3).
Spanish housing market cycle pre and post the financial crisis Anatomy of housing market boom Spain experienced strong price growth in 1986–1991 followed by a recession with a short real fall in housing prices in 1992–1994, as Figure 9.4 indicates. There was then a strong recovery that began in 1997 and continued through to 2004. The growth in real housing prices was far stronger than the previous
Jones_c09.indd 174
2/7/2012 1:55:51 PM
The Responsiveness of New Supply to House Prices
30.00
Credit crunch price reaction
Spain average New houses Existing houses Public houses
25.00 20.00
175
15.00 10.00 5.00
2011 1T
2009 1T 2010 1T
2008 1T
2007 1T
2005
2006
2003
2004
2001
2002
1999
2000
1997
1998
1995
1996
1993 1T
1994 1T
1991 1T
1992 1T
1989 1T
1990 1T
1987 1T
–5.00
1988 1T
0.00
–10.00 –15.00
Figure 9.4 Annual changes in real house prices in Spain, 1988–2010. Source: MFOM up to 1995, then Spanish Ministry of Housing.
one in the Eighties reaching a maximum annual rate of 18 per cent. The rise in house prices was experienced more by existing housing than new houses. The annual rate of real house price inflation then reduced smoothly during the period 2005–2008, but turned negative during 2008–2010. During this long upturn, market conditions were very favourable for the medium- and low-income households to buy their own home. There were several reasons. One was the low interest rates maintained over a long period (falling towards the historical minimum levels from the entry into the European Monetary Union in 1999). Second, in historical terms housing prices at the beginning of this period were moderate relative to income. A third reason was the changes in financial conditions for mortgages (favouring low interest rates) which increased the average maturity toward more European standards, from 15 years towards 25 years. The fall in interest rates increased affordability despite the rising housing prices. These processes happened with a loan-to-value (LTV) average ratio less than 80 per cent for Spain. Figure 9.5 shows the four indicators: loan-to-value ratio, debt-toincome calculated using 15 and 25 years of credit maturity as well as the price-to-income ratio. It can be seen that during the period 1997–2006 affordability of mortgages in terms of LTV reached a maximum, although from 2000 the value of the price income ratio starts to rise towards 10.3 There was an apparently ‘infinite’ flow of finance feeding mortgage loan availability. From the early 1990s mortgage loans increased dramatically in Spain (see Figure 9.6). This was the consequence of changes in the international financial system (through the globalisation of capital markets)
Jones_c09.indd 175
2/7/2012 1:55:52 PM
176
Challenges of the Housing Economy
100.00
LOAN-TO-VALUE ratio
90.00 Credit crunch shock 80.00
Percentage
70.00 60.00 50.00
DEBT-TO-INCOME ratio (20 years maturity)
40.00 30.00 20.00 10.00
2011
2010
2009
2008
2007
2006
2005
2004
2003
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
2002
DEBT-TO-INCOME ratio (30 years maturity)
0.00
Figure 9.5 Housing affordability ratios for Spain 1989–2011: Loan-to-value, debt-toincome and price-to-income. Source: INE, MFOM, BDE, own calculation. 400,000 350,000
Credit crunch shock
60,000,000
Number of loans (left axis) Total amount in thousand of euros
50,000,000
40,000,000 250,000 30,000,000
200,000 150,000
20,000,000
Thousands of euros
Number of loans
300,000
100,000 10,000,000 50,000 0
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
0
Figure 9.6 Annual number of residential mortgages in Spain, 1975–2011. Source: INE, and own calculations.
and the unification of the interbank market at European level (in the Eurozone). This led to cross-border movements of capital towards Spanish housing markets with high yields due to the strong demand. Most mortgage banks operating in Spain (both domestic and international) had easy access to such funds, creating a situation where the credit restriction effectively disappeared from the Spanish housing market, supporting the cycle expansion in housebuilding and transactions.
Jones_c09.indd 176
2/7/2012 1:55:52 PM
The Responsiveness of New Supply to House Prices
177
25 20
Credit crunch shock
15 10
2010TI
2009TI
2009TIII
2008TI
2008TIII
2007TI
2007TIII
2006TI
2006TIII
2005TI
2005TIII
2004TI
2004TIII
2003TI
2003TIII
2002TI
2002TIII
2001TI
2001TIII
2000TI
2000TIII
–5
1999TI
0
1999TIII
5
–10 –15 –20
GDP Investment in housing Investment in other constructions
–25 –30
Figure 9.7 Percentage annual change in real terms in Spanish GDP and housing investment, 1999–2010. Source: INE.
The positive position on affordability and availability of mortgages was supported by the positive reaction of local planning policies, as discussed above, which brought an increase in land availability and simplified the administrative barriers to housebuilding permissions. The perception of rising housing needs in most parts of Spain during the 1990s promoted this policy, and it was further strongly supported when immigration started to increase dramatically in the 2000s. The combination of strong demand, favourable ownership conditions, transparency on the regulation to ownership and a positive planning response from the local authorities all supported the supply expansion. A consequence of this process was a growth in the general economy, accelerated by the positive employment effects of the construction sector. Figure 9.7 shows stable real growth rates of around 3.5 per cent in the economy and 6 per cent in construction during the period 1996–2007.
The bust The global financial crisis stopped suddenly the international flow of capital into Spain both in the short termand the long term. The shock arrived at the strongest and hottest point of the building cycle in many parts of Spain. The fall in liquidity affected most institutions which specialised in mortgages, resulting in a strong contraction in credit within the housing market both to buyers and developers, reducing the amount of credit to levels applicable at the end of the 1990s.
Jones_c09.indd 177
2/7/2012 1:55:52 PM
178
Challenges of the Housing Economy
The impacts in the years since the crash have significantly affected the residential sector as follows: 1. There was a fall in housebuilding which quickly reacted to the shock from the very moment the reduction in finance occurred. The differential impact on the house industry was more severe than the rest of the construction sectors. As Figure 9.1 demonstrates, housing starts fell dramatically to an all-time historical minimum level in just one and a half years. Nevertheless there is some continuation of building activity, completing the houses under construction. 2. There was a small fall in real house prices, less than 14 per cent, accumulated over three years after the crisis, trending to zero quickly. This weak price reaction confirms housing prices as sticky, and it is the opposite to the strong building reaction, implying the existence of a high elasticity of new housing supply. 3. There was a strong reduction in employment associated with the construction sector. Most of the job losses in the economy during 2008–2010 were in construction, before the wider negative effects were transmitted to the rest of the economy. 4. There was a contraction in the housing market size in Spain. The reduction in the number of transactions to half the level from the peak has had a negative income impact as well. However, Spanish statistics do not allow observation of the market size before 2004. Using other indicators such as the price level, it is possible to argue that the 2011 level of transactions is close to the long-term equilibrium level. While there is no empirical evidence in Spain, the negative income effect derived from the fall in perceived housing wealth should play an important role in dampening aggregate consumption. The macro effects coming from the fall in both private consumption and employment have had a relevant impact on GDP growth rate. In addition, there is also the well-known Spanish financial and sovereign debt problem resulting from the shock of the financial crisis at the European level. In the aftermath of the crisis there are two positive signals for the housing market suggesting expectations of rapid recovery: first, the low level of interest rates which ensure high housing affordability (shown in last data included in Figure 9.5), and second, the existence of strong basic demand flows coming from both mobility and the population growth, which is supporting the housing market (shown in Figure 9.2). The generalised wealth effect from the previous expansionary period when households (including most with low income levels and immigrants) became homeowners could also be a positive sign, supporting future consumption in the recovery phase of the cycle, but it will act after the general economic recovery.
Jones_c09.indd 178
2/7/2012 1:55:53 PM
The Responsiveness of New Supply to House Prices
179
The increasing rate of homeownership in Spain (89 per cent of households) reinforces this argument. As has been shown, the negative reaction of housing starts, falling faster and before housing prices, probably helped trigger the consequent macroeconomic effects. Even with the financial constraints many developers could have adapted their decisions to build, investing ‘when the market required them to do’, following the market impulses. The rapid reaction was arguably unnecessarily strong in terms of speed and strength. This chapter hypothesises that the reason lies in the high degree of flexibility reached by the housing construction industry sector in some areas in Spain. The rest of the chapter seeks to demonstrate that large new housing supply price elasticities across different regions explain this phenomenon.
Estimation of the supply elasticity of new houses A standard economic model is applied here to estimate new housebuilding elasticities, and to assess the supply reaction to the credit crunch, among the regions of Spain. The model distinguishes between stock, which is the total number of units existing in the market, and supply flow which is the total number of units available on the market capable of satisfying the demand. The supply curve is considered rigid in the short run as a result of the special characteristics of housing and its production process (see, for instance, Arnott, 1987). The lack of information and financial requirements creates uncertainty and so supply (flow) is expected to respond slowly to market signals. A positive shock on any demand component causes an upward reaction to prices in the short run while housebuilding cannot increase suddenly. The supply responds with a lag and only increases step by step as starts are completed. Following a contraction in demand most evidence finds that prices do not dramatically drop but the total number of vacant units increases and new supply decreases. The interaction between demand and supply, as well as their sensitivity, is therefore a key aspect to understanding price behaviour. According to the literature, equilibrium does not occur in the short run because of the rigidity of the supply curve but does so in the long run as the curve acquires more flexibility (DiPasquale, 1999; Meen, 2002b; Topel and Rosen, 1988). This implies that the way changes in new supply react to price changes is the key to market adjustment and that supply elasticities are variable in the shortand long-term. A full justification for the statistical analysis presented here and the underpinning literature are detailed in the Appendix. The essential factors
Jones_c09.indd 179
2/7/2012 1:55:53 PM
180
Challenges of the Housing Economy
considered to be relevant in the decision to supply (new) housing units are as follows: 1. The supply of housing depends on developers’ returns, which in turn are a function of the prices achieved by the dwellings and construction costs. 2. Construction costs are an essential determinant for the supply function, including the cost of capital (interest rates) and labour. 3. In the analysis of local markets supply additionally depends on a set of specific factors, e.g. population, market activity (transactions), location, market characteristics, etc. (Goodman, 2005). The statistical analysis therefore estimates supply elasticity using a a loglinear model as follows: Ln (Qsn t t ) = α1 + α 2 ln PH,t + α 3 ln Cm t + α 4 ln Cs t + α 5 ln i t + α 6 G t k + μ t
(9.1)
where Qtsn = new supply in year t PH,t = housing prices in real terms in year t Cmt = the costs associated with construction materials in year t Cst = indicator of the payment of salaries to construction workers in year t it = real interest rates paid by developers for building credits in year t Gtk is a matrix of the regional market’s k characteristics, including physical features as well as other aspects like land and market size in year t and α2 represents price elasticity. The econometric model is designed to allow the elasticity to be time varying; that is it changes depending on the period observed. This varying elasticity is cumulative, where each observation represents the whole elasticity from the beginning of the observed period until the related date: {ε t } = (ε1 , ε 2 ,...........ε t ) Where: ε1 = elasticity estimated for the initial period, t0 to t1 (1990–1994 in this case) ε2 = the elasticity estimated for t0 to t2, where Δt = t2 – t1 refers to the longer period of an additional year, 1990–1995. εt = the elasticity estimated for t0 to tn, with tn the last year observed. This means that each observation represents the elasticity in the considered period until the last year. The last one is the whole-period elasticity, and adding observations gives a more long-term perspective of the elasticity. The outcomes of the model therefore maps the evolution of the supply response over time as this is a keystone issue in terms of developers’ changing reactions to market news, changes in prices or advancing losses, etc. To take account of the additional credit constraints/reduction in starts
Jones_c09.indd 180
2/7/2012 1:55:53 PM
The Responsiveness of New Supply to House Prices
181
Table 9.1 New housing supply ‘additive’ elasticities over time for Spain, 1994–2010. Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
(−0.72) (0.01) (−0.38) (−0.14) 0.33 0.59 0.88 0.69 0.62 0.54 0.56 0.51 0.69 0.67 0.72 1.27 1.26
e -shock adjusted
Adj R2
DW
F
0.80 0.80 0.80
0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.95 0.94 0.94 0.93 0.92 0.92
2.04 2.10 2.10 2.14 2.17 2.19 2.19 2.21 2.23 2.24 2.27 2.26 2.27 2.32 2.73 2.61 2.61
120.40 143.01 176.85 195.79 242.48 272.68 294.46 319.73 353.98 376.50 443.73 412.16 452.40 403.10 354.42 352.20 352.20
following the credit crunch a dummy variable is subsequently added to cover the period from 2007 and Equation 9.1 re-estimated. The time series quarterly data for the model covers the time period, 1990–2010, and relates to the Spanish aggregate market as well as some 16 selected regions (15 autonomous communities plus the aggregate of the two autonomous cities, Ceuta and Melilla). The model is estimated using pooled (EGLS) regression with fixed effects by regions (whole model) for both Spain and representative regions of the Spanish housing cycle. Table 9.1 reports the estimated value of new housing supply elasticities, with numbers in brackets values that are not statistically significant while the rest are significant at 1 per cent. The statistical specification of the model including full diagnostic tests and their interpretation are in the Appendix, as are the detailed results for selected regions. The calculation procedure follows an iterative process whereby each successive estimation adds information from one year ahead to the time series. As a result, the last calculated value of the elasticity refers to the long term, while the first elasticity shows the initial short term. The outcomes is a supply elasticity time path, as shown in Figure 9.8, moving from the short-run to the long-term values during the study period. In terms of the estimated values, the aggregate Spain model reports statistically significant elasticities, rising over the period 1990–2000 to 0.9, and overall for the whole period, 1990–2010 it is 1.3. However, the value of the new supply elasticity is lower when any period between 2000 and 2010 is
Jones_c09.indd 181
2/7/2012 1:56:12 PM
182
Challenges of the Housing Economy
1.4 1.2 1 0.8 0.6 0.4 Supply elasticity 0.2
Shock adjusted
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
0
Figure 9.8 Changing Spanish housing supply elasticities over time, 1994–2010.
used to estimate the model. It falls in value in the first part of the decade until 2006 when it starts to rise again. Once the dummy variable capturing the impact of the credit crunch is included in the model the ‘shock adjusted’ elasticity values remain around 0.8 (see Table 9.1). The incorporation of this added variable to take account of the credit crunch shock therefore produces constrained elasticities which could explain the strong falls in building experienced in Spain. The results demonstrate the dynamic behaviour of new supply elasticities with changing values (and the responses from the developers) over time. The estimated elasticities for selected individual regions4 shown in the Appendix reveal that many exhibit large values (higher than one) in at least some range of time within the study period. This is true for the regions of Andalucía, Asturias, Aragon and Extremadura during 1997–2002, and Murcia and Comunidad Valenciana up to 2005. However, and significantly, Madrid and Cataluña, where the two principal cities are located, have low inelastic responses of new supply to price changes: Madrid has a decreasing elasticity from 0.5 to 0 and the Cataluña region has a decreasing elasticity from 1996 to 2003 (from 0.6 to 0.3) before then rising to 0.8. The low values in Madrid, which as the capital of Spain, experiencing strong demand pressure suggests the role of other variables contributing to the housing cycle such as administrative interventions, lack of land or market power. In all regions analysed, with the exception of Castilla La Mancha, the elasticity estimations fall from the end of 1990s through the next decade. New housing supply therefore becomes less elastic during the noughties, so
Jones_c09.indd 182
2/7/2012 1:56:12 PM
The Responsiveness of New Supply to House Prices
183
the reaction of developers to prices is reducing in the period leading up to the credit crunch. The supply reaction ‘jumps’ after the external shock. In the regions with the more dynamic economies (Aragón, Canary Islands, Castilla La Mancha, Cataluña, Valencia and Murcia), the elasticities reach high values suggesting a sudden retrenchment in starts following the collapse of house price growth. Asturias, Extremadura and Madrid experienced the opposite as elasticities move close to zero with no reaction of new building to the change in prices. The model of new supply (Equation 9.1) captures how starts react to market determinants defined by the theory. The explanatory power of the estimated equation for each region, the Adjusted R2, can be seen as a measure of the degree to which market signals motivate developers in each locality (market power). Regions with high Adjusted R2 can be interpreted as markets where a high percentage of starts are produced as a response to market signals while a lower Adjusted R2 means that only a small part of the starts can be explained by such mechanisms. In regions with a low Adjusted R2 most of the house construction decisions could derive from public policy (for instance, public housing provision) or be linked to noncompetitive markets (for instance, land oligopolies). The time series of Adjusted R2 associated with the estimates of supply elasticities (as in Table 9.1) can be seen as a potential ‘walk’ through time for each region with growing explanatory power, reflecting a greater role for market signals and vice versa. For most regions the explanatory power increased dramatically during the period, 1990–2000, from a non-market position (lower Adjusted R2 < 0.5) to market (Adjusted R2 > 0.5).5 The explanatory power then remains broadly stable around what seems to be the maximum value from 2000 onwards. There are three cases where this pattern is not followed: Madrid, Extremadura and Asturias. Madrid is a distinctive region and has already been discussed. The other two can be treated as outliers or special cases. Table 9.2 classifies the reactions among the regions in terms of their (changing) market orientation over different time periods. As can be seen, all regions except Extremadura gain in market orientation, moving to a degree from a non-market position towards a market over the study period. For some of them (Cataluña, Aragón, Madrid, Murcia and Castilla La Mancha) this process continues to expand the influence of market mechanisms right up to 2007, while others (Valencia, Andalusia, Canary Islands and Asturias) reach a stable ‘equilibrium’ at the turn of the millennium. In most regions the estimated supply elasticity values move closer to or higher than one at the same time that the ‘market signals’ become more powerful to explain the expansion on house construction. After the credit crunch external shock, some regions remain in the stable classification suggesting that market signals remain dominant in explaining the strong reaction on starts, rather than prices. Five regions modify such
Jones_c09.indd 183
2/7/2012 1:56:12 PM
184
Challenges of the Housing Economy
Table 9.2
Changing role of the market in Spanish regions 1990–2010.
Until 2000
2001–2007
From non market to market (growing explanatory power) Cataluña Cataluña Aragón Aragón Madrid Madrid Murcia Murcia Valencian Community Andalusia Canary Islands Castilla La Mancha Castilla La Mancha Asturias
2008–2010
Cataluña
Stable Valencian Community Andalusia Canary Islands
Valencian Community Andalusia Canary Islands Castilla La Mancha
Asturias Extremadura From market to non market (falling explanatory power)
Extremadura
Aragón Madrid Murcia Asturias Extremadura
mechanisms going back to the non-market class (Aragón, Madrid, Murcia, Asturias and Extremadura), suggesting that interventions other than market reactions are taking place to explain the reduction of starts in this period. Overall the study shows that that most housing markets in Spain became more ‘market oriented’ over the study period with larger elasticity values showing a quicker reaction of housing supply to price signals. The responses from the development sector increased the housing supply sufficiently to dampen an even potentially stronger increase in housing prices from growing demand pressures. The impact of the credit crunch reduced the available finance to mortgages, diminishing the demand pressure and then the price signal become weak. The higher regional elasticities contributed to this market adjustment via quantities (reducing housing starts) rather than through prices.
Summary and conclusion The chapter reviews the expansion process of housebuilding experienced during the two decades in Spain, 1990–2010. It explains the growth in construction activity as a response to demographic change, an immigration
Jones_c09.indd 184
2/7/2012 1:56:12 PM
The Responsiveness of New Supply to House Prices
185
influx supported by lower interest rates and the increased availability of mortgage finance. Housebuilding was also facilitated by a relaxation of planning constraints. These changes induced the market size to rise and reach more than 900,000 housing transactions in Spain during 2006. The process has had income and wealth effects on the individual regions while contributing to the overall success of the Spanish economy during this period. The impact of the credit crunch reduced the flow of credit available to both developers and buyers, generating a sharp reduction in housing starts in all regions and inverting the process of economic growth. During the recession, the housing market experienced a dramatic shrinkage in starts while housing prices only fell, in nominal terms, by around 3–5 per cent yearly (a 13.5 per cent cumulative fall in general housing prices from 2007 to 2010). Given this asymmetric response this chapter then concentrates its analysis on the reactions of developers to house price changes. A new housing supply model is estimated for all Spain as well as selected regions. The elasticity of new housing supply is estimated for both Spain and representative regions for the housing cycle over the past two decades. Elasticities for time periods are calculated using an iterative method consisting of successively estimating the model by adding years. The results give elasticity estimations for short-run periods (first value calculated is for the period 1990–1994) to long-term periods (last elasticity value estimated is 1990–2010), which means that the last elasticity is representative of long-term elasticity. The empirical analysis finds that new housing supply price elasticity varies over time. There is a growth in elasticity values in some regions but for most of them the elasticity fell over the longer term. Over the period, the model estimations indicate that the housing market in Spain became more market oriented, following a walk from non-market to market during the 1990s and 2000s. Such a change suggests that market signals played an increasing role in housebuilding decisions in Spain and this reflects the lack of constraints on construction during the period. The size of the estimated supply elasticities explains the market correction through a dramatic fall in starts after the credit crunch crisis hit the economy. The credit crunch had a dramatic impact on housing construction, reducing the figures to the historical minimum after the hit. During the immediate aftermath the price elasticity of new housing grew in all regions at the time. The regions with a stronger fall in construction are also those with higher elasticities. After three years on from the credit crunch and having experienced the significant fall in construction activity, the Spanish housing markets are starting to feel a lack of supply. Some areas with a still rising flow of demand show housing needs derived from scarce housing supply and reduced affordability. Such flows still emanate from demographic pressures with positive immigration as well as the population moving among Spanish regions for
Jones_c09.indd 185
2/7/2012 1:56:12 PM
186
Challenges of the Housing Economy
labour market purposes. The labour market has been highly damaged due to the credit restrictions and deficit constraints with a fall in domestic demand, economic activity as well as public investment and expenditure. The unemployment rate rise approaching 20 per cent has strongly affected the young, new households and immigrants. These groups are the main housing demanders in Spain for a primary home, and it is expected that their need will translate into housing effective demand only when the general economy recovers. The short-term perspective in the market is constrained by a strong lack of housing supply, the legacy of the fall after the credit crunch. Expectations about the future of Spanish property market depend on the evolution of the Spanish economy and it is related to how real estate sectors will face the extremely strong conditions imposed by the EU to adjust the foreign debt and the restructuring of the financial sector. These will be critical to advance any upturn in housing markets which will probably slow the recovery over the coming years. However, the higher elasticities shown in this chapter could advance such a recovery which could come faster than expected if domestic demand recovers relatively soon.
Appendix: Details of statistical analysis Literature review There are a range of studies that have estimated supply elasticities, and it is widely recognised that residential supply is flexible in the long run. It is generally agreed that short-run supply elasticities are smaller than long-run ones because some time is required for building (Quigley, 1997; Topel and Rosen, 1988; Malpezzi and Maclennan, 2001; DiPasquale and Wheaton, 1994; Goodman, 2005; Malpezzi and Vandell, 2002). Nonetheless, the international experience shows that, although US research contrasts the presence of elastic values in the long and short run this may not be generalised in the experiences of other world regions. The available studies about Europe give reduced supply elasticities with a value near zero in the case of the UK during the 1990s and early 21st century (Meen, 2003; Barker, 2003; Pryce, 1999; Malpezzi and Maclennan, 2001; Bramley, 2003). Some other studies have reported extreme values of supply elasticity (Whitehead, 1974; Mayes, 1979; Bramley, 1993, 2003; Pryce, 1999; Swank et al., 2002; Mayo and Sheppard, 2001). Most of the differences arise from the way that the supply function is defined. The theory has largely related the supply to determinants of the production function. This link has generated many of the difficulties in calculating and defining a market supply function (Hanushek and Quigley, 1979). This is because, first, production (starts) is not the only source of supply; second, there is a lack of data available to
Jones_c09.indd 186
2/7/2012 1:56:12 PM
The Responsiveness of New Supply to House Prices
187
observe the whole supply as a flow, both the existing ones and the new one (Goodman, 2005); and, third, the fact that the supply function is local and specific to different regions, in many cases, a metropolitan area6 (Glaeser et al., 2005a; DiPasquale, 1999). This means that researchers use two definitions of housing supply, both the housing stock (DiPasquale and Wheaton, 1994; Whitehead, 2003a; Mayer and Somerville, 2000; Meen, 2001) or new units built (most of the research shares this focus, e.g. Mason, 1977; Malpezzi and Maclennan, 2001). This multiplicity of measures means that the elasticity obtained varies depending on which definition has been chosen. Recently some research has added complexity by estimating supply elasticities that vary over time, reflecting the swing-curves of supply which change its responsiveness depending on the different points of the cycle (Pryce, 1999; Bramley, 1993, 2003; Malpezzi and Vandell, 2002). Other researchers maintain that the difficulty in measuring supply comes from the lack of a standard price and quantity in the market because each housing unit varies in terms of its individual characteristics. The housing supply is also the result of a complex decision-making process in which both builders and homeowners participate, and there is very little evidence about how they interact (Hanushek and Quigley, 1979). DiPasquale (1999) summarised the problems related to the estimation of housing supply into two groups: The first one is the lack of suitable databases due to the information-related problems mentioned above. Analysts face problems linked to quality of the available data and use aggregated information instead of microdata, losing perspective of the local market and its equilibrium (Malpezzi and Vandell, 2002; Goodman, 2005). The second group of reasons reflects a deficit in the explanatory theory. Recent research evidence, for example, has highlighted supply variations as a result of the action of territorial factors which affect markets locally, like the climate (Fergus, 1999) or the spatial location (Goodman and Thibodeau, 1998). Finally, there is also an agreement about the existence of different conditions in the housing market that can promote quasi-monopoly or monopolistic competition (Green and Malpezzi, 2003). The potential market power by private or public actors can determine the degree of reaction of supply, normally increasing inelasticity through concentrated land ownership, reduced number of building firms, land uses under control, restrictive permit system etc. Others reasons such as the inflexibility of inputs, as well as the control that developers can apply on the production process (Coulson, 1999), can also affect the supply response to market signals. All these factors generate disparate, tangled-in-time supply responses (Goodman, 2005; Pryce, 1999) with greater dynamism when positive shocks occur than negative ones (Glaeser et al., 2005b). How to measure and model housing supply is a key issue for this chapter. Mayer and Somerville (1997) and Somerville (1999) maintain that the supply
Jones_c09.indd 187
2/7/2012 1:56:13 PM
188
Challenges of the Housing Economy
is a function of, first, changes in current and lagged prices, second, changes in input costs and, third, the current and lagged interest rates for building credits. For Goodman (2005) the supply is a direct function of the stock value and a list of characteristics, linked to the territory referred to as regional factors, such as the climate, and even family size, to name but a few, which determine the particularities of a town or city. Guirguis et al. (2005) contend that this market largely resembles an asset market and present long-run housing supply changes as a function of both the flow of dwellings that enter the market and the depreciation of existing units. According to Blackley (1999) the volume of new housing is measured as the value of (1) private residential construction in real terms, (2) the real price of construction materials, (3) the real salary of workers, (4) farming land, (5) nominal interest rates, (6) the expected inflation rate, (7) short-run real interest rates, (8) the real residential capital stock and (9) the real prices of non-residential construction. For Malpezzi and Vandell (2002) the supply is influenced by the set of construction costs as a whole, the regulations that affect the market, the existing stock and market conditions, such as vacancy rates. Malpezzi and Maclennan (2001) define supply as a function of changes in both the real and the desired stock, which is affected by the evolution of prices, population and transaction costs. In the opinion of Poterba (1991) new construction is a function of the revenues derived from housebuilding, which is reflected in the cost/price ratio, and of the depreciation rate. This approach is not the only one along these lines. According to Meen (2002b) housing supply flow is positively related to real prices, as a proxy for capital gains, and negatively to the size of the existing housing stock after reducing depreciation. For Mayo and Sheppard (2001) housing supply behaves in accordance with the earnings obtained by the developer, which in turn is a function of housing prices and construction costs, the former being dependent on the price of land in different locations. Construction constraints can become a determining factor for the supply and influence the degree of risk assumed by the developer. The inclusion of price and construction costs as input to the investment decision component can also be found in other studies by Pryce (1999) and DiPasquale and Wheaton (1994). Summarising the literature, it seems that there is agreement that prices act as a signal that encourages builders to start new housing developments, i.e. prices become an incentive for productive activity. This is the role commonly played by prices in any market and an approach to it from the perspective of the supply. If the supply side reacts to the price incentive, it could be said that the market performs within an efficiency level. Markets can have distinct degrees of efficiency depending on whether the market power exists with developers or landowners. There can also be a quasi-market or nonmarket, when incentives to build are other than prices and emanate from sources located outside the normal supply and demand dynamics.
Jones_c09.indd 188
2/7/2012 1:56:13 PM
The Responsiveness of New Supply to House Prices
189
Econometric model This study follows the models proposed by Goodman (2005), Meen (2003), Malpezzi and Maclennan (2001) and Glaeser et al. (2005b), where the housing stock supplied at a given moment in time could be defined as in (A1): Qst = f(PH,t , C t , H t −1, G kt , π H ) = eα1 PH,t α 2 Cm t α 3 Cs t α 4 i t α 5 H t −1α 6 [ηk G kt ]α 7 π H e α 8 ε t where: PH,t = housing prices in real terms Cmt = the costs associated with construction materials Cst = indicator of the payment of salaries to construction workers it = real interest rates paid by developers for building credits Ht−1 = the existing housing stock at the previous moment ηk Gtk is a matrix of the regional market characteristics, including physical features as well as other aspects like land and market size πHe = developers’ expectations regarding inflation εt = random term a1..8 are the estimated parameters. In the particular case of new supply, it can be written as Qt sn = eα1 PH,tα 2 Cmt
α3
Cst
α4
itα 5 (γ Ht,) α 6 [ηk Gt k ]
α7
π He α 8 μt
The expression (γHt,)α6 represents the existing units supplied on the market (vacant) presumed to be homogeneous in the long term. The component [ηk Gtk ]α7 captures the specificities of regional housing markets. The approach assumes differences in characteristics between regions (but homogeneity within regions). Following Meen (2003), πHe takes in factors related to expectations about prices. This model is transformed into the log-linear model set out in the main text following Coulson (1999) and Somerville (1999). The model uses aggregated data to estimate the behaviour of the housing supply. The source of residential prices is the Ministry of Fomento and data has been deflated using CPI (both aggregate and disaggregated by regions). The information on starts comes also from the same institution and it is called ‘municipal housebuilding permits’ in the original database. The construction costs are obtained from the same ministry distinguishing between costs associated with materials and those corresponding to labour. The real interest rates are calculated using mortgage market reference rates adjusted by the inflation existing both in each period and in each geographical region.
Jones_c09.indd 189
2/7/2012 1:56:13 PM
190
Challenges of the Housing Economy
The model therefore incorporates the real interest rates associated with inflation differences in each region. All the series except for housing prices have a monthly basis and have been transformed into quarterly series by means of addition. The functional form of the model has been selected after checks to the statistical features of the dataset. The existence of unitary roots in the pool is checked using the conventional tests (Table 9.A1). The IPSW test and ADF cannot allow rejection of the hypothesis of non-existence of individual unit roots, but the existence of a common one is fully rejected. The existence of cointegration among the variables is also checked (Table 9.A2) through the Pedroni Residual cointegration test for real prices and starts. Three out eight tests cannot reject the existence of cointegration and two out of three the individual correlation between dimensions. The residual cointegration test suggests the existence of a common AR(1) test among the variables which should be included in the analysis but not severe cointegration. To guarantee robust estimators of the elasticity the full model of new housing supply is applied using Pooled OLS (PLS) with a GLS specification assuming the presence of cross-section heteroskedasticity7 (called PGLS). A model for the full sample (Spain with all territory, 16 regions) is estimated including fixed effects in order to capture the spatial differences affecting to the supply equation among regions. The model is also applied for selected regions. The selection of those regions was made choosing areas from the North (Asturias), East coast (Valencia Community and Murcia), South (Andalusia), Interior (Castilla La Mancha and Aragón), main capitals (Madrid and Barcelona) and islands (Canary Islands) where the housing cycle was representative (Taltavull, 2000). The total pooled number of observations is 288.
Table 9.A1 Diagnostic tests for regression: Pool unit root test summary. Series: Log of real prices, log of starts, both by 16 regions. Sample: 1990Q1 2010Q1. Exogenous variables: Individual effect. Tests
Stat
Prob
Cross-sections
Num. Obs.
Null: common unit root process Levin, Lin & Chu t*
−3,106
0,0009
36
2755
Null: individual unit root process Im, Pesaran and Shin W-stat ADF – Fisher chi-square PP – Fisher chi-square
−0,140 82,565 113,31
0,444 0,185 0,001
36 36 36
2755 2755 2827
Jones_c09.indd 190
2/7/2012 1:56:15 PM
The Responsiveness of New Supply to House Prices
191
Table 9.A2 Diagnostic tests for regression: Pedroni residual cointegration test. Series: Log of starts and real prices. Sample: 1990Q1 2010Q1. Trend assumption: No deterministic trend. User-specified lag length: 1. Alternative hypothesis: common AR coefs, (within-dimension) Weighted
Panel v-statistic Panel rho-statistic Panel PP-statistic Panel ADF-statistic
Stat
Prob,
Stat
Prob,
3,958742 −2,702858 −1,867582 5,126537
0,0000 0,0034 0,0309 1,0000
4,957480 −2,131428 −0,388542 5,515064
0,0000 0,0165 0,3488 1,0000
Alternative hypothesis: individual AR coefs, (between-dimension)
Group rho-statistic Group PP-statistic Group ADF-statistic
Stat
Prob,
−1,657146 2,226960 9,072898
0,0487 0,9870 1,0000
Results for selected regions The results for selected regions are shown in Tables 9.A3 to 9.A7.
Table 9.A3 New housing supply ‘additive’ elasticities over time for Andalusia, 1994–2010.
Jones_c09.indd 191
Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.36 1.48 1.37 1.36 1.26 1.21 1.20 1.04 0.92 0.84 0.62 0.64 0.70 0.61 1.96
e shock adjusted
Adj R2
DW
F
0.68 0.86
0.18 0.23 0.25 0.37 0.57 0.68 0.79 0.80 0.82 0.85 0.85 0.87 0.89 0.85 0.82
2.08 1.96 1.90 2.01 2.03 1.93 1.99 1.95 2.01 1.97 2.02 2.05 2.04 2.22 2.29
1.36 1.48 1.37 1.36 1.26 1.21 1.20 1.04 0.92 0.84 0.62 0.64 0.70 0.61 1.96
2/7/2012 1:56:15 PM
192
Challenges of the Housing Economy
Table 9.A4 New housing supply ‘additive’ elasticities over time for Asturia, 1994–2010. Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0.82 0.54 0.54 1.09 1.25 1.14 1.16 1.11 1.07 0.83 0.84 0.60 0.58 0.53 0.39 0.32 (−0.21)
e shock adjusted
Adj R2
DW
0.53 0.62 0.26
0.35 0.11 0.10 0.05 0.14 0.16 0.47 0.63 0.55 0.51 0.50 0.48 0.52 0.53 0.49 0.43 0.53
2.58 2.28 2.09 2.05 2.02 2.05 1.88 1.96 1.95 1.91 1.97 1.97 2.02 2.01 2.05 2.27 2.46
Table 9.A5 New housing supply ‘additive’ elasticities over time for Castilla La Mancha, 1994–2010.
Jones_c09.indd 192
Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1.24 1.06 1.13 0.08 0.50 0.91 0.97 0.61 0.68 0.86 1.51 1.61 1.64 1.70 1.61 2.25 2.25
e shock adjusted
Adj R2
DW
1.70 2.00 2.28
0.19 0.16 0.02 0.08 0.05 0.00 0.07 0.41 0.54 0.62 0.73 0.82 0.87 0.90 0.90 0.85 0.85
2.24 2.28 1.95 2.16 2.17 2.25 2.28 2.18 2.16 2.21 2.29 2.33 2.30 2.29 2.23 2.69 2.51
2/7/2012 1:56:15 PM
The Responsiveness of New Supply to House Prices
193
Table 9.A6 New housing supply ‘additive’ elasticities over time for Cataluna, 1994–2010. Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0.60 0.51 0.76 0.60 0.56 0.35 0.31 0.36 0.19 0.20 0.31 0.45 0.58 0.69 0.69 1.67 3.75
e shock adjusted
Adj R2
DW
0.70 0.87 3.76
0.20 0.52 0.57 0.59 0.70 0.67 0.70 0.68 0.58 0.61 0.68 0.74 0.79 0.82 0.82 0.77 0.85
2.07 2.12 2.00 1.84 1.99 1.93 2.12 1.97 2.15 2.31 2.33 2.27 2.34 2.40 2.37 2.51 2.60
Table 9.A7 New housing supply ‘additive’ elasticities over time for Madrid, 1994–2010.
Jones_c09.indd 193
Year
e
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0.34 0.51 0.50 0.19 0.24 0.26 0.28 0.32 0.31 0.12 (−0.07) 0.19 0.15 0.19 0.06 0.16 (−0.10)
e shock adjusted
Adj R2
DW
0.19 0.28 0.03
0.19 0.24 0.36 0.30 0.35 0.41 0.46 0.44 0.44 0.46 0.45 0.50 0.51 0.56 0.47 0.50 0.46
2.05 2.10 2.19 2.15 2.17 2.17 2.18 2.07 2.09 2.12 2.12 2.11 2.10 2.10 2.00 2.07 2.16
2/7/2012 1:56:15 PM
194
Challenges of the Housing Economy
Notes 1. Land regulation is undertaken by local authorities in Spain, which have the responsibility to design and coordinate land uses. Valencia Community’s new Land Act in 1996 introduced an unusual flexibility allowing the municipality to approve new design and land definitions inside their boundaries with a faster administrative process. That system was subsequently included in the Land Rules of many of the other Spanish regions. 2. During 2010 the change on financial regulation related to Basilea requirements as well as the ECB recommendations converted the saving banks into banks. This transformation is still pending until the end of 2011. 3. The price-to-income ratio (PTI) is calculated inversely in Figure 9.5 in order to allow the joint representation with the other ratios. Inverse means that the value of 20% appearing in the graphic is equivalent to a PTI = 4 which is the long-term value of equilibrium for Spain. 4. Regions considered in this study are NUTS2 in the European classification. 5. A threshold of Adjusted R2 = 50% is taken to define areas where developer decisions are dominated by market signals. 6. There are in fact studies which estimated supply elasticity at an aggregate national level and by regions, obtaining very different elasticity results. For example, Mayer and Somerville, 2000, obtained an elasticity for new dwellings that overestimates with respect to the results of the calculation carried out in different local areas and underestimates the time required to respond to a price shock. 7. This method includes cross-section weights in a PGLS estimation; that is, it performs a first estimation obtaining cross-section-specific residual vectors then uses these residuals to estimate the cross-specific variances and uses them in a weighted least squares procedure to form the feasible GLS estimates. Weighting is made in iterative form, repeating this process until the coefficients and weights converge. Models compute robust covariance using PCSE – Panel Corrected Standard Error.
Jones_c09.indd 194
2/7/2012 1:56:15 PM
10 The UK Housing Market Cycle and the Role of Planning: The Policy Challenge Following the Financial Crisis Colin Jones
In the UK, a complex paraphernalia of planning structures has been created to control and shape housing development. Central to this approach until recently was the role of regional planning authorities undertaking strategic housing assessments to decide on the scale of new development needed. In addition, local authorities enter into partnerships with house builders in the form of ‘planning agreements’ to provide affordable housing. Nevertheless the number of houses built has not responded fully to rising household numbers over the past two decades. A discussion paper published by the last government in 2007 notes that ‘For a generation, the supply of new homes has not kept up with rising demand.’ (CLG, 2007). This statement sets the context for this chapter as it encapsulates the housing problem of the UK. Added spice to this quote is that the level of housebuilding in 2010 in England, only three years later, was at its lowest since 1924. The challenge for planning is to solve the weak housing supply response in a period of what is at best likely to be modest macroeconomic growth and, initially at least, tight financial constraints on lending to builders and would-be house buyers. This chapter assesses this question by reference to the fundamentals of housing market change and the role of planning over the past 20 years.
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c10.indd 195
2/7/2012 10:23:03 AM
196
Challenges of the Housing Economy
The chapter begins by setting out the context of the UK housing market as a prerequisite. It then examines the anatomy of the recent history of the long boom and bust of the housing market, focusing on the role of mortgage finance and the reasons for the lack of a significant supply response to the rise in real house prices. The next section reviews the evolution of the planning system, explaining its increasing importance to housing policy, and its (changing) relationship to market forces. The following section looks at the principal planning policies of the recent decade and the market outcomes, and based on these insights it then looks forward to the role of planning in the market downturn and beyond. In completing this task it also reviews the influence of planning on the UK housing market cycle and hence the economy.
Housing market context The overall tenure structure of Great Britain is dominated by the owner occupied sector accounting for, according to the most recent statistics available, 69 per cent of the stock in 2007 compared with 12 per cent for the private rented sector (PRS) and 18 per cent for social housing. Social housing comprises stock owned by local authorities, council housing and housing associations. The tenure structure is not cast in stone for although owner occupation is on a plateau the PRS expanded rapidly over the decade of the noughties by the order of 40 per cent. Social housing declined by 9 per cent overall during this period, even though the housing association stock increased by 72 per cent because council housing lost 40 per cent of its stock. By the end of the past decade the numbers of housing association stock had just overtaken council housing. Behind these statistics lie a number of explanations. The owner occupation rate grew significantly over the last century, especially in the latter half, until stabilising over the past decade. The rise was a consequence of rising housing aspirations brought about principally by increasing real incomes, and a subsidy/tax system that favoured owner occupation over renting whether it was in the private or public sector. A major stimulus to the growth of homeownership was given in 1980 with the gradual privatisation of much of the council housing stock through the ‘right to buy’ (RTB) policy, which allowed council tenants to buy their home at a discount/subsidy (Jones and Murie, 2006). The introduction of the RTB signalled the end of the post World War II growth of social housing. This was also because the year 1980 proved a watershed for the building of social housing with the end of massive building programmes that were started to address the housing shortages of the 1950s and slum clearance in the 1960s and first half of the 1970s.
Jones_c10.indd 196
2/7/2012 10:23:03 AM
The UK Housing Market Cycle and the Role of Planning
197
The subsequent decline in council housing as a consequence of a combination of minimal new building and the RTB, has been augmented by a recent policy of stock transfer from local authorities to housing associations (Mullins and Pawson, 2010). The revival of the PRS emanates from the removal of rent regulation in 1989 and was supported by the phasing out and ultimate abolition of subsidies via tax relief on mortgages to owner occupiers at the turn of the millennium by the Labour government. The rise of the PRS was also encouraged by the freer availability of mortgage finance and rising house prices, making it an attractive investment. It has also been partially stimulated by rising demand from households constrained by the paucity of social housing and priced out of their desired owner-occupied sector, or at least forced to delay entry (see later). These trends have manifested themselves in a surge in the number of new houses sold to private investors for renting, especially city centre flats, but the overwhelming number of new houses built since the early 1980s have been built for sale.
Anatomy of the housing market boom and bust The UK experienced a long and sustained increase in house prices from the mid 1990s lasting 12 years until the impact of the credit crunch brought the boom to an abrupt end at the end of 2007. This boom and bust can be seen as the latest episodes in a cyclical housing market that saw significant upswings in house prices during the early 1970s, the late 1980s and downturns at the beginning and end of the 1980s. A detailed review of these cycles up to 2007 is given by Jones and Watkins (2009) that examines the underlying factors and the impact on affordability. The most recent house price boom was also particularly associated with the easing of mortgage constraints making funds to buy more available as demonstrated by increasing loan-to-value ratio, loan-to-income ratio and mortgage length. The residential mortgage market became more and more sophisticated, offering not just the traditional variable interest mortgages over 25 years but also a range of new products including short-term discounted and fixed interest rate loans and even ‘110 per cent’ mortgages.
Rise and collapse of mortgage finance Given the central role of mortgage finance in the cycles of the housing market over the past two decades it is useful to consider the parallel dynamics of the banking sector over this period. The analysis also provides perspective on the future of mortgage support to the housing market. The story of the boom and bust in the UK is paralleled by the expansion and the collapse of
Jones_c10.indd 197
2/7/2012 10:23:03 AM
198
Challenges of the Housing Economy
the banking system and can be usefully seen through the well-documented example of the Northern Rock Bank. The bank was originally the Northern Rock Building Society, a solid mutual organisation, that could trace its history back to the mid 19th century. The Northern Rock Building Society was formed in 1965 as a result of the merger of two building societies and then over the next 30 years or so it gradually accelerated its expansion by taking over 53 smaller building societies (Northern Rock, 2011). Eventually in 1997 it converted to a bank and was quoted on the stock exchange. Two years later it changed its business strategy away from one dependent on retail deposits toward wholesale funding to promote further and even more rapid expansion (see chapter 1). The result was that its retail deposits fell from 62.3 per cent of total liabilities in 1997 to only 22.2 per cent by the end of 2006 (House of Commons, 2008). The shift from the traditional model to securitisation enabled the bank to expand its lending dramatically and it tripled its share of the UK mortgage market between 1999 and 2007 from 6 per cent to 19 per cent (Eisenbeis and Kaufman, 2007). Its assets over this period grew six fold up to the end of 2006, with the process accelerating so that the bank’s assets doubled between 2005 and 2007 (House of Commons, 2008, Llewellyn, 2009). The bank also achieved a substantial year-on-year rise in its profits. The strategy was heavily dependent on short-term wholesale funding which needed to be constantly rolled over to replace maturing bonds on the international money and capital markets. The business model was, in hindsight, seen as too reckless (House of Commons, 2008). A substantial portion of this wholesale funding was in maturities of less than a year when the credit crunch began to bite at it. The difficulties in refinancing this debt led to the bank seeking support from the Bank of England, and the uncertainty that this created resulted in the first run on a British bank since 1878. Queues formed outside its branches as investors sought to retrieve their funds on 14 September 2007, and its website collapsed and phone lines were jammed. At the root of the difficulties was the impact of the wholesale funding crisis that was just emerging, namely that banks could not decipher which mortgage-backed bonds were worthless because they were supported by subprime mortgages. The international markets for these securities froze at the beginning of August in that year. This problem emanated from the US market but, with such mortgages not confined to the USA, investors at the time were unsure of the spatial scale of the problem. And there were doubts expressed about Northern Rock’s asset base because of its past aggressive lending that had routinely included the offer of 100 per cent mortgages (albeit with insurance). Many newspapers criticised the bank for its irresponsible lending and it had also advertised subprime mortgages (Congdon, 2009). However, it had only introduced subprime borrowers to a third party,
Jones_c10.indd 198
2/7/2012 10:23:04 AM
The UK Housing Market Cycle and the Role of Planning
199
Lehman Brothers, and did not hold such subprime loans on its balance sheet. Furthermore its arrears were also below the industry average (House of Commons, 2008). Even if these concerns were unfounded and it had an apparently solid asset base (the problem assets only began to unravel later during the recession) in a sense they were irrelevant. It was the business model that scuppered Northern Rock and it became the first British victim of the international liquidity squeeze. After the Bank of England provided a loan/temporary guarantees and failed attempts to sell the stricken bank, it was nationalised in January 2008. Its expansive business model may have been reckless and extreme but it was not that different from most of the much larger high street British banks (Onado, 2009). The Northern Rock was simply in the vanguard of a financial tsunami, and leading banks were also eventually to succumb to the same global banking liquidity constraints. Between 2000 and 2007, the total amount outstanding of UK bank wholesale funding via residential mortgage-based securities (and covered bonds) rose from £13 billion to £257 billion. The growth in the amount outstanding amounted to £78m in 2006 alone (Crosby, 2008). The result was that such securitised credit which had played only a minor role in the 1990s rose rapidly in significance, from 2.5 per cent in 1997 to 27 per cent of UK lending in 2005 (Turner, 2009). Bradford and Bingley, another small mortgage bank which had specialised in financing buy-to-let landlords using wholesale funding, was the next to fall victim, and was similarly nationalised in September 2008. The demise of these small banks was only a prequel as the ‘cancer’ soon spread to virtually all of Britain’s principal banks in an unprecedented attack on the financial system. Subsequently, in October the government took substantial equity stakes in two of Britain’s largest British banks, Lloyds/HBOS and the Royal Bank of Scotland, to avoid the collapse of the UK financial system. As a result, British taxpayers own 43 per cent of the former and 57 per cent of the latter. Another large bank, Barclays Bank, was able to resolve its equivalent problems by attracting private equity funding from the Middle East, albeit at a higher price than that offered by the government. This was the beginning of a general re-capitalisation programme of all the high street banks in the UK that has had important ramifications for mortgage lending. Since this rescue of UK high street banks with various forms of government financial support, they have been rebuilding their capital bases and liquidity. The priority of the banks is to rebuild the probity of their balance sheets to allow them to be eventually denationalised. As a result, banks have undertaken very little lending, not just for mortgages but also general business lending to firms, etc. The weak performance of the economy and the residential market has also mitigated against an expansion of lending.
Jones_c10.indd 199
2/7/2012 10:23:04 AM
200
Challenges of the Housing Economy
Table 10.1 Quarterly seasonally adjusted residential transactions in UK, 2006–2011. Year
Q1
Q2
Q3
Q4
Total
2006 2007 2008 2009 2010 2011
401 443 301 173 221 218
404 413 254 191 220 204*
423 406 187 222 224
442 355 176 264 215
1670 1617 918 850 880
Source: HM Revenue and Customs. Note: Only transactions with a value above £40,000. * Estimate.
Housing market downturn and beyond The importance of the influence of loan finance is emphasised by how the subsequent mortgage famine that followed the financial crisis has been a central element in the dynamics of the housing market since late 2007. The mortgage lending ‘tap’ was swiftly turned off and this especially hit firsttime purchasers and buy-to-let landlords by changing the loan-to-value terms of mortgages available. The minimum required deposits for mortgagors rose literally overnight from zero to 25 per cent (Bank of England, 2010). Gross mortgage lending for house purchase fell by two-thirds over the next 15 months before stabilising (CML, 2011). The implications for market transactions are shown in Table 10.1 with the total more than halving from the peak of the market through to the beginning of 2009 before showing modest growth again. The number of transactions in 2011 still remains at just under half the figure of 2007. Broadly the same pattern applies with respect to the number of first time purchasers. Their number was already falling in 2007, but 2008 saw a drop of 46 per cent before a modest increase of 3 per cent in 2009. The trend has levelled off, with mortgages for first-time purchasers in 2010 broadly the same as the previous year (Halifax, 2010). Buy-to-let landlords had been a major driver of the housing market but they too were impacted severely by the credit crunch. Much of the funding had come from specialist lenders such as Paragon and the Bradford and Bingley that had depended on wholesale funding which was no longer available. The number of gross advances for these investments declined by 73 per cent from 346,000 in 2007 to 93,500 in 2009, while the value of these loans fell by 81 per cent from £44.6bn to £8.5bn (Paragon, 2011). The shift in lending policies was stark, and the dramatic turn-round in bank lending had inevitable consequences for the housing market in the final quarter of 2007 with price growth stalling. Prices began to fall at the
Jones_c10.indd 200
2/7/2012 10:23:04 AM
The UK Housing Market Cycle and the Role of Planning
201
beginning of 2008, falling in nominal terms, over five quarters to the first quarter of 2009, by 21 per cent according to the Halifax index (other indices differ marginally) (Halifax, 2011). The house price time series mirrors that of transactions. The short-term consequences of the economic downturn/house price falls did create adjustment problems. In the boom, many homeowners and buyto-let landlords had entered into mortgages on short-run, low fixed interest rates with a view to renegotiation on completion of the term but found the credit criteria and mortgage offering dramatically different in the post credit crunch era. Some were forced to sell, thereby adding a deflating influence to the housing market. There were also financial problems for house builders struggling to pay interest payments on land bought at the peak of the boom but with houses left unsold. Since the spring of 2009 the housing market can best be described as following a bumpy path, albeit with only modest changes in average nominal house prices. From the floor in house prices at the beginning of 2009, prices rose over the next year by 6.6 per cent. In the autumn of 2010 this upward trend was arrested with falling confidence in personal finances following the government’s announcement of forthcoming fiscal austerity measures including tax rises, cuts in different forms of social security payments (including child benefits) and redundancies in the public sector. This inevitably had a spill-over effect on the housing market, and prices fell away modestly before flat-lining in the first six months of 2011. The average house price at the time of writing at the end of the second quarter of 2011 is approximately 19 per cent below that of the 2007 peak and the market remains in a period of uncertainty. This downturn in the housing market occurred across all regions of the UK following the credit crunch although the precise timing and magnitude varied. It is sometimes portrayed as a stern but required and inevitable ‘correction’ to inflated market prices. The implication is that the boom was the consequence of a financial bubble caused by over-relaxed bank lending as exemplified by excessive subprime lending. However, the reality does not fully bear this out. The market trends from their peak around the end of 2007 suggest that this view is exaggerated. Notwithstanding the mortgage constraints, the supply of new housing available initiated in the boom and the reduction of the number of housing transactions to a low ebb, the immediate fall in house prices was limited. House prices experienced an initial relatively quick recovery, and nominal house prices are back at their 2006 levels after reaching a trough in the spring of 2009, broadly some one-fifth below their level at their peak. The major constraint on recovery during this period was the mortgage famine. The mortgage environment has therefore changed in the aftermath of the credit crunch and we are now in a period of a much more conservative
Jones_c10.indd 201
2/7/2012 10:23:04 AM
202
Challenges of the Housing Economy
approach by banks to risk assessment. This could be a fundamental regime change but is more likely to be simply part of a credit cycle. In the past, mortgage banks have collectively under-priced risks in the upturn of a cycle with very generous credit rationing rules and overpriced it during the downturn, lowering the mortgage terms available. Given the need to rebuild their capital base this process of adjustment may take longer than in past cycles. Despite the dampening effect of the severe mortgage rationing and the depressed nature of the economy real house prices rose in 2009–10 although they have fallen back a little since. This pattern compares with the previous recession in 1989 where prices fell by more than a quarter over three years in the southern regions (and very little elsewhere) with real house prices only beginning to rise after 1995. The current house price plateau is despite the Nationwide First Time Purchaser Affordability Measure showing that mortgage payments relative to take-home pay for these buyers are at least 50 per cent above that in the mid 1990s in all regions (Nationwide, 2011). This relatively quick stabilisation following the sharp price fall through 2008 compares with the very slow and sluggish market recovery response from the late 1980s downturn. It is partly down to a different policy response – including constraints on bank foreclosures, some support by the government for mortgage lending (special liquidity scheme) and a stamp duty holiday – but nevertheless raises questions about the simple overheating price bubble thesis. The nature of the short-term dynamics in the housing market suggests that there are also underlying factors at work that in particular have created a long-term fundamental imbalance of supply and demand which has seen the growth in the number of households outpace net additions to the housing stock. The recent evidence supports the thesis of Jones and Watkins (2009) that there is a long-term disequilibrium driven by forces internal to the housing market. In particular they argue that the principal cause is the lack of housebuilding which in turn is a product of the interaction of the planning and housebuilding industry.
The supply response By the peak of the housing boom at the end of 2007 affordability of homeownership was at an all-time low. Affordability is a complex issue with different facets and difficult to measure (Jones et al., 2011) but all indicators demonstrated a substantial deterioration in the noughties across all regions of the UK. The implications for the market was the pricing out of the market of many young households, and the contribution of first-time purchasers fell from a typical half of all transactions to less than 40 per cent (Jones and Watkins, 2009). The consequences were a rise in the age of entry to the owner-occupied sector and increased demand for the PRS. The impact
Jones_c10.indd 202
2/7/2012 10:23:04 AM
The UK Housing Market Cycle and the Role of Planning
Table 10.2 Year 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10
203
Housing completions in UK since 1990. Private sector
Registered social landlords*
Local authorities
162,210 160,620 144,420 146,820 156,250 156,940 153,450 160,680 154,560 160,490 152,590 153,310 164,300 172,360 184,500 189,700 192,150 187,930 139,270 113,860
19,340 21,130 30,160 36,670 37,600 38,550 30,950 28,550 22,870 23,170 22,250 20,400 18,610 18,020 21,990 23,990 26,650 28,510 31,860 32,410
16,380 9,900 4,430 3,590 3,000 3,040 1,540 1,520 870 320 380 230 300 210 130 330 260 250 850 680
Source: CLG Live Housing Table 209. Accessed August 2011. * RSL tenure includes social rent, intermediate rent and low-cost homeownership RSL new build dwellings.
of these affordability difficulties is given by the following statistics: in the period 1988 to 2004 the percentage of young people aged 25–29 that are owner occupiers fell from 64 per cent to 50 per cent while the equivalent percentage of private tenants rose from 16 per cent to 31 per cent (Andrew, 2006). A key factor in this market outcome was the slowness of new supply to respond. The housing market experienced a divergence from a typical property market cycle – supply did not respond to rising prices and more and more first-time purchasers were priced out of the market. In previous house price booms the increase in house prices has led, after a lag, to an increase in the number of houses built speculatively for sale, but Table 10.2 shows that private sector new house completions fell away in the early 1990s following the recession. Although annual house price inflation began to rise in the late 1990s new housebuilding completions remained on a plateau until 2003. There was then a brief boost to annual UK housebuilding numbers until a peak in 2006/07 but numbers fell to their lowest level in living memory in 2009/10, and has continued this downward trend subsequently. An analysis of 12-month rolling time series of starts by (HBF, 2011) concludes that the industry has settled on an even lower level of production of between 105,000 and 110,000 homes per year.
Jones_c10.indd 203
2/7/2012 10:23:04 AM
204
Challenges of the Housing Economy
The reasons for the weak supply response are to a degree disputed, or at least the responsibility for it. Part of the explanation lies in NIMBYism and the defence of the countryside and this is certainly a continuing concern for many of the British public (Evans and Hartwich, 2005; Jenkins, 2011). In particular, planning has become associated with defending the countryside which is seen as under threat from urban development. A key question is the relationship between household demand, planning, housebuilding and supply outcomes in the market. The output of the housebuilding industry has been subject to a range of criticisms about its lack of dynamism. It is dominated by a small number of major developers but has been through a period of the turbulence in its corporate structure (Wellings, 2005). It is essentially conservative, reflecting the speculative nature of the business activity with substantial outlays on land and construction costs in advance of potential sales in a cyclical market (Ball, 1999; Barlow, 1999). Perhaps as a consequence the industry is viewed as concerned more with margins than volume (House of Commons, 2007). There is a lack of competition at a local level but this is partly because of the limited amount of land with planning permission for housing (Barker, 2004; OFT, 2008). It can therefore be argued that the housebuilding industry operates within a framework set by planning and has adapted it activities accordingly. Furthermore, the planning system needs to take account of the behavioural response of the housebuilding industry in making its decisions about land allocated for housing. Planning itself has been subject to a longstanding chorus of criticism about its lack of a housing market perspective in a string of papers by different authors (see, for example, Cullingworth, 1997; Monk, 1999; Blackaby, 2000; School of Planning and Housing, 2001; Evans, 2003; O’Sullivan, 2003). Jones and Watkins (1999) in particular argue that planning suffers as a result of systematically taking no account of the spatial dynamics of local markets. Following the government-sponsored Barker report (2004), a commitment was made to take account of market signals in housing planning. The next section looks in detail at what this has meant in practice, placing it in the long-term context of the evolution of British planning. Before proceeding, it is useful to note that the government planning discussion paper referred to above, commenting the scale of housing shortages, argued for a new national drive to support more affordable housing. It set out the then government’s plans to build three million new homes by 2020 (CLG, 2007). Only around a fifth of this target was to be directly met by more social housing and at least a further tenth was to come from other forms of ‘affordable housing’ for rent or sale (CLG, 2007). ‘Affordable housing’ refers to both social and ‘intermediate’ housing. The latter encompasses homes sold at a discount to market value, various forms of shared equity, part buy/part rent of which shared ownership is the most well established,
Jones_c10.indd 204
2/7/2012 10:23:04 AM
The UK Housing Market Cycle and the Role of Planning
205
and properties let at below market rent to key workers (but above rents of social housing – CLG, 2006a; Monk and Whitehead, 2010). In other words the new housebuilding would primarily be in the private sector, reflecting the current tenure structure.
Evolution of the planning framework in the UK The planning system in the UK was established in 1947 and it has followed broadly the same principles since, exhibiting a relatively high degree of continuity, although the context in which it operates has changed substantially (Cullingworth, 1997; Brindley et al., 1996). Initially the emphasis of specific policies was aimed at solving the ills of overcrowded cities and centred on restraining urban growth and channelling new development into new and expanded towns. The overall system is comprehensive and gives local authorities the responsibility for the production of strategies for the use of land and development control. The main attribute of the system is its flexibility and discretionary nature in the granting of planning permission. Nevertheless the precise role of local/urban plans has been periodically reconfigured, partly linked to the ideological stance of the government in power. Until 2010 a two-tier planning system in cities had broadly been in operation in different guises since the late 1960s, although not universally applicable across the UK. A key function of plans is to ensure sufficient provision of land for new development and a critical input was the assessment of local housing land requirements. The system generally has a top-down approach to this task although the mechanisms vary in different parts of the country. The process has been increasingly sophisticated but essentially involves assessing future local housing needs based on population forecasts that is then fed into assessing annual housing provision requirements. As the role of social housing has diminished as a result of a combination of the RTB and of the minimal new building over three decades (see Table 10.1), so the planning system has become more pivotal to housing policy (Bramley, 1997; Malpass, 1999). At the beginning of the noughties the frictions this created led to reforms to replace the plan-led ‘predict and provide’ system with a ‘plan, monitor and manage’ approach that ostensibly encouraged planning to steer the market rather than follow it (Prescott, 2000). The changes were also bound up with a new vision for planning entitled ‘spatial planning’, a term adopted from continental Europe, that supposedly signalled a move from a passive to a positive approach to land use planning. It is also a time that unresponsive housing supply was identified as a major housing policy challenge (Bramley, 2007). The issue partially re-emerged on the policy agenda in the context of the Sustainable Communities Plan
Jones_c10.indd 205
2/7/2012 10:23:04 AM
206
Challenges of the Housing Economy
Housing Market Areas (HMAs) Determination of functional areas for planning
Strategic Housing Market Assessments (SHMAs) Assessment of housing market conditions in HMAs, focusing on demand
Strategic Housing Land Availability Assessments (SHLAAs) Assessment of the deliverability of new housing supply, including land and market conditions
Figure 10.1 Regional housing planning framework in England, 2006–2010. Source: Ferrari (2008, 8).
(ODPM, 2003) and then more fully in an independent government review by Kate Barker (2003, 2004). This review sets out a plethora of recommendations (Barker, 2004) aimed at changing the planning system. There is a focus on the need for greater use of market indicators as a basis of providing land for future development. In particular, she suggests that the traditional approach of allocating land on the basis of household projections (and related needs estimates) could be improved by setting targets for planning that are based on affordability measurements. In 2006, the planning system in England was reformed to speed up the system and make it more sensitive to market pressures. The new approach was firmly in the top-down tradition. Two main levels of plan were introduced: regional spatial strategies (RSS) and local development frameworks (LDF). Regional planning bodies took responsibility for preparing, monitoring and implementing the RSS which involved the identification of both a rolling five-year supply of developable land and up to a further 15 years of potential housing land. It was thought that this system should provide market and state actors with greater certainty (NHPAU, 2007). This complex system of allocating land for new housing involved numerous stages and inputs. At a general level, assessments were made of housing need, and future requirements for new housing were set out. This technocratic exercise outlined in Figure 10.1 was based on a range of background data including projections of future numbers of households and assessments of the capacity of regions and subregions to provide for this additional requirement. In practice, these technical processes were subject to serious criticisms as they are data driven and of debatable theoretical basis (Jones and Watkins, 2009). The problems started with the essential building block, the definition of local housing market areas. These functional areas were often simply taken to be convenient local authority boundaries and lack credibility
Jones_c10.indd 206
2/7/2012 10:23:04 AM
The UK Housing Market Cycle and the Role of Planning
207
(Jones et al., 2012, forthcoming). The result was apparent sophisticated forecasts, but scratch the veneer and their reliability was questionable. Perhaps the process was more important than the results as it promotes engagement of a wide range of actors, including house builders, social landlords and property agents as well as community and regeneration agencies. And while it was a flawed pseudo-scientific exercise it offered a vehicle for justifying increased housebuilding and a mechanism for the identification of land supply. Irrespective of these criticisms these processes took time to set up and implement and this may have contributed to the slowing of new house supply over the past decade. However, the impact of planning on the level of housebuilding is not simply through the (passive) allocation of land supply, but also through specific planning policies.
Recent planning policies Spatial planning sets itself extravagant goals or claims about what it can achieve. Planning Policy Statement 1 issued in 2006 for England stated that ‘planning aims to ensure the right development, in the right place at the right time’ (CLG, 2006b) and the planning system was also now charged with the delivery of affordable homes, establishing sustainable communities (and securing local economic development) (Prior, 2005). In the light of the foregoing analysis of the housing market it is useful to briefly review the effectiveness of the most important of these policies and their impact.
Provision of affordable housing The previous government strongly encouraged the use of ‘planning agreements’ by local authorities in England that oblige private housing developers to make social contributions (a negotiable development gains tax) in return for planning permission to build (CLG, 2006a). In the noughties the government chose to expand the supply of affordable housing by using this mechanism. The use of planning agreements to provide affordable housing placed it as the central core of national housing strategy. Contributions by developers do not only encompass housing but can also include wider infrastructure and community needs, and are legally enforceable (Campbell et al., 2000; Crook et al., 2006). There are various mechanisms for the delivery of affordable housing in this way that can incorporate the direct provision by the developer, making land available free for a registered social landlord and financial contributions. In fact, the use of planning agreements to promote affordable housing has ostensibly been honed and developed over the past two decades, supported by planning guidance that has evolved to give them more ‘credibility’.
Jones_c10.indd 207
2/7/2012 10:23:04 AM
208
Challenges of the Housing Economy
However, there have been three problems with this policy. The most fundamental was that it failed to address the conundrum of using the planning system to generate affordable housing while the affordability problem is partly created by the constraints of the planning system. Setting aside this conflict it follows that this policy of planning agreements was only likely to be applicable where there are severe planning/urban growth constraints and would not be a panacea across the country. Second, the assessment of a developer’s contribution is a complex task and planning officers do not have either these skills or sufficient negotiating skills (Crook and Whitehead, 2002). Partly because of this problem the processing of these agreements, and hence the developments themselves, became very slow and cumbersome. There are also considerable doubts about the efficacy of the system in generating funds, albeit in kind. Third, it was decided that to promote social mix, the affordable housing had to be normally provided on the same site as the private housing. This last ingredient added to the complexities. A major area of friction was the persistent differences between developers and planners about the type and location of affordable housing to be built (Rowlands et al., 2006). The success of the policy is also very dependent on the buoyancy of the private housing market and new housebuilding rates. It is a function of the willingness of builders to build houses for sale and in the first half of the noughties the numbers of affordable housing units provided in this way did increase (Crook et al., 2006). There is a circularity here. The accelerating house prices (with a relatively weak supply response) led to higher land values, and hence more potential leverage of land values as planning obligations (Crook et al., 2008). The dependence of affordable housing supply on planning agreements and the inherent problems are reflected in the regional pattern of houses built by registered social landlords (housing associations) shown in Table 10.3. Only in London and South East regions are there significant rises in the noughties; in other words success was heavily concentrated in areas of high housing demand/value. There were very few affordable housing units provided in northern regions of England where prices were lower. Similarly McMaster et al. (2008) showed that over 2004–07 very little affordable housing was provided through planning agreements in Scotland, where again demand pressures were less intense. Nevertheless there is an argument that the application of planning agreements has given registered social landlords access to land in areas of high demand – possibly the only way they could (Crook et al., 2006). The dependence on planning agreements therefore has a number of drawbacks. As a tool for providing affordable housing they promise more than they deliver. This is illustrated by a study of rural and semi-rural areas by the Countryside Alliance (2011), which found that in such local authorities
Jones_c10.indd 208
2/7/2012 10:23:05 AM
The UK Housing Market Cycle and the Role of Planning
Table 10.3
Year 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10
209
Regional housing association building completions in England 1999–2010.
Yorks North North HumberEast West East of South South East West side Mid-lands Mid-lands England London East West 660 790 660 510 100 330 490 420 480 420 660 840
2280 2410 1910 1760 930 740 650 760 900 660 870 900
1540 1160 820 810 540 370 370 270 490 510 650 880
1270 1270 940 690 700 590 800 1200 1300 1370 1430 1740
2230 2210 1800 1140 1310 1190 1160 1260 960 1550 1640 1660
2410 1700 1640 1450 1170 1550 2100 2620 2950 3340 4290 3480
3360 3280 4080 3420 4080 4320 6180 5200 8320 7750 6950 6920
3640 2880 2800 2900 2760 3280 3400 4610 4480 5090 5920 5710
1500 1580 1780 1420 1500 1300 1520 1830 1870 2410 3110 2620
Source: CLG (2010a).
only just over a fifth of the estimated annual affordable housing target was to be built in 2010/11. From a wider perspective the use of planning agreements slow down the process of housebuilding because of the necessary negotiations, and potentially discourage private development. As a consequence, they dampen the upswing in housebuilding expected, with a rise in house prices and therefore contribute to exaggerating the house price cycle.
Sustainable development Sustainable development is the stated overarching goal of planning in the UK yet the term is subject to interpretation. The planning profession has emphasised environmental and ecological concerns, and both the government and the planning profession have placed a strong emphasis on urban design as part of ‘sustainable development’ within a compact city form (Urban Task Force, 1999). This has been translated into a policy that encouraged high residential densities by following a stringent defence of the green belts surrounding the cities, a minimum recommended development density of 30 dwellings per hectare in England (since 2006, although recommendations were higher between 2000 and 2006) and the reuse of brownfield land. There has been a target in England of building 60 per cent of new homes on reused urban land. The effectiveness of these policies can be gauged by Table 10.4 which demonstrates how sacrosanct the green belt has been in every corner of England. The protest rhetoric of the political lobby against such development expressing concerns about how long we will be able to enjoy the pleasures of our green and open spaces has been very effective (Das, 2011).
Jones_c10.indd 209
2/7/2012 10:23:05 AM
210
Challenges of the Housing Economy
Table 10.4 Percentage of new dwellings built within designated 2008 green belt by region in England. North North Year East West 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
1 1 1 1 0 1 1 1 1 1 0 0 0 0 1 1 1 1 1 1 1
2 3 2 2 2 3 3 2 3 3 4 7 2 4 4 2 4 3 2 3 3
Yorkshire East West East of South South and Humber Midlands Midlands England London East West England 1 3 2 3 2 3 3 2 4 4 4 4 4 6 5 3 2 2 3 3 5
1 1 0 1 1 1 1 1 1 1 0 2 1 0 1 1 1 1 1 1 1
5 5 2 2 2 2 3 3 4 4 4 5 4 5 5 3 4 3 2 4 3
1 2 2 2 3 2 2 2 3 4 5 6 6 3 4 3 2 3 2 2 2
0 1 0 1 0 0 1 1 1 4 4 5 2 3 2 1 1 1 1 1 0
3 3 3 2 2 2 3 2 3 3 5 5 5 3 3 3 2 3 2 3 3
1 1 1 1 0 1 1 1 1 1 1 1 2 1 1 0 1 0 1 1 1
2 2 2 2 1 2 2 2 2 3 3 4 3 3 3 2 2 2 2 2 2
Source: CLG (2009).
The consequences of these policies were seen in a transformation in the nature of housing development over the brief period of just five years as Table 10.5 demonstrates. The contribution of flats to new housebuilding in England rose from a fifth of the total at the beginning of the 1990s to 46 per cent by the end of the noughties. If social housing completions are added, the proportion of flats increased to half in 2008/09 This move toward flats built for sale induced a major reduction in the proportions of three- and four-bedroom housing units built, in favour of smaller units, especially twobedroom properties. The data for the latest financial year shows a minor reaction to this, probably reflecting the internal market pressures of the downturn (see below). The building of so many small flats, especially in city centres, raises longterm questions for the housing market and sustainability (see later). However, in the short term this trend was facilitated by the emergence of buy-to-let private landlords on a large scale who bought up many of them. At the same time, over the past decade as the house price boom increased the unaffordability of house purchase, many households have had to delay homeownership and there was a growth of long-term private tenants (as discussed above).
Jones_c10.indd 210
2/7/2012 10:23:05 AM
The UK Housing Market Cycle and the Role of Planning
211
Table 10.5 Percentage breakdown of private housebuilding in England by type and size, 1991–2010. Year 1991–92 1992–93 1993–94 1994–95 1995–96 1996–97 1997–98 1998–99 1999–00 2000–01 2001–02 2002–03 2003–04 2004–05 2005–06 2006–07 2007–08 2008–09 2009–10
Flats
1 Bedroom
2 Bedroom
3 Bedroom
4 Bedroom or more
21 19 16 14 15 13 12 14 15 17 22 26 33 40 45 44 45 46 40
15 13 11 9 8 6 5 5 5 5 6 6 8 9 9 9 10 12 11
31 31 30 29 28 26 24 24 24 24 24 28 32 37 41 40 41 41 35
31 33 35 36 36 37 38 37 36 35 31 30 29 29 28 28 27 26 29
23 22 24 26 28 31 32 35 35 36 39 36 32 25 22 22 22 21 25
Source: CLG live housing statistics 254. Accessed August 2011.
Table 10.6 Proportion of new dwellings on previously residential land in different regions of England, 1994–2009. Region North East North West Yorkshire and Humber East Midlands West Midlands East of England London South East South West England
1994–1997
1998–2001
2002–2005
2006–2009
9 7 9 8 10 14 11 16 13 11
5 9 11 9 11 16 15 19 16 13
7 10 16 11 15 18 18 26 22 17
15 18 22 21 23 26 26 32 29 25
Source: CLG land use change live statistics. Accessed August 2010.
The combination of land constraints and planning policy had increased development densities in cities. Much of this intensification of land utilisation incorporated redevelopment of existing housing including their gardens. Table 10.6 shows that the recycling of residential land rose by around 50 per cent over the latter half of the noughties. This is emotively referred to as ‘garden grabbing’. This reflects the increasing drive toward higher
Jones_c10.indd 211
2/7/2012 10:23:05 AM
212
Challenges of the Housing Economy
density across the whole of England. Regions where there are high demand pressures are consistently above the average. Looking at the local authority level statistics it is clear that they reflect not just demand pressures but also the availability of gardens/residential space to redevelop, so northern core cities such as Liverpool and Manchester have proportions in the mid teens whereas suburban areas nearby are just above the national average. In contrast, the boroughs of Camden and Kingston upon Thames in central London have half of new housing provided from previously residential land. The local authority areas with the highest proportions are perhaps surprisingly the relatively wealthy low density residential areas of Chiltern (71 %), Elmbridge (69 %) and South Buckinghamshire in the South East and East Dorset (72 %) in South West. The implications of these urban sustainability building policies with their focus on high density development on brownfield sites inevitably means that construction activity is more complex than building on greenfield land. Potential remediation of the land (including any contamination/pollution) or redevelopment of buildings is generally more expensive than simply building on agricultural land. All these factors sum to a low elasticity of supply for new housing in a market upturn.
The planning policy consequences of the recession and beyond The recession and the downturn in the housing market exposed a number of issues about planning policy. First the core use of planning agreements to build affordable housing requires high land values and a housing market on an upward curve. As land values plummeted with the downturn, partly driven by the financial difficulties of house builders, the scope for the use of planning agreements to provide ‘affordable’ housing disappeared. The traditional argument that planning does not provide sufficient land was also seen to be too simple as land in the short term is no longer a constraint on development. The channelling of supply toward flats was also shown to have serious limitations in the short term with the dependence of this sector of the market on the highly geared buy-to-let market and its retreat. The impact of the financial crisis hit buy-to-let more severely than the owner occupied housing market, as shown earlier. There is also long-term legacy of planning policies that have led to a high proportion of new housing in the form of city flats over the past decade. Yet most of the long-term tenants of the noughties that have lived in these flats have not lost their aspirations for homeownership, and sooner rather than later will also be seeking non-flatted family housing. As noted above (Table 10.5), the latest statistics show some retreat from the emphasis
Jones_c10.indd 212
2/7/2012 10:23:05 AM
The UK Housing Market Cycle and the Role of Planning
213
on flats but they still constitute two fifths of total output. This adds to the uncertainties about the future. The foreseeable continuing low new supply, both in terms of market or affordable housing, will mean that the strains are most likely to be placed on younger adults. Unaffordability will frustrate their ambitions for homeownership and they will struggle to achieve the housing equivalence of their parents who will benefit in continuing use of equity release to fund retirement. There will be increasing demand for social housing. Fiscal advantages of owner occupation are set to continue but it is difficult to foresee a rise in this tenure, given the balance of unaffordability that will be generated by continuing supply constraints. Demographic trends are a fundamental long-term influence on the housing market. The balance of the population is shifting toward a more elderly structure. There are expected to be 27.8 million households by 2031, a rise of 6.3 million on 2006 levels, driven by population growth and an increase in the number of people living alone. The demand/supply imbalance that already exists is set to deteriorate without a sea change in planning policies.
New planning agenda The election of the coalition government has seen some significant changes to the planning system, notably the swift abolition of the Regional Spatial Strategies and their replacement with the ‘localism agenda’. As part of this new approach, minimum densities for housing developments have been abolished and gardens are to be reclassified and will no longer be treated as brownfield land. In July 2011, it announced a consultation on a new draft national planning policy framework. The new proposals have already generated much controversy partly because of their nebulous nature. The reforms are still a work in progress, but the top-down approach to forecasting/allocating housing demand/supply has been abolished. These forecasts/allocations will be undertaken at the individual local authority level although they are encouraged to work with neighbouring authorities. The fear is that this new localism will be a NIMBYism charter, and housebuilding will be more constrained. There is an apparent new commitment to a presumption in favour of sustainable development although the distinction with the past policy is unclear. Any change will depend on the interpretation of sustainability in this context, and some critics have argued that the associated legislation is worded as a developers’ charter (Jenkins, 2011). The reason for this is that ‘presumption’ means that the default position is that development goes ahead unless it can be shown to be contrary to local plans or the framework (e.g. protection of the green belt/areas of outstanding natural beauty). This
Jones_c10.indd 213
2/7/2012 10:23:05 AM
214
Challenges of the Housing Economy
is still a consultation document at the time of writing but the government argues that the reasons for the presumption-in-favour approach are to speed up decision making, encourage more development and encourage both councils and local communities to focus on the content of plans (CLG, 2011a). There has also been the introduction of two new initiatives the ‘Community Right to Build’ and the ‘New Homes Bonus’. The latter is designed to engender a more positive attitude toward local population growth by match funding the additional council tax (property tax) raised from new homes and properties brought back into use, with an additional amount toward the construction of local affordable housing for the following six years (CLG, 2011b). Combined with the reduced public funding for affordable housing it is difficult to see this new planning agenda making any real difference to the long-term imbalance between supply and demand. It is planning policies not the framework that has created this problem as well as the longstanding stance of the planning profession, and these seem set to broadly continue. In addition the disruption of a new system is likely to slow housebuilding in the short term.
Conclusions The extended upward trend in house prices in the UK was supported significantly by the expansion of credit and the increasingly sophisticated mortgage products available. In addition, planning policies (rather than the planning system) shaped the housing market boom by slowing the supply response. The operation of these planning policies is dependent on a buoyant housing market but they did not deliver sufficient housing supply despite the unprecedented long housing upturn. Planning policies have been more concerned with the nature of the housing built rather than the quantity. This emphasis continued even after policies were charged with becoming more market oriented over the past decade. The power of mortgage finance on the housing market was demonstrated forcibly with the international financial crisis and the immediate loss of wholesale funding. Four years on there are still severe credit constraints on the housing market, particularly in the form of the requirement of low loan-to-value ratios by mortgage lenders. This has created barriers for households, primarily young adults, wishing to buy a home. There are knock-on consequences for the demand for rented housing and increased crowding. The mortgage constraints have also contributed to record low housebuilding in the UK. However, the long-term imbalance of supply and demand has moderated the housing market downturn and there has been a relatively quick stabilisation of house prices in comparison with the last recession.
Jones_c10.indd 214
2/7/2012 10:23:06 AM
The UK Housing Market Cycle and the Role of Planning
215
To fully meet the needs of the predicted long-term demographic trends will require a step change in the supply of housing. The problem is that a substantial increase in funding for social/affordable housing will not be sufficient without a rethink of planning policies. Planning for sustainable development is ostensibly about meeting the needs of different generations (see chapter 6) not just the promotion of compact urban forms, but it is singularly failing to address these needs at present. It needs to take up the challenge in a more effective way that both addresses sustainable urban design and significantly increases the rate of new housebuilding.
Jones_c10.indd 215
2/7/2012 10:23:06 AM
11 Developments in the Role of Social Housing in Europe Christine M E Whitehead
One of the most important challenges in terms of future housing provision is the role of social housing in a more market environment. Many European countries have invested in large social sectors aimed both at accommodating lower-income employed households and ensuring adequate housing for those outside the labour market. The models by which social housing has been developed vary considerably across countries, and the policy emphasis also changes over time, particularly as the overall housing situation has improved. This chapter looks at the development of social housing, especially since the 1970s when finance and housing markets started to be deregulated. The first section provides some background to the politics and economics of social housing. The following section identifies major trends and their potential impact on social housing. The chapter then looks at some of the more important outcomes of change. On the basis of these trends and outcomes the chapter draws out some future developments and their implications, followed by some concluding comments.
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c11.indd 216
2/7/2012 10:22:48 AM
Developments in the Role of Social Housing in Europe
217
Looking back: social housing and the welfare state in Europe The early years Social housing has been provided in Europe since at least the 12th century when the churches, abbeys and monasteries built almshouses for the elderly and other vulnerable people. This type of provision formed the vast majority of social housing until at least the late 19th century. Thereafter, as regulations were introduced to ensure minimum standards, and the costs of providing adequate housing proved too high for those on low incomes, governments across northern Europe responded to varying degrees by developing policies to increase supply at lower rents. Even so, it was not until after 1945 that social and municipal housing became a major element in new housing provision and an important part of the welfare state concept that spread across much of western Europe in the wake of World War II (Esping-Andersen, 1990; Malpass, 2008a). The immediate post-war objective was to address the large-scale numerical shortage of dwellings across Europe as well as to clear damaged units and slums. The emphasis was therefore on new construction, often in the form of flats, including clearance and renewal in large urban areas. Municipalities played a significant role in ensuring development through the provision of land, and government guaranteed finance. In many countries municipalities also owned and allocated the resultant dwellings, and the provision of housing for lower-income employed households in particular was seen as an important political objective (Malpass, 2008b). More generally in the welfare states of northern Europe the stated aim was to ensure ‘a decent home for every family at a price within their means’. This involved not just building large estates on the periphery of urban areas but also clearing areas of 19th century buildings mainly in the private rented sector. These were then replaced with higher density housing, often high-rise modern apartment blocks, within urban areas. Housing was provided at below market rents and subsidised by national and often local government. In the main this public investment served to support urban labour markets rather than to assist households who were particularly vulnerable. The role of helping households with additional needs often remained with traditional charities and other non-profit providers – or it was simply left to the market to provide very poor quality and overcrowded dwellings (Priemus and Dieleman, 2002).
Attributes of the social housing sector The social sector in the welfare states of northern Europe had a number of attributes in common, although the mechanisms by which the housing was
Jones_c11.indd 217
2/7/2012 10:22:48 AM
218
Challenges of the Housing Economy
provided differed in the detail. The approach was fundamentally an alternative to market provision rather than a partnership with private providers. There was a special circuit of finance, normally involving either national government borrowing or central guarantees to municipalities. Sometimes, notably in France and the Netherlands, government guarantees were additionally made available to not-for-profit providers. This reduced the costs of borrowing and was usually further supplemented by direct subsidies to interest rates or other outgoings to enable even lower rents to be charged. The scale of subsidies was determined by government and allocated administratively to local providers on criteria that were often related as much to political pressure as to housing need (Kemeny, 1995). In some countries subsidies tended to be tenure neutral; in others they were only available for municipally owned rented accommodation (Turner and Whitehead, 2002). In addition social rented housing was often built on publicly owned land made available at low or even zero cost – further reducing cost-based rents. Because rents were held down below market levels there were queues for the available accommodation and administrative rules on allocation. The allocation rules varied between countries – and sometimes localities. At one extreme it might simply be based on waiting time and at the other on the extent of housing need. Often employed households, particularly those working for the public sector, were given priority (Priemus and Dieleman, 2002; Whitehead and Scanlon, 2007). The policy basis for the development of large social housing programmes across northern Europe was both practical and political. After 1945 most economies did not have fully operating private development industries and there were shortages of materials, labour and finance. In this context it was simply easiest to meet the overwhelming housing needs of national populations through direct intervention/government provision. At the same time, socialism dominated the politics in many northern European countries and involved implicit contracts between government and citizens to ensure investment in the welfare state – notably health and education, but also housing, as pillars of that welfare state (Kemeny, 1995). Although the objectives were very similar across Europe there were different models of provision across countries ranging from the communist states where housing was seen as part of the social wage, through countries where subsidies were concentrated directly on social provision (notably the UK, Germany and France) to countries particularly in Scandinavia and the Netherlands where there was, in principle, neutrality of government assistance between tenures (Turner and Whitehead, 1993; Priemus and Dieleman, 2002). In this new development phase there was little discussion of issues relating to efficiency of production or even to equity and redistribution. The immediate requirement was to address numerical shortages and to clear slums as quickly and cheaply as possible. Only later, as extreme shortages
Jones_c11.indd 218
2/7/2012 10:22:48 AM
Developments in the Role of Social Housing in Europe
219
were overcome, did commentators start to ask more conceptual questions about the relative efficiency of different institutional frameworks for ensuring adequate and affordable housing for all and the value of supply rather than demand subsidies (Galster, 1997; Yates and Whitehead, 1998; Stephens et al., 2005).
The changing environment of the 1970s and 1980s By the 1970s the absolute shortage of dwellings had been overcome in most northern European countries even at the regional and local level. As a result the pressure to maintain large-scale public sector building programmes lessened (Hills et al., 1990). The worst of the slums had also been removed and there was a shift in emphasis away from renewal towards renovation which was seen as having a more positive impact on communities (Paris and Blackaby, 1979). At the same time there was increasing pressure to reduce the role of the state and the scale of public sector expenditure overall. This was a particular pressure for housing where the core needs were seen to have been met. Moreover social housing provided a large unencumbered asset, the value of which had been massively increased by the rapid inflation of the early 1970s – leading governments to think of more innovative approaches to financing. The potential for developing new ways of funding social housing was further increased by the rapid deregulation of financial systems and the removal, in many countries, of special circuits of housing finance (Turner and Whitehead, 1993; Priemus and Boelhouwer, 1999). At the same time there were increasing concerns about the quality of the social housing that was being provided and the capacity to manage and maintain the stock. Most countries were therefore having to look for further sources of revenue to support investment in existing assets (Droste et al., 2008; Van Kempen et al., 2005). On the demand side increasing incomes led to growing demand for more and better quality housing. This demand was in turn supported by financial deregulation which enabled larger proportions of households to borrow to purchase their own homes. At the same time there were increasing concerns that poorer households were unable to afford adequate housing – and as better-off employed households moved out of social housing there was also the opportunity to provide better housing for more vulnerable groups. This shift in emphasis towards those less able to pay for themselves was supported by a shift towards income-related demand-side subsidies and away from the generous supply subsidies of the early post war period. At a more conceptual level this changing environment resulted in a move away from regarding housing as a social good and a core element of the welfare state towards seeing it more as a matter for private decision. The fundamental reasons for this shift in both public and individual attitudes
Jones_c11.indd 219
2/7/2012 10:22:48 AM
220
Challenges of the Housing Economy
were twofold: the vast majority of households were well housed and were able to afford their own accommodation; and households wanted greater choice over where and how they lived – because a large proportion of the benefits from housing go to the owner or occupier who consumes that housing (Whitehead, 2003a). As a result, housing above a socially determined minimum standard, which rose as incomes in general rose, came to be seen more as a private good. In some welfare states, social housing continued to be seen as an important part of mainstream housing provision (Lévy-Vroelant et al., 2008). But in others – and increasingly the majority – allocations became more concentrated on those who could not access adequate standard market housing as well as far less concentrated on employment-related housing. Whichever approach dominated, middle-income middle-aged households tended to move out of social housing towards more suburban and lowerdensity dwellings in the private sector. The extent of these pressures obviously varied between countries but were generally supported on the one hand by continued income growth and on the other by the shift in emphasis from bricks and mortar subsidies towards income-related subsidies which were available in both the social and private rented sectors (Whitehead, 2003b; Whitehead and Scanlon, 2007; Stephens and Fitzpatrick, 2008).
The major trends The chapter now examines these changes in greater detail, identifying some of the most important trends in the development of social housing across Europe. Over the past three decades six related trends can be identified which have helped change the nature and importance of social housing: ●
● ● ●
●
●
privatisation not only of the housing stock, but also of the financing of that stock which remains in the sector; increased targeting of government support to people and areas most in need; a growth in debt financing in all sectors; the pressure to use housing and land assets more effectively and for housing to become more self-sufficient; the increasing need to upgrade both the social housing stock and neighbourhoods dominated by social housing; and the continued shift towards a wider range of tenures that better meet the full and, closely linked, range of household requirements.
Privatisation The movement towards greater private involvement in housing has its roots in improved opportunities for individual households on the one hand and the need for governments to reduce public expenditure on the
Jones_c11.indd 220
2/7/2012 10:22:48 AM
Developments in the Role of Social Housing in Europe
221
other. It is reflected in both the growth in owner occupation and in greater partnership between public and private stakeholders in the provision of subsidised housing. Privatisation would not have been a viable option if it were not for both increasing affluence in general and the development of financial mechanisms that allowed housing costs to be spread more effectively in relation to lifetime incomes. On the demand side, as incomes rose, what people looked for in housing became more complex, with increasing emphasis on location, neighbourhood, facilities and the type of dwelling as compared to internal space and the number of rooms (Whitehead, 2003b). As a result, the benefits of choice also increased, making administrative allocation of housing less desirable. In this context, households tended to put a higher value on the specific attributes of single-family homes and owner occupation because they give them greater control over their home environment. Second, deregulation of the general finance market and the integration of housing finance into a global market enabled the development of more sophisticated financial instruments to support both private ownership and the transfer of ownership from social landlords to either individuals or private and non-profit landlords. Turning to housing policy development, shifts in political power toward the right across much of northern Europe led to increasing movement to restrict the role of the state through large-scale denationalisation of transport and utilities. Housing became the ‘wobbly’ pillar of the welfare state as governments cut back public expenditure in the light of European Union budget constraints. Governments saw housing both as an appropriate policy area where they could withdraw and as an asset which could be used to raise funds for other priorities (Kemeny, 1995; Malpass, 2008b). In eastern Europe privatisation, and particularly the restitution of housing assets to their previous, pre-war owners, was seen as a core element of post-Communist ideology. It was also an important means of reducing the government’s commitment to investing in what had often become a very run-down housing stock. In some cases, housing was also seen as a cash cow, enabling funds to be raised to pay off municipal debt and to provide support for broader-based local government services (Hegedus, 2007; Lux, 2003). Targeting Once absolute shortages had been overcome and the vast majority of households had indeed obtained ‘a decent home at a price within their means’, governments moved to restructure their housing subsidies in three main ways: 1. reducing supply-side subsidies overall, modifying the incentive for municipalities and non-profit organisations to develop and increasing rents;
Jones_c11.indd 221
2/7/2012 10:22:48 AM
222
Challenges of the Housing Economy
2. shifting away from open-ended supply subsidies available to all relevant providers (e.g. interest rate subsidies), towards capital grants available only to landlords and areas where particular needs could be identified; and 3. increasing the targeting of demand-side subsidies, so that those on the lowest incomes were adequately supported but those further up the income scale were expected to pay more for their own housing (Turner and Whitehead, 1993; Stephens et al., 2002). As a result, public expenditure on housing supply was reduced in most northern European countries and that which remained was targeted more on investment in areas where there were identified shortages of affordable housing. At the same time the form of subsidy was changed so that revenue subsidies were reduced or removed and replaced by upfront grants which gave greater certainty and therefore more government control over future expenditures (Whitehead, 2008). Reductions in supply subsidies led to increased rents and, in some countries, enabled rents better to reflect the value to the tenant rather than the history of financing costs. Where rents were raised significantly, this increased the capital values of tenanted property. This in turn provided a larger asset base from which additional housing investment could be funded without further direct subsidy. Most of the pressure for greater targeting came from national governments and indeed the EU. However, municipalities also faced both greater responsibilities – notably in the context of ensuring accommodation for homeless households – and enforcing often rising standards across the sector (Whitehead and Scanlon, 2007). The pressure for greater targeting was not only about cutting direct public expenditure but also complemented the move towards allocating social housing to more vulnerable households and gave greater emphasis to income-related demand-side subsidies (Stephens and Fitzpatrick, 2008). Yet while housing subsidies were becoming more targeted there was much less emphasis on reducing tax benefits. The result was that government support in countries such as the Netherlands and Sweden, with reputations for neutral tax systems between tenures, shifted towards systems that favoured owner occupation and indeed higher income households (Turner and Whitehead, 2002). Increased use of debt finance Traditional social housing was funded by public sector borrowing and subsidy particularly in the 1950s and 1960s. As the debt matured and the historic value of borrowing declined, especially in the 1970s as a result of rapid inflation, rents could be held down well below market levels. This both increased the demand for social
Jones_c11.indd 222
2/7/2012 10:22:48 AM
Developments in the Role of Social Housing in Europe
223
housing and generated problems about how to fund maintenance and improvement of the existing stock. As financial markets were deregulated and incomes rose, it became possible for many more households to buy their own homes in the private market and fund their purchase with increasingly available mortgage finance. Equally the availability of mortgage funding (sometimes with government guarantees and other support) enabled the transfer of dwellings from social housing into homeownership. This policy was most effective (in terms of the extent of transfer) in Great Britain where 2.3 million units were sold to tenants, mainly at significant discounts (Jones and Murie, 2005). But most other northern European countries also introduced some incentives to transfer and so lever in both individual equity and debt finance. These transfer policies were a significant factor in the relative decline of social housing (Gibb, 2002; Whitehead, 2007). Financial deregulation also enabled social housing sectors to be funded through private debt finance (Oxley, 2009). This in turn allowed governments to further reduce direct subsidies and public borrowing. The increasing value of the asset base, together with rising rental income, provided the security for borrowing on the global market. Independent social landlords in the Netherlands and the UK were in the forefront of this initiative. Once the market understood the nature of social housing activity, interest rates fell towards risk-free rates and social landlords were able to borrow effectively to expand supply (Priemus and Boelhouwer, 1999; Whitehead, 1999). In the UK, the market also enabled the transfer of a large proportion of the social sector stock from municipal to independent landlords (housing associations), allowing further reductions in public borrowing. Other countries, such as France, although having long traditions of social landlords borrowing from the private sector, still depended on significant government support and a special circuit of finance for their funding (OECD, 2011b). The most extreme examples of this approach have been seen in Germany where not only private debt finance but also private equity has been introduced. Dresden, Kiel and a number of other cities in eastern Germany have sold their housing stock to private equity organisations subject to licence agreements which specify allocation rules, the regulatory framework for rents and standards, and rights to demolish or sell the social housing stock. The resultant capital inflow has then been used to reduce municipal debt and to pay for local services. In western Germany, similar approaches have also been used to transfer management and ownership to the private sector either in part or fully. As a result, the ownership of the stock has in some cases moved to other countries, notably the USA and the UK, making social ownership part of a global market (Stephens et al., 2008; Droste and KnorrSiedow, 2011).
Jones_c11.indd 223
2/7/2012 10:22:48 AM
224
Challenges of the Housing Economy
Using housing and land assets more effectively As the social housing sector matured, it became increasingly possible to use available housing and land assets to increase the housing stock and to reduce rents. As prices and costs rose, rents on properties built earlier could be increased and the revenues could cross-subsidise additional dwellings and/or the renewal of existing dwellings (Kemeny, 1995). In some countries (e.g. Denmark) the potential to cross-subsidise is limited to particular estates but in countries such as the Netherlands and the UK it is possible to pool rents across the landlord’s whole stock – and in some circumstances even across the total social sector stock in the country (Hegedus, 2007; Priemus, 2010; Whitehead and Scanlon, 2007). A second approach involves modifying the social landlord’s portfolio by selling units into the market sector and buying other more appropriate units, as well as restructuring estates through demolition and renewal to increase densities and match dwelling type and size more effectively with housing needs. This can link to borrowing against increasing housing, land values and rent streams to enable additional housing to be provided or the existing stock to be improved (Knorr-Siedow, 2008). It can also help to support desegregation policies and ensure increasingly mixed communities (Droste et al., 2008). A third important approach has been the use of cross-subsidies through the transfer of land values to enable additional housing to be provided. In many northern European countries, the proportion of total costs that is made up of land costs has increased rapidly over the past three decades as economies have grown and good quality land with access to transport and services has become increasingly scarce. Land in public ownership can be provided at a subsidised, or even zero price, to support affordable housing. This has become an increasingly important issue as public accounting systems have moved to reflect current rather than historic costs and public prudential requirements have come to include selling public assets at current values (RICS, 2011; ODPM, 2006). An extension of this approach involves the ‘taxation’ of private land to provide funding for affordable housing by one means or another. In some countries, for example Spain, a specified proportion of land is allocated to affordable housing in each private development. In the UK, it is a proportion of housing as a result of negotiation between the municipality and the private developer (Crook and Whitehead, 2002). Similar approaches, sometimes nationally based, sometimes locally, can be found in many countries across the world (Scanlon, 2010). The benefits of this approach are twofold: first, the subsidy, whether in the form of land or finance, directly supports affordable housing as opposed to more general land taxation which is unhypothecated; second, when the subsidy is in the form of land within
Jones_c11.indd 224
2/7/2012 10:22:49 AM
Developments in the Role of Social Housing in Europe
225
private developments the allocation supports the mixed communities agenda – with affordable and social pepper-potted within market provision. A more general approach to using land more effectively has been the move to increase densities within mixed communities and mixed use developments (Droste et al., 2008). Much of the post-war housebuilding across Europe tended to be in the form of high-rise flats often with few local services in the immediate vicinity. Even so, the sites often included considerable amounts of underutilised, and sometimes derelict, land. Restructuring these estates to provide higher quality and mixed tenure housing together with retail- and employment-related uses and better access to local and social services has become an important Europe-wide policy approach to the effective provision of social housing in the 21st century. It is also seen as a means of addressing the emerging issues of climate change and the potential benefits of compact cities that better use transport and local services (Van Kempen et al., 2005). Rehabilitation and redevelopment Much of the social housing stock in northern Europe was built in the 1950s and 1960s to standards that were high at the period but often used what are now seen as poor quality materials and design. Much of this stock was anyway built to have only a 30–50 year lifespan. Much post-war development therefore needs upgrading, with respect to both the dwellings themselves and the surrounding neighbourhoods (Droste et al., 2008). What has become clear over the decades is that the private market is particularly ineffective in the context of renovation and renewal (Whitehead, 1998, 2003a). Most fundamental is the issue of property rights, and the range of objectives and incentives among individual owners in circumstances where many of the benefits are quite general while costs and decisions have to be made by the individual owners. There are also major issues of information, risk and contract enforcement. The result tends to be underinvestment especially in the neighbourhood as well as long drawn out inefficient processes for achieving improvement (Rothenberg, 1967). This has become particularly obvious in eastern Europe where the restitution of anyway poorly maintained properties has led to massive problems of ensuring adequate maintenance, let alone regeneration and improvement (Tosics, 2006). Policy instruments are therefore required to overcome the full range of these market failures as well as the more obvious administrative failures observed in social housing. Most northern European countries have developed public–private partnership approaches which support large-scale urban redevelopment (Droste et al., 2008). These tend to be funded by borrowing against the projected land value increments associated with well-organised, demand-oriented regeneration (CLG, 2010b). These policies also address environmental and
Jones_c11.indd 225
2/7/2012 10:22:49 AM
226
Challenges of the Housing Economy
climate change concerns mainly through regulations which apply to all investment whether in new build or rehabilitation and redevelopment. Equally they attempt to address issues of segregation and exclusion by mixed tenure and mixed use development that provide access to the full range of required services. Tenure change Initiatives linked particularly to privatisation, financial deregulation and general income growth, together with the need to reduce public spending, have all pointed to the expansion of owner occupation through both policy and choice. At the same time, there has been growing evidence of administrative failure, especially in large high-rise mono-tenure estates with inadequate service provision which pointed to a case for restructuring. The big shifts have been in the growth of owner occupation purchased with a mortgage; the transfer of municipally owned housing stock to owner occupation, independent social landlords and private finance which have introduced private incentives and constraints into the management of the social sector; and the use of public–private partnerships to support investment in both new and redeveloped housing (Whitehead and Scanlon, 2007). As these changes have worked through the system, other gaps in provision between social renting and mortgaged owner occupation have emerged. This has led to important policy developments in intermediate, shallow subsidy, housing in the form of both low-cost homeownership and rented accommodation often provided by private landlords. This has mainly been in response to increasing affordability problems among lower-income employed households who, two or three decades ago, would have been accommodated in social housing (Monk and Whitehead, 2010). It has also led to changes within the social rented sector to attempt to better meet tenant aspirations (Munro et al., 2005). The core role of social rented housing in many countries has therefore become to provide for people further down the income scale and particularly for more vulnerable households. Private housing (both owner occupation and private renting) is increasingly seen as the main means of providing for those with adequate income, while universalist systems are under increasing pressure to target subsidy more effectively.
Outcomes Although there are many differences in how social housing has developed in Europe over the past decades there are also some general patterns which arise from the six trends that have been identified above.
Jones_c11.indd 226
2/7/2012 10:22:49 AM
Developments in the Role of Social Housing in Europe
Table 11.1 and 2007.
227
Social rented housing as percentage of all housing, early 1990s, early 2000s Early 1990s
EU countries Austria Belgium Czech Republic Denmark Finland France Germany Greece Hungary Ireland Italy Luxembourg Netherlands Poland Portugal Slovenia Spain Sweden United Kingdom
Year
Early 2000s
Year
2007
21 7 15 10.7 15 15*
1992 1990 1987
13.9
1987
40.2
1993/4
17 7
2002 2003
4
2001
Non-EU countries Australia United States
6 2
1990 1991 1994/5
1994 1990
20 18 19 6 0 8.5 5 2 35
16
1.1 22 22.6
2001
1
2002
2001
21
5 3
2001 1999
3 4 1 21 18
2.3**
Note: *former West Germany only; **2008 figure. Source: All Danish figures calculated from Statbank Danmark, Table BOL 30. Other early 1990s figures from Freeman et al. (1996). Other early 2000s figures Fitzpatrick and Stephens (2007). EU figures for 2007 from Czischke and Pittini (2007), except for the UK, which is from Department of Communities and Local Government Live Table 101 (Dwelling stock by tenure). US figure calculated from American Housing Survey and A Picture of Subsidized Households, US Department of Housing and Urban Development.
Social housing supply First and perhaps most importantly, the scale of social housing has declined in proportional, and often in numbers, terms. Resultant proportions range from as high as 35 per cent in the Netherlands to single digits in a significant minority of mainly smaller EU countries. The eastern European transition economies have shown the most rapid declines as state-owned housing has been transferred either to their original owners or to current tenants. The Czech Republic has retained the largest proportion at around 20 per cent of total stock. Quite a number of traditional welfare state economies now have social rented sectors at a little below 20 per cent, including the UK, France, Austria and most Scandinavian countries (Table 11.1) (Scanlon and Whitehead, 2011). Germany is atypical in that, under a definition of social housing based on ownership, the proportion is among the lowest. Based on
Jones_c11.indd 227
2/7/2012 10:22:49 AM
228
Challenges of the Housing Economy
Table 11.2 Social housing as percentage of all housing completions in EU countries.
Austria Belgium Czech Republic Denmark Finland France Germany Hungary Ireland Luxembourg Netherlands Poland Spain Sweden United Kingdom
Per cent
Year
30 6 20 13 12 9 9 4.5 6.3 0.6 12.8 8.3 10.3 16 19
2005 2005 2005 2005 2005 2005 2005 2005 2005 2003 2005 2005 2005 2005 2008
Source: Czischke and Pittini (2007).
subsidy or allocation rules, however, the proportion is much higher, as private owners provide a large proportion of social homes under a subsidy system that requires that the accommodation is let to households in need, for specific time periods (Knorr Siedow, 2008). Within the social housing sector there remains a wide range of providers. However, the general tendency has been to move away from municipal providers towards not-for-profit and even private landlords who act as agents to meet municipal responsibilities (Whitehead and Scanlon, 2007). Social housing is, however, still a significant part of new construction (Table 11. 2) as governments continue to provide subsidy to meet specific needs. France at one extreme provides interest rate and capital subsidies to independent social landlords (HLMs) almost on demand (OECD, 2011b). At the other, in the Netherlands no direct government subsidy is now available and new investment is funded from past public support and by continued guarantees where debt finance is used (Priemus, 2010). In between, there are a range of targeted approaches usually seeking to limit the amount of subsidy required per unit built (Whitehead and Scanlon, 2007). Even though output levels in many European social sectors are significant, they rarely add much to the net total of social housing because of losses through demolition and transfers to the private sector. France is a typical in that the continued universal subsidies, the maintenance of a special circuit of finance, and limited transfer programmes have led to increases in both the number and proportion of social housing units. In some other countries, notably the UK, there have been increases since the financial
Jones_c11.indd 228
2/7/2012 10:22:49 AM
Developments in the Role of Social Housing in Europe
229
crisis, as social housing has been an important part of stimulus packages. However, this expansion is unlikely to be maintained.
Allocation of social housing The demographics of who lives in social housing have changed significantly over the past three decades. In the immediate post-war period, social housing across Europe was intended for a wide range of mainstream households, especially family households where the sole, normally male, breadwinner was in full-time employment. In many countries, local government employees were given some priority, especially if they were moving into the area to take up that employment. In some countries employers of all types contributed to municipal housing through taxation and, in return, received nomination rights. Universalist regimes like those in Scandinavia, the Netherlands and France saw social housing as, at least in principle, open to all. Other countries such as the UK prioritised those living in poor quality housing whatever their incomes. As a result, an analysis in 1970 (Holmans, 1970) showed a significant overlap in income, socio-economic group and other households between social renting and owner occupation. Over the decades, as housing shortages were overcome and opportunities in other tenures improved, there was a growing emphasis on accommodating lower-income and more vulnerable households. In some countries, notably the UK, government policy has reinforced these tendencies in order to better target assistance on those without adequate alternatives. As a result of demographic change, social housing tends to house relatively large proportions of older households (who often moved in decades ago), lone parents and increasingly single adults and migrant households as well as lowerincome couples with children (Table 11.3). In some countries there are clear distinctions between who is housed in municipal housing (more vulnerable households) and by independent social landlords who still tend to accommodate more mainstream households. This is particularly true in countries which maintain universalist objectives, notably Austria and France (OECD, 2011b). A number of countries also have policies specifically to support key workers (usually in the public sector) and sometimes employment more generally. This is one area where investment in shallow subsidy, intermediate housing have been concentrated (Scanlon, 2010). Within these general allocation rules there have also been attempts to better meet the aspirations of tenants through policies such as choice-based lettings in the UK (Munro et al., 2005). The generally observed shift towards accommodating poorer and more vulnerable people and towards more mobile and migrant households has led to increasing concerns about residualisation and segregation. This is observed in almost all countries at the estate level and, in some, far more generally across the social sector overall
Jones_c11.indd 229
2/7/2012 10:22:49 AM
230
Challenges of the Housing Economy
Table 11.3
Austria
Denmark England France Germany Hungary Ireland Netherlands Sweden
Demographics of social housing in Europe. Age/household type
Income
Young families on new estates; older people/singles on older estates Children and young people; households with one adult Young and old; single parents, single persons Single-parent families and couples with children Older people
Municipal housing: working class/ disadvantaged. Housing associations: middle class Low-income and households receiving transfers Low incomes – less than half owner occupier income Average household income 74% of national average Lower income Low income and social status 62% have incomes less than 60% of median (cf. 22% overall) Lower than average and falling Lower than average
Single-parent families and couples with children Older/smaller than average Single parents, elderly single
Source: Whitehead and Scanlon (2007).
(Whitehead and Scanlon, 2007). This has been one of the main reasons for the introduction of policies supporting mixed communities and of shallow subsidies to help employed households unable to afford the increasing costs of market housing (Monk and Whitehead, 2010). Management of social housing The heyday of social housing in northern Europe was made possible by political commitment, large-scale public funding – including the provision of land at below market values and sometimes at zero cost – and limited alternatives. Initially in most countries administrative allocation of both resources and the housing stock replaced market provision and allocation for a large proportion of the total available. In others the private sector played a role in partnership with municipalities and other social landlords but still based on administrative objectives and regulation. Over the years there was increasing evidence of administrative failure particularly with respect to the management and improvement of the stock but also in terms of the location and quality of additional housing (Whitehead, 2003b). There has been some movement towards introducing market or quasimarket incentives particularly with respect to the introduction of private finance, subsidy allocations based on cost effectiveness and value for money; and through rents related more to market values than to financial costs – in an environment where income-related subsides have become more important. However, this movement has been limited to a relatively small number of countries – those that have gone furthest down the road towards privatisation and debt financing are notably the Netherlands and the UK.
Jones_c11.indd 230
2/7/2012 10:22:49 AM
Developments in the Role of Social Housing in Europe
231
At the other extreme, in countries such as in France and Austria, more traditional administrative systems have remained in place almost unaltered (OECD, 2011b). A core shift in emphasis has been towards addressing the problems associated with a maturing – and ageing – social sector stock. This has involved different skills and increasingly complex partnership and financing arrangements to enable improvement and regeneration. Important in this context has been the increasing emphasis on sustainability and energy efficiency programmes across Europe (Cecodhas, 2011). Overall there has been relatively little attempt to improve the efficiency of management to orient provision more towards consumer demand, either in countries where demand pressures remain strong or in countries where the stock available outweighs the demand for that housing (More et al., 2005). In the first context there is little need to take account of tenants’ preferences as opposed to management objectives; in the second, there is usually inadequate funding available to bring in higher quality management and make decisions more responsive. The most important mechanism by which pressure towards greater efficiency and redistribution has been generated has been through the transfer of stock to owner occupation and, to a lesser extent, to other non-profit landlords. But, however successful these transfers have been, there always remains a significant element provided by traditional administrative means. Only in some of the transition countries has the social sector stock fallen to such low levels that it can no longer be regarded as a significant player in housing provision.
Future developments Looking to the future, there is clearly a continuing political commitment to ensuring a decent home for every household and therefore a need to provide effectively for households unable to afford market housing. There is far less clarity about whether this implies continuing direct government involvement in the ownership and provision of housing, which remains the fundamental of traditional local rented sectors. The evidence of Germany and much of southern Europe shows that, at least in principle, social objectives can be filled by partnership with the private sector. However, that evidence also suggests that pure demand-side incomerelated subsidies cannot of themselves be an effective approach to ensuring adequate housing for all. Although distributional issues are particularly important, and increasingly so as affordability measured by price and rent income ratios worsen across much of Europe, they are not the only rationale for government intervention (Whitehead, 2003b). Market failures, while not as extreme as for other welfare services, are pervasive, especially with
Jones_c11.indd 231
2/7/2012 10:22:49 AM
232
Challenges of the Housing Economy
respect to locational externalities, neighbourhood and contractual failures, and are not effectively addressed by regulation alone. Perhaps most importantly, the supply of housing appears to be becoming increasingly inelastic as economies become richer (Barker, 2003). As a result demand-side subsidies tend to be ineffective in generating supply and, to a significant extent, simply raise land and house prices (Galster, 1997, Yates and Whitehead, 1998). On the other hand there is wide-ranging evidence of administrative failure in social housing provision. Resources are used ineffectively not only with respect to new investment but in the management and allocation of the existing stock and the capacity to maintain and improve that stock. The ideal in economic terms would be to enable social landlords (whether municipal, non-profit or private) to operate in competition with other providers and indeed owner occupation on a ‘level playing field’. In these circumstances social landlords would operate only in those fields where they had comparative advantage. These would be likely to include providing for those who need additional residence-related services; where social landlords can obtain funds in a free market at lower interest rates than private landlords; and where they can address risk more effectively than the private sector. This suggests that there is a case for some social housing in most if not all localities but that to achieve scale, social landlords would operate in many different locations across the country and possibly even internationally. No European social housing system comes close to this ‘ideal’. Moreover, the model is strongly based on a hypothesis that markets are likely to work better than administrative systems in richer economies where housing now packages a wide range of benefits included in the dwelling, its facilities, the neighbourhood and access to employment and services. Choice and control over one’s own residential environment are therefore increasingly highly valued. Rather, most European systems have become relatively residualised and constrained in ways that neither achieve the fundamental objectives of social housing nor operate efficiently even within their limits. As such they in no way reflect the best that administrative systems could achieve. What is clear, however, is that there is a growing interest not in administrative systems that replace the market but rather in those that act within partnership systems. These arrangements aim to put together the relative strengths of administration and market approaches to achieve better quality housing and neighbourhoods for all. However, in most European countries this approach is still not even an aspiration, let alone a reality. The continuing financial and economic crisis in Europe since 2007 has led to shifts in emphasis in both conceptualising the social rented sector and in housing policy. Investment in social housing has been seen in some countries both as a means of meeting increasing needs but also as a useful
Jones_c11.indd 232
2/7/2012 10:22:49 AM
Developments in the Role of Social Housing in Europe
233
mechanism for stabilising the housing system (Houard, 2011). More immediately, it has been an acceptable mechanism for stimulating the macro economy. On the other hand, where policy requires private finance, the crisis has increased costs and risks, making social housing more difficult to achieve. In the medium term the need to reduce both private and public borrowing is likely to generate further pressures for increasing rents and targeting assistance more carefully. It is also generating downward pressures on housing standards exactly at the time when the sustainability agenda is looking to increase energy efficiency in particular, in both new and existing housing. The new coalition government’s housing policy in the UK and the suggested changes in the Netherlands reflect where such policies might be going toward more market-oriented economies, while France provides an example of a more positive approach to continued investment in social housing (CLG, 2011c; Priemus, 2010; OECD, 2011b).
Conclusions What we can predict with some certainty in relation to European social housing is that: ●
●
●
●
●
Jones_c11.indd 233
there will be continuing diversity in the extent, nature and role of social housing across European countries, arising from the history of how social housing has developed in each country; in those countries, mainly in eastern Europe, where governments have withdrawn rapidly from social provision, there is likely to be some attempt to develop new approaches to ensuring adequate housing for all; however, the more general trend associated with rising incomes and economic growth continues to be towards more market provision and shallower subsidies to support public–private partnerships in providing for those unable to afford market housing; at the same time, it has become increasingly obvious that market systems, together with demand-side subsidies, are not able to achieve desired levels of investment either in additional dwellings or in maintaining and improving the quality of the existing stock; practical politics therefore points to continued government involvement in supporting additional investment in housing although not necessarily in ensuring that this is undertaken in the social sector; in ‘ex’ welfare state countries, housing will undoubtedly remain an important element in the implicit contract between government and the governed; at times of crisis, as after the financial meltdown starting in 2007 this tends to be reflected in policies to maintain development (and indeed the development industry);
2/7/2012 10:22:49 AM
234
●
●
Challenges of the Housing Economy
more generally, as costs have risen relative to incomes in most countries, there is a sense of failure to provide effectively for the next generation; support for access to adequate housing among younger employed households has once again come to be seen as a policy priority; and finally, helping the vulnerable to achieve minimum housing standards and the wider benefits that flow from these standards remains a fundamental objective, although whether it should be achieved through social provision is an important part of the current debate.
Thus the policy of direct provision of social housing is on the decline in most countries across Europe – but it is by no means dead. Indeed, in a significant proportion of these countries, up to one in five households still live in the social rented sector, often a higher proportion than those in private renting. Social housing is likely to remain as a positive and proactive, if somewhat inefficient, part of most northern European housing systems.
Jones_c11.indd 234
2/7/2012 10:22:50 AM
12 Delivering Affordable Housing in the UK Kenneth Gibb
The delivery of low cost – or in some way below market priced – housing, normally dubbed affordable housing, has been a key policy challenge both internationally and in the constituent nations of the UK for several decades. At its heart it is an indicative test of the ability of the market to meet the housing requirements of households and for households’ resources to be in some way sufficient to pay for the housing costs of some threshold form of decent housing. This rather general definition is advanced in order to capture the breadth of issues underlying the affordability debate. It encompasses normative ideas about levels of housing cost, capacity to pay and household income after housing costs, as well as levels and types of subsidy that make affordability constraints less binding. It is also concerned with a quality threshold and the different manifestations of affordability, if owning or renting, and the pricing of, for instance, social or other affordable housing. As the literature indicates, there are many dimensions to housing affordability within which it is possible to come unstuck. It is important furthermore to distinguish the concept of affordability from affordable housing policies. In the UK, affordable housing policy encapsulates expanding supply of low-cost housing to meet need, intermediate housing either to part own, such as shared equity, or to rent at a level in between social and market rents, and it has also, via the Barker agenda (see chapter 10), included general supply targets aimed at reducing long-term
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c12.indd 235
2/7/2012 10:22:32 AM
236
Challenges of the Housing Economy
real house price inflation. All of these policy instruments and policy goals involve slightly different meanings and applications of affordability. This conceptual ambiguity is a feature of academic commentary on the topic and one that undoubtedly complicates debate. Alongside the varied judgemental norm-based ways of thinking about affordability are different and relatively unconnected ways of measuring or indicating affordability. Age-old debates about what constitutes an affordable rent or starter price for which subset of the population for practical purposes remain unresolved.1 In 2011, a Scottish government housing strategy paper was widely criticised over new models of affordable housing because it did not make this critical decision fully explicit. Similar disputes are echoed in England where the search is also on for a viable low (public) cost sub-market rent delivery model for new ‘affordable’ housing (in this case 80–85 per cent of market rents is being mooted). The importance of affordable housing, its meaning and the capacity to fill the gaps left by market allocation was heightened by the long boom in house prices in the UK and other countries that followed broadly the turn of the millennium. To an extent in the UK and in other deregulated mortgage markets higher prices were lessened by the willingness of lenders to offer more and more generous terms to first-time buyers. However, what might be termed an ‘affordability crisis’ has been intensified by two major developments since 2007. First, the global financial crisis has of course greatly reduced mortgage credit availability both to home buyers and to social housing providers (in the UK, social housing landlords in the form of housing associations take out long-term market loans to supplement capital grants and their internal resources when they build new social housing). Potential first-time buyers confront tougher terms and conditions for loans and in particular face much larger deposits. This access problem more than offsets the generally lower nominal interest rates and is the symptom that sparks the demand for low-cost homeownership niche products such as shared equity, which only require loans of typically 50–70 per cent of the value and nominal down payments, the remaining equity being held by the partner, normally a housing association. Equally, social landlords face a shortage of funds at a price that allows building to be viable, and a feature of the post-crisis period has been the search for innovations in private funding to fill an obvious gap (Gibb and Leishman, 2011). The second theme is the age of austerity and the fiscal crisis. English and Scottish governments have made deep cuts to capital funding of housing (a predictable target) as a contribution to managing overall cuts of more than 20 per cent in public spending required by the UK coalition government’s four year comprehensive spending review. As we shall see, this effectively ends the primacy of more than 20 years of national capital programmes for building housing association homes in Scotland. Combined with the credit market challenge, this creates the substantive question for this chapter – how
Jones_c12.indd 236
2/7/2012 10:22:32 AM
Delivering Affordable Housing in the UK
237
can policy promote new and financially sustainable models of affordable housing in the face of these challenges? We look at the evolving approach taken in Scotland since the election of the UK’s coalition government in order to answer this question. In a general sense, there are four principal ways to fund new affordable housing, each of which impacts on the distribution and management of risks associated with the development of new supply: ●
●
●
●
long-term finance – often mortgage debt repaid out of rent of social housing but could involve equity participation provider or purchaser contributions – these could be internal resources such as unencumbered reserves, individual deposits, the recycling of pooled surpluses from existing stock, capital receipts from sales of stock or planning contributions where private developers make a cash or in-kind contribution such as land for social housing in return for planning permission subsidy from the state, which can take many forms: capital grants, interest subsidies, tax concessions, direct revenue subsidies, public guarantees or subsidised loans to individuals to allow them to make a sufficient deposit to get a mortgage revenue contributions from social security (in the UK this is mainly through ‘housing benefit’ – HB) – individual personal subsidies may be used additionally to reduce the cost to the individual household and protect cash flow on debt repayments.
The focus of the chapter is how new models utilise these fundamental concepts but in a context of reduced public funding and tight credit markets. It assesses whether they can deliver affordable housing to the levels required by political imperatives: society’s responsiveness to levels of unmet housing need, as measured by government, commentators and housing lobbies. We pursue this by looking at the contemporary case of Scotland within the UK. This is a case study of a devolved country but a situation – trying to be creative and innovative in the face of financial crisis – that has relevance to the wider world, searching for new ways of building affordable housing. The chapter begins with an explanation of the historic context to social housing provision in Scotland as a prerequisite to assessing, in the next section, the challenges facing the Scottish government, both in terms of its value-for-money crisis over new affordable housing and its consequent pragmatic search for better long-term alternatives. The following section closely examines the main models emerging at the beginning of 2011 as the principal way forward. The final section concludes with a discussion about the overall effectiveness of Scottish and UK affordable housing mechanisms and a wider assessment of the adequacy of present government practice in specifying what affordability means.
Jones_c12.indd 237
2/7/2012 10:22:32 AM
238
Challenges of the Housing Economy
Historic social housing context in Scotland A brief historical account of social housing in Scotland helps to put the present debate in context. As a share of all housing, social housing accounted for 54 per cent in 1981 (all but 1.8 per cent of this was council housing) and this share had fallen to just 24 per cent by 2009 (just under half of it was housing association stock), mainly as a result of the growth in homeownership and stock transfer from the public sector to the housing association sector (Pawson and Wilcox, 2010, Table 17b). Until the mid 1970s, new social housing development was dominated by mass council housebuilding before the last major public finance crisis in the UK ended large-scale capital borrowing by councils – never to be reinstated on the same levels.2 Very few council houses were built for general needs after 1980. In parallel, the third sector housing association movement grew from small charitable beginnings into the main provider of new social housing after legislation in 1974 and in particular in 1988. The latter legislation enabled housing associations to access private finance in the form of mortgage loans which supplemented capital grants but only the grant counted as public spending. This enabled housing associations to lever-in private funding. Rents were sub-market and were deemed affordable if they were within the reach of local in-work low-income households. In practice, social housing rents have been easily affordable for the majority of low-income tenants because they also receive social security payments in the form of HB – at a level that pays all eligible housing costs for those who qualify for the broadest support (and HB also meets all rent increases for eligible applicants, also serving to massively reduce risk for lenders).3 The rent was required to repay the loan interest and meet current costs and an element for long-term repairs or depreciation. Clearly, the lower the grant, the higher the loan and rent required (see Young, 2001). In England and in Scotland housing associations have been closely regulated in order to, among other things, protect tenants but also to provide comfort to lenders. By the mid 1990s, housing associations were providing essentially all of the UK’s new social housing. In Scotland, grant rates were higher and rents lower than in England; the English system encouraged competition between associations through bidding down grant rates for new development and also greater use of crosssubsidy from profits from properties built for sale. Table 12.1 shows the long-term level of social housing starts and completions relative to all housing starts and completions in Scotland from 1996/7 to 2009/10. In the period since 2001, social housing completions have varied from 3000 to 5000 units. At the same time, official levels of unmet annual housing need varied from 7000 to 10,000 (Bramley et al., 2006) – although not all need requires to be met by new homes.
Jones_c12.indd 238
2/7/2012 10:22:32 AM
Delivering Affordable Housing in the UK
Table 12.1 Year 1996/7 1997/8 1998/9 1999/00 2000/1 2001/2 2002/3 2003/4 2004/5 2005/6 2006/7 2007/8 2008/9 2009/10
239
Scottish house building 1996–97 to 2009–2010. Social starts
Social completions
Total starts
Total completions
4101 3142 3515 4282 4699 4658 3707 4621 4406 5127 5584 6214 5765 5580
3204 4603 1873 4033 3916 4262 3809 3368 4024 4698 3237 4125 4913 5919
22,014 21,677 20,510 22,646 22,315 23,178 22,274 27,049 27,003 26,367 28,419 26,592 19,593 15,372
20,696 22,587 20,657 23,107 22,111 22,571 22,747 23,822 26,468 24,947 24,247 25,781 21,019 17,474
Source: Scottish Government Scottish housing statistics webpage. Accessed September 2010.
Background to the current Scottish affordability challenge Devolved affordability Scotland’s evolving policy framework for social and affordable housing is similar to, but distinct from, that of England. This divergence began before devolution in 1999 but has accelerated thereafter. Many of the basic elements are the same: mixed funding for housing associations, aiming to develop affordable contributions through planning agreements with private developers and encouraging shared equity models. However, each mechanism is in practice different in important respects north of the border. More generally, there are several important ways in which Scotland’s housing system has evolved differently from England (Wilcox, et al., 2010; Gibb and Leishman, 2011). First, for council housing, Scotland has historically enjoyed greater financial discretion than in England, and this translates into greater investment capacity. Scotland did not have to experience the centralised finance regime controlling English housing local authorities (particularly after 1988), and also enjoyed (for most of the period) greater financial freedoms over the use of rental surpluses, rent-setting and capital receipts. There have been no national rent structures or rent ceilings as has been the case in England. Consequently, there was less demand for stock transfer than in England, and higher levels of capital spend from council’s own resources in Scotland. By the turn of the century rents had risen sufficiently that almost all councils received no revenue deficit subsidy.4 Second, the mix of providers in Scotland has remained essentially a combination of councils and housing associations. Only a handful of
Jones_c12.indd 239
2/7/2012 10:22:33 AM
240
Challenges of the Housing Economy
councils opted for stock transfer (although this does include the largest council, Glasgow) and there are more than 200 housing associations ranging from the neighbourhood-level or community-based to regional providers and subsidiaries of UK social landlords. Unlike England, there has never been a funding route for arms-length housing organisations running council housing with additional funding or the political will to support public– private models like PFI adopted south of the border. Third, the English have now more than 20 years’ experience of planning agreements generating affordable housing (reviewed in Crook, et al., 2006; Crook and Whitehead, 2002). The last two Scottish governments, having come to this more recently, have expressed considerable hope that this route will generate a lot of the numbers they wish to see developed – but as a strongly pro-cyclical policy, it has as yet neither had the impact nor generated a critical mass of practical experience of making it work across much of Scotland. That is not to say that it will not become an important element of policy in the future, but it remains relatively untried as yet (see Newhaven, 2008, for a critical analysis of this policy framework). Finally, compared with England, there has been much less use of crosssubsidy to fund housing association development, and in part this is because of the tradition in Scotland of a much less competitive grant system, higher grant per unit costs and less emphasis on driving down the cost per unit (at least until 2007). Grant rates are high in Scotland (around £70,000 on average over the last few years until 2011/12’s revolution in grant rates – see below). This long period of arguably inefficient ‘plenty’ makes the recent break with the past in terms of grant support particularly stark. A further point to stress in this introduction to social housing in Scotland is the wider context of devolution. The housing budget is a component of the wider ‘Scottish block’ funding that is approved by the UK Government each year. The block rises in relation to wider UK spending, and there is a formal relationship – defined by the Barnett formula – that provides additional funding when English only spending increases, though it also reduces the Scottish block when spending is cut in England/UK. There are also assumptions built in about benefit spending including assistance through council rent rebates. An important feature of this block grant is that the opportunity cost of programme spending decisions should be relatively clear because of the fixed nature of the overall spend (i.e. they cannot run a deficit). A feature of the current political settlement is that current and recent governments have committed Scotland to a range of universalist social policy programmes: free medical prescriptions, an element of free social care for the elderly; free transport for the elderly; a commitment to free higher education, etc. These commitments put further strain on remaining budgets such as housing and this is worsened by political pledges to protect voter-friendly areas such as
Jones_c12.indd 240
2/7/2012 10:22:33 AM
Delivering Affordable Housing in the UK
241
the National Health Service. Something has to give – and so far one of the key losses has been affordable housing. Although local authorities can borrow for capital spending, the Scottish parliament does not currently have the powers to borrow for its capital spending programmes – capital funding comes directly out of the block grant. This is why the Scottish parliament is seeking these borrowing powers in the medium run – and will get them through UK legislation currently under discussion – but is also incentivising councils to borrow for housing purposes, as explained below. Scotland also has money spent on it by the UK government – reserved spending and policymaking, some of which directly impinges on housing. Most of the taxation system remains determined and controlled by London. The same is true of the benefits systems. A third key area concerns the public spending rules set by Her Majesty’s Treasury – what is public spending and to what extent can a devolved government stretch the definition of what it is allowed to do? In a period of severe public spending constraint, these issues grow in importance. The Scottish government has used this discretion on several occasions: council housebuilding (examined below) relies on council borrowing at public sector rates and also receives government grants of £25–30,000 per unit. Normally, this would be classified as supported borrowing and the entire borrowing would be part of the Scottish block control total. However, a devolution concordat allows such council borrowing to be scored as annual managed expenditure and hence to be additional and non-controlled. An English council could not do the same thing. However, the broader picture is one of constraint and limited room for manoeuvre. As we discuss below, it is this lack of freedom that has led government to a wide-ranging pragmatic search for new models of affordable housing.
Affordable housing in crisis: 2007–2011 The Scottish National Party (SNP) minority government launched a housing policy consultation (Firm Foundations) shortly after coming to power in 2007 (Scottish Government, 2007). The document centred on an aspiration to increase new housing supply from typically 20–25,000 annual completions to around 35,000 by 2015 (a policy broadly in line with the Barker supply agenda). Table 12.1 above shows the pattern of new build since 1996–97. The government wished to reduce the perceived high grant per unit received by housing associations; one way to achieve this was through regional consortia who would bid for new affordable supply (Gibb and O’Sullivan, 2008).5 At the same time, the right to buy of tenants (see Jones and Murie, 2006) future new council properties would be outlawed in order to clear the way for a new council housebuilding programme supported by a grant of £25–30,000 per unit.
Jones_c12.indd 241
2/7/2012 10:22:33 AM
242
Challenges of the Housing Economy
There was also a desire to expand the low-cost homeownership package, specifically through shared equity opportunities. Furthermore, the government assumed that planning agreements for developer contributions to affordable housing (known as Section 75, after the relevant part of the planning legislation) would grow quickly and take on the share of social housing found in England. These were modest but significant changes to the affordable supply firmament. However, they were blown aside by the global financial hurricane and resulting recession which was keenly felt in Scotland, as elsewhere, within the housing and construction sectors. House building, house and land prices, transactions and mortgage lending all fell (O’Sullivan and Gibb, 2008). In the light of the crisis and subsequent recession, the Scottish and UK governments responded with specific expansionary housing sector policies to stabilise and act as counter-cyclical measures (at least temporarily). The government reintroduced a generous income support measure to help homeowners in repayment difficulties and an insurance guarantee to incentivise lenders not to repossess households in arrears. They supported the mortgage lenders more generally and the bank lending rate was brought down to an unprecedented low rate. The UK government also temporarily loosened stamp duty to encourage transactions (though arguably all this served to do was to alter purchase timings). They also reduced value added tax for a fixed period which has some relevance to those parts of the housing sector not exempt for such taxation (O’Sullivan and Gibb, 2008). In Scotland, the government pursued a number of measures to sustain affordable supply and the construction sector. These included temporarily increasing the grant rate to housing associations, running additional waves of subsidised council housebuilding, expanding new supply shared equity (later partly redirected onto social housing investment at the margin) and supporting specialist housing advice to keep marginal owners in their homes. The Scottish government also allowed housing associations to make land purchases and bank land with a promise of grant in future years to allow building (known as forward funding). The most radical measure was the decision to accelerate affordable housing funding by bringing forward funds allocated to the final year of the three-year spending review period – £120 million – though this left a significant shortfall for the final year and hastened the search for radical alternative models of affordable supply. The Scottish government embarked on a further round of consultation in the summer of 2010 leading up to a strategy paper in February 2011 (Scottish Government, 2010; 2011). The 2010 document at its heart sought to consult on innovative or creative ways to deliver more affordable supply given the reduced public finance envelope (hastened by the election and then the deficit reduction strategy of the UK coalition government). The starting point (see Table 12.2) was a stark assessment of the value for money of existing
Jones_c12.indd 242
2/7/2012 10:22:33 AM
Jones_c12.indd 243
2/7/2012 10:22:33 AM
£2–4k
£23k
£55k
£45k
30
Can be sold on capital receipts recycled to Scottish government Can be sold on capital receipts recycled to Scottish government 5–10
£30k
60+
Yes
Yes
Yes
Yes
Yes
Yes
New Supply?
Short assured tenancies. Rents and targeting largely as for MMR. 1000 units are approved for 2010/11.
Currently being piloted. Target customers as above. 108 units are planned for 2010/11.
New lets almost wholly to lowest income deciles. Planned approvals fell from 3421 units in 2009/10 to 2117 in 2010/11. Severely constrained by LA borrowing limits, new lets as above. Planned approvals grew from 1276 in 2009/10 to 1784 in 2010/11. Currently not on any significant scale. Uses short assured tenancies. Rents at 80 per cent of market rent and targeted at economically active but low to moderate incomes. Targets customers similar to MMR. Planned approvals are of the order of 260 units in 2010/11.
Comment
Source: Scottish Government (2010) – note the comments column is reprinted from the original with the author’s additional comments in italics. Short assured tenancies are tenancies that have a legal minimum of six months’ lease and rents set by the market.
National Housing Trust (guarantee)
New supply shared equity (developer)
New supply shared equity (RSL)
£70k
60+
Duration (years)
Average Scottish government component per unit
Current funding/subsidy options.
Housing association grant Council housebuilding incentive Intermediate rent (MMR)
Grant type
Table 12.2
244
Challenges of the Housing Economy
models of affordable housing. Although contentious, it made a powerful point about the high relative cost of the mainstream housing association development programme. Interestingly, this approach was avowedly nonideological. It was as much about finding creative new ways to lever in private funding, including from the financial institutions, as it was about successful pilots by various housing bodies and making innovative use of scarce public resources. In all, the government suggested 17 small and large proposals, including a number of policies already in place. Table 12.2 indicates that the top-line grant per unit for housing associations (£70,000) compares very unfavourably with the council option (£25–30,000), even if a detailed examination might suggest that the opportunity costs to the public purse were a little higher, e.g. the resulting council rents are often relatively high and this has HB public expenditure implications; the use of prudential borrowing capacity for new housing also reduces capacity for council to invest in other activities. It is also noticeable that the housing association shared equity and housing association intermediate rent products (the latter known as mid-market rents) require higher grant for an expected short life span than do the council build programmes. We discuss the different models in more detail below but the final model based on public loan guarantees – the National Housing Trust – is cheaper still, even after one accounts for its short expected life (although it also makes use of council borrowing capacity). There then followed a consultation process through the autumn that led to the 2011 document. The strategy document in fact combines a small number of concrete proposals (both new and the retention of other existing schemes), alongside support for a number of pilot experiments and the aspiration to support a number of longer-term ideas such as the intention to involve pension funds in social housing. Our focus in this chapter is on the concrete models that are scheduled to be taken forward in the near term. These are developed fully below but consist of: ● ● ●
●
the retention of council housebuilding through further programmes; new supply and other variants of shared equity; housing association development premised on a small overall programme and only £40,000 grant per unit;6 and the National Housing Trust intermediate rent model based on a private– public partnership between councils and the private sector underwritten by a loan guarantee.
The Scottish government appears to have decided to end the existing affordable housing programme for new association development. This was the grant funding component that had operated since 1988. The existing pre-committed spend will be disbursed (e.g. for earlier forward funding
Jones_c12.indd 244
2/7/2012 10:22:33 AM
Delivering Affordable Housing in the UK
245
initiatives where associations purchased land in lieu of the promise of grant) but thereafter it looks as if the support model for councils and associations will be the much smaller Innovation and Investment Fund, which assumes a target of £40,000 grant per unit for associations and £30,000 per unit for councils. The new Innovation and Investment Fund will only have £50 million in 2011/12: £20 m to councils, £20 m to housing associations and the remaining £10 for competitive innovative bids. It may well be a larger fund in future years (once the older programme ceases) but it will still be very challenging in terms of grant rates. The Scottish government argues that its programmes combining new affordable social supply, intermediate rent and shared equity will generate 6000 affordable new homes each year for the next three years. This would be close to the generally accepted levels of unmet need in Scotland (Bramley et al., 2006). Before we evaluate this proposition, we need to develop criteria with which to assess the main models.
Assessing the emerging models Assessment criteria The process of the narrowing down a large range of ideas into concrete proposals involved the more or less explicit adoption of a range of criteria by the Scottish government. The 2010 consultative document Fresh Thinking, New Ideas, laid out specific criteria with which to assess the suggested proposals: ● ●
● ●
●
value for money i.e. budgetary cost and the level of risk to the public sector; the potential scale and deliverability of each initiative and the extent to which it would help meet housing need; the likely length of time required to deliver new or refurbished homes; the initiative’s contribution to the Scottish government’s overall purpose of securing sustainable economic growth; and the scale and nature of benefits associated with other government priorities such as tackling climate change (a major continuing commitment for the Scottish government).
These dimensions together embody arguably a rather narrow perspective (Gibb and Leishman, 2011). The government’s criteria do not consider impacts on the wider housing system or unintended consequences, such as displacement of activity that would otherwise have taken place. This is important because of the government’s underlying pragmatism – there is less concern with the system as a whole and with longer-term second round effects and more with low-cost ways of building and sub-market pricing now – i.e. the future is heavily discounted and the perspective is strongly
Jones_c12.indd 245
2/7/2012 10:22:33 AM
246
Challenges of the Housing Economy
bottom-up. As is developed below, this rather non-strategic approach may be successful and foster local innovation and creativity but the policies as a whole will be more difficult to manage from the centre, and there is a lack of a wider coherent vision to such an approach. Hall and Gibb (2010), as part of the wider Joseph Rowntree Foundation Housing Market Task Force, propose an alternative set of criteria relating to the reform of housing subsidy systems for social housing, arguing that any housing subsidy system (or indeed any set of affordable supply funding and delivery vehicles) should possess the following characteristics: ●
●
● ● ●
● ●
be able to withstand shocks over the often quite long periods of repayment of the initial investment; be able to work efficiently (e.g. are well targeted) and without creating distortions or other unintended consequences; be consistent with society’s norms about affordable rents and prices; be transparent and simple to understand to the relevant parties; work with the grain of other related social policies relating to, for example, labour incentives and building mixed communities; enable new supply to be quickly forthcoming; and offer good value for money for the public purse, keep costs under control and allow for an appropriate financial and regulatory relationship.
This set of criteria, compared with the government’s choice, consider additional points about transparency, regulatory relationships, efficient design, unintended consequences, working with the grain of other policy spheres and norms relating to what we mean by sub-market rents of housing costs. In the analysis below it is this broader perspective that is adopted, but in practice the emphasis is on robustness of the system in the longer term, the efficiency of the delivery model and whether the rents produced are acceptably affordable.
Assessment Council housebuilding A first key model is one of the success stories of the SNP government’s period of office – the return of council housebuilding on a relatively large scale since its demise in 1980. Table 12.3 shows the spatial allocation of the first three waves of new council building. Building has been significant in the Lothians, Grampian, Lanarkshire and other parts of the central belt. More than 3000 units have been built. We saw in Table 12.2 that the Scottish government public cost was established at £25–30,000 per unit. However, one should be clear where the wider funding for the new programme comes from in totality. In addition to the grant from the Scottish government, councils used capital receipts and
Jones_c12.indd 246
2/7/2012 10:22:33 AM
Delivering Affordable Housing in the UK
Table 12.3
247
Scottish council building programme, 2008–10.
Local authority Aberdeen City Aberdeenshire Angus Clackmannanshire Dundee City East Ayrshire East Dunbartonshire East Lothian City of Edinburgh Falkirk Fife Highland Midlothian Moray North Ayrshire North Lanarkshire Orkney Islands Perth & Kinross Shetland Islands South Ayrshire South Lanarkshire Stirling West Lothian Scotland
Population (2008/9) 210,400 241,460 110,310 50,480 142,470 119,920 104,720 96,100 471,650 151,570 361,890 219,400 80,560 87,770 135,920 325,520 19,890 144,180 21,980 111,670 310,090 88,350 169,510 5,168,500
Rounds 1–3 total allocation (units) 151 155 81 24 40 50 52 267 280 132 160 106 352 172 63 218 78 81 20 68 106 127 515 3,298
Rounds 1–3 grant (£m) 3.93 4.1 2.24 0.6 1.04 1.3 1.43 7.155 7.48 3.48 4.31 2.925 5.255 4.92 1.645 5.625 2.0 2.25 0.6 1.83 2.775 3.45 9.615 79.955
Source: Scottish government housing website (www.scotland.gov.uk/Topics/Built-Environment/Housing) accessed September 2010. Note: Local authorities who had undertaken whole stock transfer are excluded (Western Isles, Argyll & Bute, Glasgow, Inverclyde, Dumfries & Galloway, Scottish Borders), plus others that chose not to participate: East Renfrew, Renfrewshire and West Dunbartonshire. There are a total of 32 localities in Scotland.
surpluses from the ring-fenced housing revenue accounts. Councils also tend to own the land they build on. The residual figure would be part of normal council borrowing – so-called prudential borrowing, which allows public sector borrowing and debt management of borrowing where there is an identifiable new stream of secure income to repay the loan up to the council’s overall borrowing capacity. Clearly, this financial capacity is strongest where need or demand is itself operating at a high level (generally this is particularly true in the east of Scotland). Combining the different elements of public land and prudential borrowing capacity, it is clear that the opportunity cost is far higher than £25–30,000 per unit. However, this is the headline figure that the contrast is made politically with housing associations. The abolition of the tenant’s right to buy a new council property further removed an important disincentive for councils to build. While there was scepticism about the ability of councils to return efficiently to housebuilding, the experience on the ground has not been a
Jones_c12.indd 247
2/7/2012 10:22:33 AM
248
Challenges of the Housing Economy
problematic one. In fact many councils have worked in partnership with associations, drawing on the latter’s development expertise. They have also been willing to innovate, for instance in one case using their borrowing powers to act as a source of cheap private finance for a housing association who was a development partner in eastern Scotland. This model should be able to withstand financial shocks, given its prudential borrowing basis, the discretion over future rent levels and the requirement for such borrowing to have a clear new additional income stream, i.e. areas of demand pressure. Of course, the financial models over a long period of time are susceptible to the quality of the assumptions adopted about the future; in the same way, stock transfer business models have to predict future levels of voids, arrears, etc. There are risks associated with HB change, as there are across the social rented sector, but insulation from private lending clearly helps. It is too early to say whether housing management lessons have been learnt from the past and in particular how effective asset management of the new council stock will be, but initial programmes look encouraging. Bramley et al. (2010) conducted an analysis of the financial capacity of councils and associations to assist with future housing development. This detailed study was controversial, and has remained so, in that the government has enthusiastically looked to its findings as evidence that public resources can be stretched further. The authors’ estimates are based on explicit forward assumptions that the council sector has the capacity to build nearly 2500 per annum with a grant per unit of £32,000 (£78 million a year in current prices) and the remainder coming from Housing Revenue Account surpluses and council borrowing. The critical issue, aside from the accuracy of the assumptions, is how much potential there is for building once the estimated capacity is qualified by a council’s political aspirations, its prioritisation of borrowing capacity and the evolving local sense of the best ways to solve housing problems. The capacity study does not measure the effective but rather the potential supply, and there is no reason to think that take-up will match this potential figure. As was indicated above, the council programme offers relatively good value for money – at least if it is based on Table 12.2. However, it does reduce borrowing capacity for non-housing activities and there remains the accounting rules question of whether it is right for what is clearly supported borrowing to count outside of the block department expenditure limit? The model is simple and transparent and familiar – issues that cannot be so easily said of shared equity or the National Housing Trust. It should also be able to continue to be readily forthcoming and to augment supply where it is needed. While we might expect new build housing to have a rent premium, there is anecdotal evidence of some new build rents significantly higher than comparable local social rents. Bramley et al. (2010) also note that if
Jones_c12.indd 248
2/7/2012 10:22:33 AM
Delivering Affordable Housing in the UK
249
low-rent councils converged their rents on the average Scottish council rent, this would also expand the capacity to borrow and build. There is no clear evidence base, but there may be concerns about rental affordability in the future, particularly given planned reductions in the generosity of HB. Housing association general needs A second broad area or model is the rather uncertain future for housing associations as developers, both for social and intermediate rented housing. Housing associations have an overhang of development caused by forward funding and other specific commitments. Perhaps the most important example is the investment and new build programme underway in Glasgow, agreed as part of the massive stock transfer of council housing to a housing association in the city. But after these prior commitments are exhausted, the sector will be expected to rely on a benchmark £40,000 per unit grants for general needs social housing and a very small programme of £20 million in 2010/11. As indicated, this may be a larger overall programme in future years, e.g. in excess of a £100 million, but with this binding grant per unit constraint there could be additional requirements to meet the government’s innovation fund’s expectations about creativity and partnership etc. The key question is can associations build social housing for a much lower grant and keep rents at social rent levels? The government assumes that the financial capacity study is accurate and there are spare funds and unused equity etc. that can contribute to development costs; and that crosssubsidy from properties for sale can also bring down the private borrowing required to equate with acceptable social rents. There is much scepticism about either the willingness to use internal resources or the implicit financial capacity. Development officers in the sector also argue that the depth of cross-subsidy required would mean that new estates would have to be largely private for them to work. At the same time, the earlier point questioning the extent to which housing associations will make use of the financial capacity identified by Bramley et al. (2010) remains to be tested. Instead we may expect to see a greater penetration of intermediate or mid-market rent housing with this new supply filling the niche between social and market rents, normally thought to be around 85 per cent of the prevailing local housing allowance, the median private rent for benefit purposes (equivalent of HB in private sector). Lower grant rates can be traded off against higher rents which should still attract the full level of HB should the tenant need to claim. The innovation and investment fund expects housing associations to work with a higher grant per unit than councils but it is still a significant reduction compared to previous years – so there is a relatively large value for money gain – if associations build. It remains to be seen whether they will take up the challenge. If the sector is obliged to move to 80 per cent of local
Jones_c12.indd 249
2/7/2012 10:22:33 AM
250
Challenges of the Housing Economy
housing allowance rents in order to be able to build, then this will call into question what the sector means by a social and an affordable rent. The expected expansion into intermediate rent will also create a degree of interaction with other providers notably in the private rented sector and potential first-time buyers/shared equity purchasers. As has been pointed out both in England and in Scotland, this policy assumes that there is a big difference between market and social rents but where private demand is weak, private and social rents may be broadly the same. In such a context there is little scope to generate extra finance by ‘raising’ social rents towards market levels. Such policies will only work in high demand areas and therefore raise questions about the kind of need that will be met in future. Other cross-tenure impacts may result in less need being met overall because of displacement and substitution effects across tenures. Intermediate rent may also be a counter-cyclical policy that may not sustain demand in a recovery when mortgage lending loosens again. Like the other providers, housing associations are also exposed to the risk of HB reform (i.e. generosity cuts), particularly so for the intermediate rent sector. Shared equity Further along the intermediate housing spectrum, shared equity is a third model – the only subsidised model for low-cost homeownership in Scotland now that the government has decided to abandon other such initiatives (e.g. shared ownership which involved a rental component). Shared equity comes in several variants but can essentially be distinguished between new supply and off-the-shelf versions and also between those with housing associations as the partner and those versions, more speculatively, where the private sector might be the partner. Shared equity is a policy that attempts to overcome specific symptomatic failings in the housing market – current tight lending rules mean that firsttime buyers need more assistance to purchase – in this case to part-purchase with effectively a free equity loan. Government is now also developing policies in Scotland and the UK to help first-time buyers secure a deposit to allow them to take out loans of around 70–80 per cent of valuations. The point is that as the housing market recovers and lending restrictions ease, demand will naturally fall, for what is a niche product. If shared equity demand is counter-cyclical then this may call into question the government’s assumptions about its uptake over a period of years. From a demand-side point of view shared equity is a counter-cyclical policy offering alternatives when the owner-occupied starter home firsttime buyer sector is not functioning. On the supply side it does something similar through helping developers to sell properties and providing development activity for social housing providers. It is not at all clear that the policy is as required across the full market cycle. Although government will get its grant back on resale, the association new supply variant of shared equity is
Jones_c12.indd 250
2/7/2012 10:22:34 AM
Delivering Affordable Housing in the UK
251
quite an expensive subsidy in terms of upfront capital grants which cannot be used for anything else. Apart from the holding period and opportunity cost dimensions there is also an issue of prioritisation of scarce funds between marginal homeowners and low-income renting households. It is a normative question – but is the balance right? National Housing Trust The fourth model is undoubtedly the most radical – the National Housing Trust (NHT). Part of the origin of the NHT comes from the Scottish government’s desire to find a not-for-profit alternative to public–private partnership infrastructure projects. The Scottish Futures Trust developed a housing model in partnership with local authorities that will create joint ventures locally between a council and (normally) a private developer. The Trust provides intermediate rented housing, with rents of 85 per cent of the local housing allowance, for a period of 5–10 years when the properties are then sold off. The council borrows from the Public Works Loan Board and faces two primary risks – that the properties are not let or there are arrears and, second, that there is a capital loss on resale. These losses are insured by the government through a loan guarantee of £2500–4000 per unit. As with the council building incentive programme, the subsidy is not the full opportunity cost but it looks like a good value for money deal for the public purse even with the shorter timescale – provided suppliers actually deliver. The NHT model was initially funded for 1125 units across Scotland and these have been fully subscribed by the eight participating councils (see Table 12.4). However, they are not all in the ground yet and one must indicate a degree of caution about whether the Scottish government’s ambitious plans for further expansion of this segment of the intermediate housing
Table 12.4
National Housing Trust round 1.
Local authority Aberdeen Edinburgh Dumfries and Galloway Dundee Falkirk Highland Scottish Borders Stirling
Maximum number of homes sought 50 600 100 75 100 50 100 50
Source: Scottish Government website, accessed March 2011 http://www. scotland.gov.uk/Topics/Built-Environment/Housing/supply-demand/nht Note: Glasgow, East Renfrewshire and East Lothian dropped out after initially expressing interest. However, at least one of the three, Glasgow, has definite plans for future NHT activity.
Jones_c12.indd 251
2/7/2012 10:22:34 AM
252
Challenges of the Housing Economy
market will be achieved. The Scottish government is also working up a housing association version of the NHT but no details are yet forthcoming of how exactly it would operate (other than that the association would replace the private developer) and interest has thus far been lukewarm. NHT rent-setting is a potential problem. The model expects rents to be around 85 per cent of the median local market rent (i.e. the local housing allowance used for HB purposes). This is above social rents but within benefit ceilings. Or at least it was until benefit cuts in the summer of 2010 which redefined the local housing allowance calculation, reducing it from the median to the 30th percentile. This means that 85 per cent of the local housing allowance is now 85 per cent of a smaller denominator (depending on the local distribution of rents). This change may yet call into question the viability of the NHT if rents are still to be safely within benefit ceilings yet remain financially feasible. The NHT model will stand or fall on consumer demand for this segment of rental housing. This again raises questions about how local rental markets will interact with intermediate products, about the level of private demand and rents relative to the social sector, and again raises possible questions of crowding out and displacement. To the extent that intermediate rent demand is a symptom of failings on the first-time buyer market, it may not have the scale of demand assumed by government (though this may not be viewed as a significant problem if the market does recover). In the long run, the NHT may become a low-cost intermediate rent option for housing associations – depending on the details of how it is to work. The government thinks the NHT will generate thousands of new intermediate rent units over the next three years – but it will depend on demand in a presumably recovering market and on councils’ continuing willingness to use up borrowing capacity in this way.
Concluding discussion This chapter has presented an overview of the fairly typical and representative struggle faced by the Scottish government to find new value-for-money ways to fund affordable housing supply while confronting private credit shortages and a dwindling future source of public finance. There are several integrative conclusions one can reasonably reach about the government’s strategy for affordable housing in Scotland. ●
Jones_c12.indd 252
While there is much to be said for local bottom-up solutions (innovation, sensitivity to context, etc.), the decision to move away from centrally controlled resource allocation towards a series of local models implies the loss of strategic central government control over the delivery of affordable housing policy and the shape of the spatial allocation of spending.
2/7/2012 10:22:34 AM
Delivering Affordable Housing in the UK
●
●
●
●
●
Jones_c12.indd 253
253
An area where the housing system has massively changed concerns the position of housing associations. The new funding models and the shift to other providers and market segments threaten to fragment the sector further. It also calls into question what vision or purpose government has for the housing association community. These rapid changes after a long period of stability herald an uncertain future for the housing association sector. A genuine unknown is whether the new models will generate the level of supply (18,000 units in three years) anticipated by government. There are many reasons to be sceptical but it is not possible to be precise, such are the many contingencies. Will there be the borrowing, development and management capacity for new council build? Is there the appetite and demand for the increasingly crowded mid market involving the NHT model, shared equity and intermediate rent, let alone other parts of the private rented sector? Allied to these increasingly important local interdependencies in terms of the demand for the offering from different types of providers, just what the government considers to be a social rent, an affordable rent and an intermediate rent needs much more clarity than has yet been provided. The financial capacity study is a major step forward in thinking about what can be achieved in a mixed finance setting – however, and quite reasonably, it is open to wide interpretation because of the assumptions it uses and the types of sensitivity analysis presented. Moreover, for both council and housing associations, there are wider political and professional choices to make about whether or not to exercise the financial capacity that they possess and to develop homes – there are many reasons not to convert potential into effective capacity. It is probably the least developed end of the policy framework: long-term partnership models, bond finance, pension funds and other players as equity investors – that is the most important long-term route from government’s perspective to develop new models. Bond finance is already well established among large UK housing associations (also acting in Scotland through subsidiaries), the government is piloting pension fund equity involvement and there remains considerable interest in private sector participation in the grant-funding of affordable housing more generally – but that does not mean it can be made to work in more than nominal numbers. However, these are not easy and require careful thought about governance, regulation, accountability, the position of tenants, rent levels etc. It is also still likely that the main players will be larger developing associations and this could further fracture the housing association sector in Scotland. Many associations will not be able to participate with lower grant rates or lack of access to cheaper long-tem funding. This is not a static situation. There have been cuts but currently looking forward, government is planning on spending levels fixed in ‘flat cash’
2/7/2012 10:22:34 AM
254
Challenges of the Housing Economy
terms, i.e. real cuts year on year for four years to come. The Scottish government will in time receive additional spending, taxing and borrowing powers, though these will only come into play later in the next Scottish parliament (Calman, 2009). This fundamental lack of powers and the public accounting rules are a key (pragmatic) reason for the opportunities afforded to councils (who do have the relevant borrowing powers). The impact of HB reform may yet undo several of the plans developed by the Scottish government, such is the intertwined relationship between funding affordable housing and the benefits system. Two final points should be noted. First, the contemporary discussion and debate around affordable models in a context of fiscal shortage and unmet need illustrates clearly that, in Scotland at least, there is no agreed practical application from government regarding what social and affordable means in terms of the level of cost faced from housing. After the decades of analysis and debate it is a remarkable and salutary finding. Second, although the focus has here been on Scottish attempts to wrestle with these wicked problems – the principles apply universally and will have resonance across the world and not just in Scotland and in England.
Notes 1. There is an extensive academic debate. See, for example, Bogdon and Can (1997), Fingleton (2008), Leishman and Rowley (2012), Maclennan and Williams (1990), Newhaven Research (2008), Stone (1993, 2006a, 2006b) and Whitehead (1991, 2007). 2. Councils borrow 60-year loans from the public works loan board at below market interest rates. Rents running costs and debt are pooled across the whole council stock. 3. The UK coalition government is making major cuts to housing benefit, initially mainly in the private rented sector but it does spill over into social renting too, e.g. the greater use of lower size-related rent ceilings for benefits, penalties for under-occupation and greater reductions in benefits if non-dependents are present. 4. The English councils on the other hand had a succession of policy constraints that governed their use of rental surpluses, controlled capital spend and debt repayment in generally much more severe and restrictive ways, culminating in the present proposals to radically dismantle the fiendishly complex and over-centralised system and replace it with something that may end up looking a bit like the Scottish system. 5. This did not happen in the wake of sector-wide resistance. 6. Council new build and the housing association development programme (i.e. new money) have been combined into an Innovation and Investment Fund – worth £50m in 2011–12. Alongside spending previously committed from earlier years (e.g. forward funding), this compares with a 2010–11 budget for housing association affordable general needs development of £221m.
Jones_c12.indd 254
2/7/2012 10:22:34 AM
13 The Private Rented Sector As a Source of Affordable Housing Michael Ball
This chapter aims to provide an assessment of the potential for the private rented sector (PRS) to provide an alternative to homeownership for households either priced out of owner occupation or for whom the risks of home purchase are too high. This statement takes private rented housing to be that provided at market rents, as is the case with most in the UK. Subsidised rental provision and ‘intermediate’ affordable forms of tenure are discussed in the previous chapters so the emphasis here is on private unsubsidised market-driven rental provision. This definition is designed to cut through the varied international definitions of the tenure (see Oxley et al., 2011). Owner occupation is regarded as the default tenure to which most people aspire, which seems a reasonable supposition (Gyourko, 2003; IPSOS, 2009; Wallace, 2010). At the most simple level, the answer to the question about the affordability of private renting is found simply by looking at who lives in the PRS and noting that the rents being paid must be affordable because the people in question are paying them and few are defaulting. But a more interesting investigation is that of exploring who benefits from the tenure’s existence and what types of agency supply market rental accommodation and why. That analysis highlights many issues underlying policy debates over rental housing and affordability. It helps to unravel whether potential tweaks to the rental offer, subsidised or not, make a significant difference. Of great
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c13.indd 255
2/7/2012 10:25:46 AM
256
Challenges of the Housing Economy
importance is the institutional framework under which rental provision takes place: that is the range of suppliers in existence, the range of legal and social rules under which they operate, and their place in the wider framework of housing availability. This institutional dimension means that discussion has to be grounded in one country, and the UK is used for that purpose, although the results have applicability elsewhere. The assessment here is based on looking, first, at some features of the PRS within the overall housing system: the ‘what and why’ of its customer base; the prospects for PRS demand and rents. The next section investigates the impact of extending security of tenure and of introducing second generation rent controls in parts of the sector. The final section considers the potential for attracting large investors to the PRS by assessing the role of economies of scale. International comparisons are used where relevant. The chapter is completed with a summary and conclusions.
The private rented sector within the housing system Growing importance of the private rented sector In the UK the PRS ended its long secular decline following liberalisation in the late 1980s. By 2008, well over a million more English households were private renters than two decades earlier (Figure 13.1). The pattern of expansion between the late 1980s and 2008 had three broad stages: 1. Tenancies grew by around a fifth after liberalisation up to the mid 1990s. Interestingly, this was during a period when real house prices were falling nationally, suggesting that the sector has distinct cyclical dynamics from those in owner occupation. 2. Expansion then slowed sharply between 1997 and 2001. This was despite the introduction of the buy-to-let mortgage in 1998 and was probably due to an increase in first-time buyers entering owner occupation at a then favourable stage in the house price cycle, many of whom were quitting renting. 3. Rapid expansion from 2003 to 2008. This period is popularly known in the UK as the ‘buy-to-let boom’. Rental demand increased, partly because of the strength of the economy but also because rising house prices put off first-time buyers during the strongest phase of a market upswing in the years prior to the financial crisis (Andrew, 2006). So, new investors entered during a period of booming rental demand and rising house prices, using newly introduced mortgage products that now had similar variable interest rates to those enjoyed by owner occupiers, whereas previously landlords had had to pay an interest rate premium and experienced more onerous mortgage terms.
Jones_c13.indd 256
2/7/2012 10:25:46 AM
The Private Rented Sector As a Source of Affordable Housing
4500
Thousands
257
16
% All dwellings
4000
14
3500
12
3000
10
2500 8
%
2000 6
1500
4
1000
2
500
-
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
Figure 13.1 Growth of the private rented sector, England, 1991–2008. Source: Department for Communities and Local Government.
The phrase ‘buy-to-let boom’ indicates the common assumption that the growth of the UK PRS over the past two decades was supply driven. Yet levels of rents and the number of tenancies are a function of the interaction of demand and supply, and a strong case can be made that developments in the PRS, once deregulation is in place, are driven mostly by variations in demand to which supply adapts with a lag. The prime reason for this is that vacancy has a high cost to landlords, and so supply rarely races ahead of demand. When demand falls, supply will also lag behind as it takes time to sell off properties. Variations in demand are influenced by general economic and demographic factors and also by what is happening in owner occupation markets, which are over five times larger in scale than the PRS and so have significant impacts on cyclical variations in private rental demand. The size of the PRS thus depends on demand drivers within it but, in addition, on what is happening in other housing tenures. Supply factors will also vary cyclically across the tenures. However, there is no literature at present on the detailed cyclical relationships between renting and owning. So, in the context of this paper, it is probably best simply to accept that housing market cycles in renting and owning are different. Whatever the cyclical dynamics, the assumption here is that owner occupation will remain for the foreseeable future the predominant UK housing tenure at roughly the current percentage share that exists now. Justification for this statement is based on people’s preferences, mortgage market and economic trends, and demographics.
Jones_c13.indd 257
2/7/2012 10:25:46 AM
258
Challenges of the Housing Economy
Housing cost interlinkages between renting and owning In a competitive market with ‘tenure neutral’ taxes and subsidies, there is a clear and simple long-run relationship between rents and the prices of dwellings, in that prices are the discounted values of future net rent streams (DiPasquale and Wheaton, 1996), where ‘net rents’ take account of administration, transaction costs and adequate repair costs to maintain dwellings. There are some differences in the costs of owning and renting, in their taxation, and also some suggested external benefits of owner occupation. So the precise long-run relationship is more complex in reality, but that does not alter the fundamental connection between discounted rents and property prices nor does the impact of cyclical factors in the short run alter it either. Looking at the long run gives an important insight into the essential point that the cost of housing is broadly the same in either renting or owning. So, if prices are high in one of them because of supply shortages that situation cannot be resolved in the other. Subsidies and regulations that try to do that are ineffective, inefficient and expensive. So, if housing is in particularly short supply, moving from one tenure to another makes little difference to that uncomfortable fact. This fundamental economic insight is of great relevance to discussions of renting and intermediate housing. Housing demand has been growing at a much faster rate than supply in the UK for many years. As a result, national trend house prices have been rising at 2.7 per cent a year in real terms over the past 50 years while earnings rose at only 2 per cent (Lloyds Bank, 2010). Forecasts of even greater shortages over the next 20 years suggest that prices may rise faster relative to earnings, once the recovery is firmly in place. Without substantial increases in housing supply, prices go up. As house prices go up, so will rents.
The complex dimensions of consumer choices There are important trade-offs that households make when paying for housing, because housing constitutes a complex bundle of attributes. Dwelling space, dwelling and public amenities and location are the main determinants of the relative differences between house prices. People can economise, or spend more, by altering all or any one of those characteristics. This multi-dimensionality makes it inherently difficult to discuss housing affordability issues and tenure options, because like is not necessarily being compared to like, unless care is taken. In order to be able to economise on housing costs in the face of rising prices (or rents), in the absence of subsidies, people have three choices: limit the amount of space they use, accept worse quality accommodation or move to a cheaper location. In such circumstances, people vary in their responses: for example, some move to cheaper locations, while others prefer to crowd up and gain the benefits of the location. Such differences may be reflected in
Jones_c13.indd 258
2/7/2012 10:25:46 AM
The Private Rented Sector As a Source of Affordable Housing
259
tenure differences with, say, younger people crowding up in lower quality but centrally located rental accommodation, while others move to the suburbs to find cheaper but more spacious owner-occupied accommodation. It is important from a policy analysis perspective to recognise such choice decisions inherent in tenure disparities.
Subsidies, policy responses and impact leakage Subsidies offer another housing possibility; say via social housing, through which some people’s housing can be provided at below market cost in a locality. Yet, unless there is additional local supply, there will be a crowding out of the housing opportunities of others. House prices will rise as ‘crowded out’ people chase shrinking opportunities for accommodation. As a result, everyone’s housing costs increase. Prices rise because demand is enlarged, through subsidy, without any commensurate increase in supply; with the binding supply constraint being related to land. Actions in one tenure, or a part of it, may consequently have spillover effects elsewhere. So, any policy evaluation must include such potential spillovers and feedback effects. Therefore, broad demand and supply situations matter a lot when discussing policy options related to a specific tenure. Moreover, the required level of subsidies to meet target aims, when there is crowding out, rises over time as overall housing costs rise. Subsidies and tax breaks also do not necessarily end up with their intended recipients. A good example was the removal of stamp duty (sales taxation) in the March 2010 UK Budget for first-time buyers on properties worth under £250,000 (HM Treasury, 2010). Given UK supply elasticities, this tax break was likely to have been quickly absorbed into higher purchase prices, so that the sellers of properties rather than their first-time-buyer purchasers gained. What is more, landlords wishing to purchase homes will have to pay the resultant higher house prices. So, it could have been the case that potential first-time buyers currently living in the PRS may be actually worse rather than better off because of the tax break designed to help them move into homeownership.
Life cycle characteristics and the role of the PRS When providing an assessment of the potential for the PRS to provide an alternative to homeownership for households who are either priced out of homeownership, or for whom the risks of homeownership are too high, four types of ‘priced out’ household need to be distinguished: 1. those that cannot become homeowners because of their current life cycle stage, i.e. young with insufficient savings
Jones_c13.indd 259
2/7/2012 10:25:47 AM
260
Challenges of the Housing Economy
2. those temporarily leaving owner occupation, because of say a job move/ temporary loss or a divorce 3. those that cannot afford owner occupation because of temporary sharp surge in house prices, as occurred in the last housing market boom, or because of increases in deposit requirements, as has occurred following the 2007–8 credit crunch 4. those who because of the levels or variability of their incomes cannot ever expect to be able to afford to become homeowners The role of the PRS in providing unsubsidised accommodation for the first three of those groups is well documented (see below). With respect to the last group, evidence is more variable. With that in mind, the following sections will outline the current role that the PRS does play.
The demand for housing in the PRS Younger and more mobile people in work are the main clients of the modern UK PRS. Over four fifths of private tenants are in employment or higher education (see Figure 13.2). Many are relatively prosperous, with median or higher incomes, and stay in the tenure as long as they value mobility. Taken together, private tenants have a broadly similar distribution of income to the population as a whole, with a somewhat higher share of middle and lower income groups and a markedly smaller share of those with the highest
70
%
All private renters
All tenures
60 50 40 30 20 10 0 Full time
Part time
Unemployed
Retired
Other inactive
Figure 13.2 Economic status of private tenants and all households in England, 2008. Note: Household reference person. Other inactive includes students. Source: Department for Communities and Local Government from Labour Force Survey.
Jones_c13.indd 260
2/7/2012 10:25:47 AM
The Private Rented Sector As a Source of Affordable Housing
261
incomes.1 The share of low incomes is influenced by the fact that around 10 per cent of tenants are students. The average age of private tenants is much younger than that in other tenures as shown in Figure 13.3. Even the fifth of private tenants that claim housing benefit, due to low incomes, tends to be far younger than the general age structure of the country’s population. The proportion of renters on housing benefit has fallen to about a fifth from around a third in the mid 1990s, so that the number has not risen in line with the growth in the PRS as a whole. The significance of mobility can be seen in moves data. The median length of a private tenancy is 1.7 years, whereas it is almost 8 years for social tenants, and owner occupiers on average move only every 12 years, though those with mortgages move more frequently at around 7 year intervals. Some 80 per cent of private tenants have lived in the same tenancy for five years or less. The remaining 20 per cent include those still living in a rump of rent-controlled, secure tenancies or in long-term work-linked accommodation, so that the number of long-term tenants in the free market is very small. The scale of mobility in the PRS is indicated by the fact that 40 per cent of private tenants have moved at least once in the previous year.2 Most of this high PRS mobility is driven by tenant choice rather than landlord push.
60
Priv. renters Soc. renters Mort. owners
50
Outright own 40
30
20
10
0 Under 35
35–44
45–64
65 or over
Figure 13.3 Age distribution by tenure in England, 2007/8. Note: Household Reference Person. Source: Department for Communities and Local Government.
Jones_c13.indd 261
2/7/2012 10:25:47 AM
262
Challenges of the Housing Economy
Once households become less mobile, they tend to move into owner occupation or, less commonly, into social housing. Another group of tenants, typically somewhat older, are those who have been in a relationship that has broken down and have moved into rented property as a consequence – at least for a while. So, living in the PRS is a life cycle stage for many tenants. As circumstances change, most become owner occupiers. This phenomenon can be seen in moving surveys: 27 per cent of recent owner movers were previously private tenants, which is a substantial share of existing renters, given the relative absolute numbers of homeowners and private tenants.3 This suggests that most private tenants become homeowners for the first or second time after some years as renters. A fifth of recently moved social tenants were also previously in the PRS, again suggesting that potentially eligible social tenants live in the PRS only for as long as necessary in order to obtain a social home. Consequently, relatively few people at present are long-term renters in the market sector. It is hard to identify them precisely from the available data. The small sample sizes on which published PRS information is based mean that specific, relatively small, subgroups within the sector are subject to large sampling errors. The available tables also only provide two-dimension cross-tabulations, which are insufficient to identify particular subgroup experience. The main influences on whether households opt for renting or ownership relate to household characteristics. Childless, unmarried, younger households are much more likely to rent privately, and those in any tenure tend to move within it rather than switch to another one. Incomes are also important, with those on higher incomes more likely to be owners. Final key factors are the ease of getting a mortgage and the size of the deposit required to purchase (Andrew and Meen, 2003; Gyourko, 2003). There is a two-way relationship between the availability and quality of rental accommodation and people’s desire to live within it. It is now feasible for younger British households to rent decent accommodation on a scale impossible for older generations, because of the prior existence of rent control. This has helped to delay the age at which people bought their own homes. One of the biggest influences on the growth of renting, however, has been worsening housing affordability, especially building up enough equity for a deposit. Rising affordability problems have delayed moves into owner occupation: the UK share of young adults aged between 20 and 24 entering owner occupation fell from 31 per cent to 20 per cent between 1991 and 2006 and from 63 per cent to 46 per cent for those aged from 25 to 29 between the same dates.4 The prime driver of the size of the PRS is demand rather than supply. The growth of the PRS has not been associated with a rise in vacancy rates, so
Jones_c13.indd 262
2/7/2012 10:25:47 AM
The Private Rented Sector As a Source of Affordable Housing
263
that landlords in the main have not mis-forecast demand. Rather the demand drivers have been strong, especially in London and the south, due to several factors: ●
●
● ●
growing numbers of households in the age ranges with highest demands for renting, i.e. younger households. This has partly been fuelled by domestic demographics but also by immigration, because immigrants tend to be in the younger age ranges; growing affluence and associated mobility among younger people has encouraged them to set up independent households; mobility requirements in some occupations have risen; and greater price barriers to entering owner occupation and, to a lesser degree, lower availability of social housing.
If there is a demand in a free market, supply will follow as long as it is profitable to do so, and since deregulation in the late 1980s, that has generally been the case in the UK’s PRS. There is a common misconception that the bulk of investment in the PRS over the past decade has been related to newly built flats.5 Most landlords find that tenants prefer cheaper accommodation, so unsurprisingly the vast bulk of the private rental stock is found in the older stock of housing, especially in inner city and older suburban areas, where investors have purchased and often improved existing properties (Ball, 2006).
Is the private rented sector peaking? Demand drivers have shifted in a variety of ways, as a result of the 2007–8 financial crisis and its aftermath: ●
●
Jones_c13.indd 263
Mortgage availability has shrunk. This forces some households to remain longer in the PRS (HCA, 2010). However, this phenomenon has to be viewed over a five-year-plus horizon, when mortgage availability should be improved as financial systems gradually emerge from crisis. Mortgages are unlikely to be as easy to obtain as in the final years of the pre-crisis boom but they may not be that much more difficult than in earlier periods. The cost of mortgages may rise somewhat with greater regulatory requirements but it is the cost of entering owner occupation in the form of the scale of the personal equity required that is the greatest barrier, not monthly outgoings. The long-run price of housing is rising at a faster rate than earnings, so that the burden of the deposit on first-time buyers may grow somewhat. This may also force some households to remain longer in the PRS. However, rents will be rising as well, deterring consumption there.
2/7/2012 10:25:47 AM
264
●
Challenges of the Housing Economy
Off-setting this price dynamic may be a slower than previous trend growth in real house prices; longer stays at parental homes by younger people; further reliance on parental funding for first-time buyer deposits; more low-cost homeownership mixing renting and owning; and financial innovations that reduce the impact of down payments. Demographic forecasts for England suggest that the number of younger households is going to rise at a much slower rate than that for older households: 90 per cent of the predicted annual 252,000 increase in household numbers between now and 2031 is expected to be in 35+ age groups. Those groups are currently not big clients of the PRS and are unlikely to be so in the future.6
The demographic point is the most significant. The mobility data quoted earlier suggest that households move less as they age. Older people are particularly reluctant to move out of their homes (Venti and Wise, 1990, 2000). Staying as homeowners makes sense for them, because they are shielded from rising house prices in that way and benefit from the absence of taxation on imputed rental income (Rouwendal, 2009; Ball et al., 2011). Demographic pressures may be partly offset by the growing housing expectations of younger households as their incomes rise over time, with some prepared to pay higher rents for better accommodation. However, as rents are likely to rise broadly in line with house prices, other renters may choose to crowd up or quit and stay with parents instead, so that the overall effect is probably neutral.7 Weighing up the various demand factors overall suggests that the tenure share of UK private renting may stay at the same level as present, assuming that housing benefit and social housing remain broadly as they are. Private renting will still expand absolutely as overall household numbers grow. In this analysis, a big unknown is the size of entry barrier for first-time buyers arising because of the need for deposits. Nowadays, typical first-time buyer down payment ratios are of the order of 25 per cent rather than the 10 per cent seen at the end of the last boom, while debt servicing ratios of 29 per cent indicate that monthly repayments as shares of income remain above trend levels.8 Such ratios may improve somewhat when mortgage markets free up from their currently highly constrained states in the foreseeable future but they are unlikely to bounce back to pre-crisis levels (FSA, 2009). Does this mean that far more frustrated first-time buyers will remain as renters? The answer is yes and no, but the overall conclusion strongly suggests that there will not be a surge of younger households entering or remaining in the PRS. This conclusion arises for a mix of the reasons already mentioned. The secular decline in first-time buyers has already occurred in the UK, down from almost 600,000 a decade ago to 200,000 in 2009, which helped to
Jones_c13.indd 264
2/7/2012 10:25:47 AM
The Private Rented Sector As a Source of Affordable Housing
265
boost the size of the PRS over that period as noted earlier. Any further trend reductions in first-time buyer numbers, if they do occur (which is debatable) will be from a lower base and so have less proportionate effect on the size of the PRS relative to owner occupation. Demographic developments do not suggest major pressures either, as already noted. Substantial future increases in the size of the PRS therefore seem unlikely to be realised in practice from this source. The remaining substantial source of potential expansion of the PRS is if social housing were reduced in size and the PRS took over part of its rental role, but this scenario is beyond the scope of this chapter.
Potential policy developments Would better security of tenure assist low income groups? The widespread use of assured shorthold tenancies (AST) is now firmly entrenched within the UK’s PRS. These were introduced in 1989 for new tenancies when the sector was deregulated and rents of these tenancies are set by the market. Under these contracts, security of tenure exists for only the initial six months. Some fear that AST’s lack of security creates significant long-term problems for tenants. Rugg and Rhodes (2008) argue that a climate of uncertainty is generated for tenants far out of proportion to actual notices to quit. However, whether or not AST actually generates significant problems is debatable and there are substantial costs to having long-term tenant security, which may in practice benefit only a few households and damage prospects for others. The evidence is mixed on the basis of international comparisons. Many European countries have greater security, either through medium-term contracts, which may be renewable, or through explicit or implicit rules. Yet they simultaneously have rent controls as well, so that the existence of greater security of tenure may be an outcome of such controls, rather than a clear benefit in its own right. By contrast, other European countries adopt a similar stance to the UK, e.g. Ireland and Finland, as do most areas in the USA or Australia, even though both those countries have cohorts of longterm, moderate income renters and small or non-existent social housing sectors (Haffner et al., 2008; Ball, 2009). There are also reasons to think that market mechanisms work to enhance security implicitly. Tenants requiring long-term accommodation will gravitate to landlords that are happy with such arrangements; either through active search or through the fact that reluctant landlords will evict them. Importantly, landlords have incentives to value good existing tenants willing to pay somewhere near the going market rent. When tenants quit, landlords face the risk of a prolonged vacancy – vacancy rates average a month or more
Jones_c13.indd 265
2/7/2012 10:25:47 AM
266
Challenges of the Housing Economy
according to the Association of Residential Landlord Agents data. Landlords have to spend on redecoration and minor maintenance in order to attract new tenants. New tenants are also unknown quantities. They may fail to pay the rent on a timely basis or generate maintenance and administration. New tenants may themselves be suspicious of landlord motives and so be more reluctant to cooperate. Such asymmetrical information between tenant and landlord is diminished over time as each gets to know the other better. To avoid these risks associated with new tenancies, existing tenants may enjoy lower rents, especially in situations where rents are generally rising. In the literature, there are a variety of cases where it is shown that longer-term tenancies are associated with lower rents in free rental markets (see later). Security of tenure is not offered in other important aspects of life, such as in employment. In fact, a lack of complete job security is seen as essential to create the appropriate workforce incentives. What is generally regarded as far more important in the labour market is that a plentiful supply of alternative employment exists. A similar incentive view can be argued for in housing. The Hills (2007) Report also argued that total security of tenure may not be ideal in English social housing. Security of tenure needs to be distinguished from a typical long-term contract. Security of tenure is asymmetrical in its terms, because under most specifications the tenant can costlessly quit at any time but not the landlord (Law Commission, 2007). By contrast, typical long-term contracts outside of the housing sphere specify penalty clauses for early termination. For example, in other areas of real estate such as the commercial sector, tenants take out binding leases for long periods of time (Crosby et al., 2006). Residential landlords may be more willing to negotiate binding long-term contracts that guarantee a long-term income stream, rather than negotiate on the basis of asymmetrical contracts solely guaranteeing tenants’ security of tenure. The latter leave landlords with a vacancy risk as well as reduced rights of property possession, whereas the former contain only default risk. Symmetrical contracts are commonplace when private investors rent their properties to local authorities, under private sector leasing programmes, which offer contracts for several years or more in exchange for a discount on market rents. For example, in London around 40,000 dwellings are leased in this way (Find-a-Property, 2011). It may be argued that social housing institutions have managed to achieve security of tenure for their tenants. However, they avoid vacancy risk in most areas by charging rents that are significantly below market levels, often backed up with housing benefit. These arrangements often create queues for properties that obviate vacancy risk. Yet, in fact, those queues often do not actually remove the existence of vacancy risk but transfer it to a private landlord, in cases where new social housing tenants relocate from the PRS.
Jones_c13.indd 266
2/7/2012 10:25:47 AM
The Private Rented Sector As a Source of Affordable Housing
267
Security of tenure can also create unfortunate filtering rules. If tenant security exists, rational landlords are still likely to value the benefits to them that are associated with limited security of tenure. They include the low cost removal of poor-paying or badly behaved tenants, or of just being able to get back vacant possession of their property, for example when they wish to sell it. So, they are likely to act strategically and try to select types of tenant that limit the potential adverse costs of extra security. Therefore, if they are required to offer security of tenure, they will prefer to house tenants with a low probability of staying long, or who will be ‘good’ longterm tenants, and will consequently filter out potential applicants for whom tenure security would be more worthwhile. Paradoxically, therefore, the risk of tenure security becoming a binding constraint may shrink the housing opportunities for those wishing to live in rental property for long periods of time, especially the most vulnerable. What is more, landlords will not have detailed information on potential tenants and so are likely to use rough and ready personal perceptions about particular types of tenant and their behaviour. The potential for undesirable discriminatory practices is thereby enhanced, because the perceived costs to landlords of non-discriminatory behaviour are greater if they cannot cancel tenancies at short notice. Empirical evidence is relatively limited, but some does exist. For example, evidence has been cited for Germany, where the filtering, discriminatory effect is argued to be strong (Basu and Emerson, 2000). Selecting only tenants that are likely to move within a relatively short period of time is one noted strategy. There are very high rates of tenant turnover in some European cities with high levels of tenure security, such as in Zurich, where a fifth of households move each year (Thalmann, 2010). Tenure security combined with rent controls can also alter tenant behaviour. One commonly cited effect is reduced mobility, because tenants do not wish to give up secure, low-cost accommodation (Krol and Svorny, 2005). This can be particularly detrimental to the participation of economically vulnerable groups in labour markets, because they are unlikely to search far for a job if unemployed. Security of tenure is closely associated with rent control, because without some form of rent control, landlords could simply circumvent tenure security rules by raising rents until tenants left. Rent control is common in some parts of the USA on a localised basis (Roistacher, 1992; Early, 2000), but is widespread throughout Europe on a national basis, particularly in the form of second-generation rent control whereby initial contracts are freely negotiated but subsequent rent increases are linked to some marker approximating to general price inflation (Haffner et al., 2008). In some European countries, rental contracts are associated with symmetrical obligations in terms of contracts. For example, most contracts
Jones_c13.indd 267
2/7/2012 10:25:47 AM
268
Challenges of the Housing Economy
in Belgium are three-year contracts, after which the landlord can regain possession, if desired, and rents can be freely renegotiated for another term otherwise. Consequently they are argued to be quite close to free market conditions (Baudewyns, 2007). In contrast, security of tenure in Germany is virtually permanent in practice, because courts are reluctant to permit eviction except in circumstances of tenant default or poor behaviour and, even then, the process is costly and time consuming. The precise consequences on housing investment of both security of tenure and rent controls depend on housing supply elasticities. If housing supply is relatively plentiful and controlled rents are not held down by some exceptionally strong price control, the benefits to tenants of staying in secure accommodation in relation to a freer rental contract situation may be limited, because they can potentially easily move to an equivalent free market property. The costs to landlords are thereby lowered by the greater probability that tenants will do so. Where supply responses are good, long-term market rent rises are likely to be limited; whereas in places where demand strongly outpaces supply, the difference between controlled and free market rents is going to be greater. Major cities are a case in point. This means that the incentive of tenants to stay in their existing tenancies is higher, while the interest of landlords in selecting tenants who move on rapidly is all the more substantial. Another crucial factor is the local supply and demand balance. Where the demand for rental housing is falling relative to supply, most of the constraints discussed above would not be binding. This could occur possibly because of a movement of more prosperous households out from rental (inner city) districts to (owner occupier) suburbs. This has taken place in many cities in advanced economies at various stages over the past 50 years. In some cases, such as Brussels, that situation is still occurring. There might also be quite long-term national declines in house prices and rents because of over supply, which again makes the constraints non-binding. This happened in most of Germany for a decade after 1996 and also in Switzerland and Austria for an extended period around the same time (Ball, 2009). Another potential influence is the ease with which rental housing can be converted into owner-occupied property. In Italy, there has been a largescale transfer of rental flats into owner occupation, so that the homeownership rate has risen to 80 per cent. Rent laws changed in the late 1970s and the policy inadvertently enabled landlords to sell out. The Fair Rent Act of 1978 established a common four-year renewable lease and continued rent controls. This made the dwellings far more valuable in owner occupation than in renting and enabled landlords to sell out when leases came up for their four-year renewals. A similar sort of process has been occurring in Sweden and to a lesser degree in Denmark for the past ten years. Blocks of previously rented properties have been converted into tenant cooperative
Jones_c13.indd 268
2/7/2012 10:25:47 AM
The Private Rented Sector As a Source of Affordable Housing
269
owner flats, which is a slightly complex, institutionally determined way of creating de facto owner-occupied flats out of previously low-return securetenure rental property. This ‘tenant owner’ tenure now represents over 20 per cent of the Swedish housing stock and more in its major cities. Thus, the impact of security of tenure depends on the length of rental contracts; their symmetry in terms of obligation between landlord and tenant; the ease with which landlords can remove tenants, if desired, at the end of contracts; the type of rent control mechanism that is associated with security; the price elasticity of housing supply; the local demand and supply balance; and the ability to transfer private rental housing into other tenures. Undoubtedly, with such a large number of potential variables, the overall consequences of them are likely to be highly variable when estimating either the impact on housing investment or the return landlords would require in order to offer greater security of tenure. One clear conclusion to emerge is that the adverse impact on investment is likely to be far worse (and the returns required by landlords for more security greater) in situations of considerable housing shortages and high housing costs. A corollary, given the poor British supply responsiveness, is therefore that the consequences would probably be adverse if adopted in the UK but still highly variable across British cities and particularly bad in London, with its housing shortages and high privately rented share. It is by no means clear that the introduction of greater security of tenure would offer benefits to long-term, potentially vulnerable, private tenants. In fact, the market consequences of introducing such regulations may well make them worse off, as they would face diminished supply and be discriminated against in competition for what remains. Any inducements to encourage greater security of tenure are consequently likely to be ineffective in meeting their stated aims.
Would second-generation rent controls offer benefits? The previous section was vague on the precise content of rent control. Some types have been grouped together and termed ‘second-generation’ rent control, which is supposed to be smarter than the ‘first-generation’ fixing of rents at the nominal level of a particular date in an era of rising price inflation, which then rapidly erodes their real value (Arnott, 1995). There are three characteristics of second-generation rent control: 1. free market agreement of initial rents; 2. subsequent ‘cost based’ rent increases only; and 3. security of tenure for sitting tenants for fixed or indefinite time periods. Together they imply a particular pattern of rent adjustment. However, landlords and current and potential tenants react to the existence of such
Jones_c13.indd 269
2/7/2012 10:25:48 AM
270
Challenges of the Housing Economy
Path of free market rents
Rent
Path of tenant rents set at beginning of each period
1
2
3
Time/period
Figure 13.4 The growth of rents under second generation rent control illustrated. Tenancy starts in period 1 and are reset each ‘period’ of, say, three years, at a time when market rents are rising strongly.
controls. The consequences are that the apparent tenant-benefits of second generation rent control may be limited. This can be seen with a simple sequence of argument. Initially, assume that everyone has full information. With perfect foresight about the time a tenant is going to stay and perfect forecasts of future rents, landlords would ask for, and new tenants would accept, rents that match the discounted flow of rent in a free market over the time of the tenancy minus the permitted increases in rents during the tenancy. In a context of rising rents, initial rents would be higher than market rents because the rent control measure means that landlords will frontload rents. This pattern of rent change is illustrated in Figure 13.4. Each tenancy lasts a fixed period of, say, three years and rents are not permitted to increase during that period. At the start of the period, when the rent is negotiated with a new tenant, the rent is higher than the market one and then declines relative to it over the contract period, only to rise above it when rents are negotiated again. In a competitive market, landlords will not earn excess returns through this strategy, because average landlord net returns are influenced by the flow of investment into the rental sector, which will continue until it equals the risk-weighted prevailing return on all investments. Apart from the altered pattern of rent payments, the overall effect of such rent control legislation in the perfect foresight context on housing building, trend rent levels and maintenance standards is limited. This is primarily because landlords have set rents that incorporate all potential tenant rent reduction
Jones_c13.indd 270
2/7/2012 10:25:48 AM
The Private Rented Sector As a Source of Affordable Housing
271
gains occurring during the course of the controlled tenancy (McFarlane, 2003). But, at the same time, the point of having the legislation itself is considerably diminished as well. In the more general situation of uncertainty, the impact of secondgeneration rent regimes is less clear than in the world of perfect foresight. Rent controls act as a barrier to unexpected increases in rents. Obviously, because they live in the real world, tenants only demand rent control legislation in situations of uncertainty rather than of perfect foresight. But the principle that landlords will try and compensate themselves for any potential loss of return by frontloading the rents agreed with new tenants still prevails. In effect, tenants pay an initial ‘insurance premium’ for having fixed rents during a given period, whether they want the resultant fixed rent or not. For example, research has found that tenants living in secondgeneration rent control properties in New York initially pay higher rents than those in the free market sector but tend to stay longer and enjoy increasingly lower-than-market rents as the lengths of their tenancies increase (Nagy, 1997). The more landlords bear the costs of uncertainty over future real rent increases and inflation and the less they know about the length of time a tenant is going to stay, the higher initial rents are likely to be set in order to compensate them for any potential expected loss of return. As a corollary, less housing will be supplied on the rental market unless landlords are confident of earning sufficient return. Moreover, the tenant selection effects discussed earlier will operate so that landlords are likely to prefer applicants that they expect will stay only a short time or be ‘good’ long-term ones, when tenants have a right to remain in the property for long periods of time. Landlords will consequently tend to discriminate against riskier or potentially more costly households with a higher probability of staying for a long time. Typically, younger and childless households are likely to stay for a shorter time and subgroups of them be perceived as good long-term tenants. Therefore, other types of household may find it particularly difficult to find accommodation because of landlord fears that they will stay or be high cost. So they may be forced to move less frequently, perhaps only be able to rent in a higher priced ‘free’ sector and stay in worse accommodation because of lower market opportunities. What is more, despite their attempts at broad categorisation of tenant types, landlords will still not know which tenants are actually going to stay longer. There is asymmetrical information, because tenants will have much clearer views of their plans. An adverse selection situation may well occur as a result, with higher equilibrium rents for all and lower supply compared to a free market context (Basu and Emerson, 2000).
Jones_c13.indd 271
2/7/2012 10:25:48 AM
272
Challenges of the Housing Economy
Potential attraction of large investors Who invests in rental housing? Mainly small-scale landlords Around the world, most residential investment is owned by private individuals in one form or another. This preponderance is often put down to such motives as speculation on house prices, but the PRS is, and has always tended to be, a small individual investor sphere of activity, especially in free market contexts. The principal causes are economic, related to the nature of the rental business and the investment characteristics of the asset, as is explored further below. Three-quarters of English landlords are private individuals and couples; about 10 per cent are property companies and the rest a mix of other organisations.9 This small-scale characteristic was preponderant in the PRS’s heyday, prior to 1914 (Daunton, 1983; Ball and Sunderland, 2001), and has been reasserting itself over the past two decades as residential renting has once again become an attractive investment. In the mid 1990s, less than half of landlords were individuals but the long housing boom saw many firms sell up or run down their property holdings, while the opportunity to buy to let through rapidly rising house prices encouraged many thousands of new investors to become landlords. Surveys suggest that many landlords own only one or two properties, giving rise to the impression that the majority of tenants’ landlords are small-scale investors. However, the growth of the small landlord should not be exaggerated, precisely because most of them own only one or two properties, while larger investors and companies each own many, so that ownership of the stock is probably still concentrated among those that have at least 10 or more properties each. While unfortunately the data do not allow this important question to be adequately answered this ownership structure blurs the oft made distinction between small ‘amateur’ and larger ‘professional’ landlords. When asked, only a quarter of English landlords actually regard housing as their prime source of income. Surveys have also suggested that landlords have different motives, with some less return-oriented than others. Self definition may be a worry as people tend to want to be seen as ‘nice’ rather than ‘money grubbing’. But even if this behavioural difference is true among investors, rents and rental housing supply are, at the margin, more likely to be determined by prevailing market rates of return rather than some putative ‘satisficing’ one. There may be some filtering within the market of particular tenants living in properties owned by specific types of landlord but that is unlikely to alter market outcomes much. Moreover, it must be questionable whether the behavioural experiment can be extended to one where all those claiming to be willing to accept lower returns for social or other
Jones_c13.indd 272
2/7/2012 10:25:48 AM
The Private Rented Sector As a Source of Affordable Housing
273
reasons can be grouped together and actually given lower returns. In other words, it is risky to base analysis and policy on potentially uncertain beliefs in investor altruism. In the UK, private landlordism has come to be associated with buy-to-let mortgages, yet the roughly 1 million such mortgages in existence account for only around a third of private tenancies. Larger landlords will typically take out other types of residential or commercial mortgages, on which there is far less information. In addition, it is likely that much of the PRS is held by individuals with no mortgage debt on some or all of their properties. However, many own multiple properties and have mortgages on some of them. Given high house prices, the values of investor portfolios can be substantial. Lower quartile (i.e. the price of a dwelling at the 25 per cent point in the house price hierarchy) house prices are likely to reflect typical landlord investments. In the south-east in 2008 it was £162,000, so that a landlord portfolio of ten dwellings at that price would be worth £1.6m and a portfolio of a 100 would be worth £16m. Similar calculations can be done for other countries. Residential renting offers flexible leverage and labour-oriented opportunities to investors, as well as potential tax benefits (Glascock and Turnbull, 1994) even though the tax situation varies widely across countries. Small- to medium-sized landlords holding portfolios of from 10 to around 100 or so properties seem to be cost competitive relative to larger ones over wide portions of the rental housing stock.10 They can flexibly invest their own time; do not have high overheads; can move fast in response to market signals; and the relatively larger ones holding scores of properties can easily spread their investments over a variety of local markets and thereby spread their risks. In particular, small and medium-sized enterprises can compete effectively with larger ones within the diverse and scattered stock of properties that constitutes most of the private rental stock. The majority of rental housing is spread around neighbourhoods of quite old, terraced and semi-detached houses and flat conversions within this house type (Ball, 2008; Glascock and Turnbull, 1994). New properties in blocks of flats – though often thought to be typical landlord territory – are, in fact, comparatively rare within the private rental stock. New rental properties are often derived from renovations and conversions of previously single-family properties into flats. This has the spin-off benefit of making intensive use of the existing stock in places where it might be otherwise underutilised and poorly maintained. Such conversions and modest upgrades are often found in neighbourhoods of relatively moderately priced properties. This can assist in regeneration and in avoiding neighbourhoods slipping over into cumulative decline (Parkinson et al., 2009). If, as there seem to be, limited economies of scale in the business, many small landlords can offer as good a service as larger ones and be competitive.
Jones_c13.indd 273
2/7/2012 10:25:48 AM
274
Challenges of the Housing Economy
Table 13.1
Characteristics of owners of multi-family properties in the USA. Units:
Type of owner (% – numbers do not add to 100) Individual investors, husband/wife Trustee for estate Limited partnership General partnership Real estate investment trust (REIT) Real estate corporation Other corporation Non-profit/church-related institution Owner residence on this property Owner owns one rental property
Total US 86 2 3 3 1 2 1 1 24 48
Less than 5 units 90 2 2 2 1 1 1 0 27 52
5–49 units
50+ units
74 3 6 5 1 4 3 2 10 30
31 1 25 13 3 11 6 6 3 15
Source: US Bureau of Census.
Current survey evidence in England does not suggest that larger private landlords or social housing institutions have more satisfied tenants than small ones (Rugg and Rhodes, 2008).
Role of large investors in the private rented sector internationally The PRS is generally a small-scale operation around the world. This is the case in Australia, the USA and Europe including in the UK. Institutional investors exist in many countries but they tend to have special characteristics, operate in particular parts of the rental market and have small overall market shares. The USA is an illustrative example. Table 13.1 shows the ownership structure of rental housing in the USA. It shows that only around 8 per cent of the stock is owned by large-scale investors, despite a tax friendly environment and the potential for widespread initiatives at state and local as well as the national level. However, there are some major firms, and they generally own and manage large, standardised apartment blocks for young, middle- to low-income households but, given the overall size of the US rental housing stock, their impact is still small in aggregate (Jones, 2007). Residential real estate investment trusts (REITs) have failed to make a significant impact in US rental housing, with only 1 per cent of the stock. A focus on the mobile, affluent market has been the experience and REITs have been active in the growth cities, especially in the south and west. They have, however, had little impact on private rental housing in the older industrial cities. The key is that REITs in the USA are dividend-driven vehicles, and older cities, where the need for affordable housing is strong, do not have clients that can pay for new REIT housing. However, growth areas such as Dallas and San Diego, have a younger wealthier clientele who wish
Jones_c13.indd 274
2/7/2012 10:25:48 AM
The Private Rented Sector As a Source of Affordable Housing
275
for more upscale housing and demand flexibility of terms: e.g. month to month leases after a first six- or twelve-month base. REITs can and have provided rental housing to this segment of the market. These growth cities also have the demand to afford large-scale projects that ideally fit the REIT structure of funding and management (Ball and Glascock, 2005). Residential REITs hardly exist in Europe. The one in Belgium has remained a small niche player. When German REITs were introduced several years ago, they were explicitly restricted from active involvement in the residential sector after strong lobbying by those that feared they would be less tenant friendly than existing landlord owners (Ball, 2010). This highlights the fact that the idea of superior service from large-scale providers may not be believed by tenants, who in the German case were frightened, rightly or wrongly, that the bottom line would take precedence over their interests. Large-scale investors exist in Germany because of the stipulations of social housing legislation, which converts social housing to private ownership after a stipulated time period and due to privatisations by municipalities and employer-owned estates. Some of these were bought up by foreign investors with the aim of upgrading and converting properties through tenant purchase. In the main, these investments have not proved to be particularly successful, with tenants reluctant to take on additional costs. A number of investors sold out after only a few years (Ball, 2010). Elsewhere, insurance companies have been pulling out of the PRS due to unattractive returns, as in Finland (Loikkanen and Lonnqvist, 2007). In the Netherlands, financial institutions own less than 5 per cent of the housing stock, concentrating on the upper market ‘free rent’ sectors. Rabo Bank has a major real estate division, which includes residential and an active development arm within its portfolio. It operates across much of Europe. This indicates that it is possible for financial institutions to build up successful real estate operations, though Rabo Bank has a long historical background in the area. It is notable that it spreads risks by being active across a wide range of real estate activities across a variety of locations, while retaining a specialist core in the Netherlands and links with social housing institutions. Switzerland is virtually unique in that its major pension and insurance institutions have substantial holdings in Swiss rental property. Switzerland, of course, is renowned for its high share of private rental housing at almost 70 per cent of the housing stock. Yet, even in Switzerland the majority of rental properties are owned by small-scale individual landlords. It is unclear why institutions own such atypically large holdings but it may well relate to taxation, regulatory matters and corporatist-style state business relations. The existence of large-scale investors does not lead to cheap accommodation. A recent study identified the high price of Swiss housing as a reason for the low homeownership rate (Bourassa et al., 2009). It is worth
Jones_c13.indd 275
2/7/2012 10:25:48 AM
276
Challenges of the Housing Economy
noting from the Swiss case that a preponderance of renting did not lead to housing market stability – with a substantial market crash occurring in the second half of the 1990s.
Constraints on large-scale investor roles The biggest constraint on large investor involvement in the PRS is their relative competitive position. The predominant mix of private rental property and the costs of managing it often favour small-scale landlords. There are also problems with regard to the typical rental property. There is substantial ‘churn’ with properties being bought and sold by landlords. They sell to each other as well as sell to and buy from owner occupiers and, additionally but less so, from housebuilders. It is consequently impossible to identify a given rental stock but rather a changing pattern of private rental tenancies within the UK’s total housing stock. This trading pattern is not conducive to the operations of large-scale investors. Typically, they want to hold a large block of property as a unit over the long term; although companies like Grainger Trust have a trading role as well which often involves selling previously regulated (i.e. subject to rent control) rental stock that have become empty (Jones, 2007). There are segments where large-scale investors are likely to have clear competitive edge in residential investment. Such opportunities probably exist mainly when scale economies or enhanced skill sets, or both in combination, create distinctive and cost-effective rental products. In other words, large residential operators can prosper particularly well where they can take advantage of potential scale advantages arising through branding, in purchasing, via specific forms of quality control or in other management fields. Empirical evidence in the UK of such investment possibilities already exists and its potential for expansion is shown by the several new fields of direct residential investment developed by large-scale investors over the past decade. A number of firms and funds now exist with £100–500m portfolios of UK residential property. They have grown to be major players in such areas as accommodation for students, key workers and retirees. Other growth areas with significant potential are serviced apartments for mobile, high-income professionals and healthcare-related housing. Student housing illustrates the scale and growth of such investment opportunities. Many students rent accommodation from small-scale landlords, with some neighbourhoods in university towns almost entirely used for student housing. Even so, there are over 450,000 bed spaces in student halls and over a 130,000 of them are in halls owned and run by private firms. This market was estimated to be worth £6.6bn in 2007, according to Savills, and had almost doubled in value over the previous two years. This indicates
Jones_c13.indd 276
2/7/2012 10:25:48 AM
The Private Rented Sector As a Source of Affordable Housing
277
the scale of new investment taking place, and forecasts point to further strong growth over the next few years. Growth is being encouraged by substantial increases in rents over the same period, due in part to the greater provision of en suite accommodation (Savills, 2007). The small-scale and the large investor both operate in the student housing market and the largescale ones have been able to prosper due to an ability to differentiate their product and offer a distinctive service package. Generally, large-scale investors regard residential property less favourably than commercial property because of several characteristics of the asset class. Not all of these will hold for all investors but taken together they offer a picture of a less attractive asset class than commercial property: ●
●
● ● ● ●
high investment transaction costs linked to the too-small-for-them scale of residential investments a lack of matching investment income with investor liabilities, because returns contain a significant capital element that cannot be realised without selling the properties intensive management and maintenance high tenant turnover and vacancy rates potential reputational issues illiquid, thin markets that make it difficult to sell blocks of flats on a timely basis or to work out what is their fair market value at some points of the property cycle
Yet, despite these drawbacks, advantages remain in the long-run capital returns associated with housing and the relatively uncorrelated nature of those returns with other types of property, which facilitates portfolio diversification. The problem for large investors remains how to mobilise those benefits, while minimising the costs. Indirect residential investment (i.e. holding the shares of property companies that own rental stock) offers some benefits. Yet a recent study argued that a low-level equilibrium may exist because of problems in matching the demand for, and supply of, residential property. On the supply side, the existence of firms and funds holding residential assets requires willing purchasers of their offerings but, on the demand side, potential investors may be wary unless there are truly deep and transparent markets in indirect residential, which there are not at present. A chicken and egg situation may easily arise, leading to a relatively low level of investment activity (Ball, 2007). However, this is not a problem that governments can easily resolve and the social returns from doing so may in any case be limited. Markets may themselves overcome some market failures, especially when given time. There have been policy attempts to increase large-scale investor interest in the past but these have foundered, partly because of the complexity and
Jones_c13.indd 277
2/7/2012 10:25:48 AM
278
Challenges of the Housing Economy
narrowness of the subsidy offer but also because they did not overcome investor concerns (Crook and Kemp, 1999). Residential REITs have been absent to date in the UK and they are relatively small in scale in other countries. There has been recent investor interest. For example, the Aviva insurance group has been contemplating launching a £1bn fund to invest in blocks of a hundred or so units in regeneration areas in the south-east and near transport hubs (Jones, 2010). The aim is to introduce mass-produced rental housing, which is common in some US cities, for middle market tenants such as younger office workers. However, even if there were 10 or 20 developments like this, they would have a relatively small impact on the rental market as a whole. All the same, such developments are to be welcomed as they offer new supply at a time of considerable constraints. This supply is not targeted at the lower end of the market but obviously helps ease rent pressures by providing extra accommodation, which can be particularly beneficial in local areas of high demand. In summary, there are benefits for large-scale investors in residential – and this is encouraging some current interest – but there are also dis-benefits and substantial competition from smaller-scale landlords. This relative imbalance is likely to explain the comparatively low scale of investor activity, and also the failure of previous policy attempts to expand large-scale investment, as much as any inherent issues with them as policies.
Summary and conclusions The arguments in this chapter can be summarised briefly as: ●
●
●
●
●
●
Jones_c13.indd 278
The PRS is relatively good at doing what it does at present, but there are limits to what it can do. It is unlikely to increase in relative size by much in the UK, despite growing affordability problems, though it will expand absolutely with rising household numbers. It is impossible to compensate for general supply-side-induced housing shortages through adjustments in the operation of specific tenures, such as social renting or intermediate affordable housing. In situations of shortage, subsidising one group or tenure crowds out housing for others. Currently, the PRS provides lifetime homes for only a few households; instead it is a ‘life style’ or ‘stage in life’ choice. If the PRS were to increase its role as a permanent tenure for some households, moving towards greater security of tenure or towards secondgeneration rent controls are unlikely to improve, and will probably worsen, the housing situations of those households.
2/7/2012 10:25:48 AM
The Private Rented Sector As a Source of Affordable Housing
●
●
●
●
279
Imposing rent and tenure security constraints in a country like the UK would have particularly adverse effects because of the extent of general housing shortages. Institutional investment in rental housing is limited internationally, because of the nature of the investment and the attractiveness of the sector to small-scale investors. Institutional investment in the PRS is typically low and as a share of all rental properties is likely to remain so, without substantial, unjustifiable subsidies. The social costs of a lack of institutional investment in the PRS are small or non-existent.
The PRS works comparatively well at what it does. This statement may seem unnecessary to make but the UK has only had a free rental market for a relatively short period of time and institutional arrangements have only become embedded over the past decade. There is also a continuing chorus of criticisms and requests for greater regulation, some of which have been successful. As no housing system is perfect, the most appropriate policy question should be to ask whether alternatives to a low regulation stance in relation to the PRS would be cost effectively better. For example, Scotland introduced landlord registration some years ago. Has this raised the quality and quantity of the rental offer there compared to England? Similarly, has HMO (house in multiple occupation) registration legislation in England, introduced several years ago, with extensive and costly compliance requirements and local authority monitoring, actually achieved its aims? How have new local powers in England to limit high concentrations of shared rented homes affected local housing markets? No worthwhile research has been undertaken on these questions but scepticism regarding improvement seems most appropriate. Regulations, and the threat of more regulation, put off investors. This is generally due to the costs of compliance. Rogues ignore rules, so it has to be demonstrated that the total costs of compliance by all landlords are outweighed by improvement in the quality offer by the few. The paradox may arise where regulations deter good-quality investors, and the resultant accommodation shortages generate substantial financial incentives for those prepared to flout the rules. In those situations, the outcome of legislation would then be the opposite of what was intended with the quality offer falling sharply. Feedback effects are important and have been the basis of the criticisms made here of enforced security of tenure and rent controls. Renting cannot resolve housing shortages. These are determined by the overall relation between demand and supply, which currently sees chronic shortages in the UK across the whole range of the market. If rents rise in the
Jones_c13.indd 279
2/7/2012 10:25:49 AM
280
Challenges of the Housing Economy
PRS, as they are likely to, the cause will be these growing shortages. Simultaneously, some tenants in order to economise in the face of rising rents will want to crowd up more and take lower quality accommodation. So, the prognosis for housing conditions without improved housing supply is bleak. In addition, subsidising specific types of housing can crowd out other options, including free market renting. The policy prescriptions with respect to housing are, therefore, as complex as ever and the watchwords are please look out for spillover and feedback effects in other parts of the housing system when promoting one policy measure. For example, intermediate affordable housing schemes may dent private rented supply. In policy terms, the case for supporting additional large-scale investor involvement in the PRS seems scant. Of course, any new supply of rental property is to be welcomed, but with regard to private investors targeting the general rental market, there seems little justification for such a policy intervention. Solvable market failures or potentially merit-worthy social redistribution do not exist. It may be the case that large investors begin to find residential investment more attractive but, if this is the case, any expansion of large-scale investors would be at the expense of provision by smaller-scale ones, i.e. there would be a supply crowding out effect. Whether this would be a social benefit or not would depend on competition and the reaction of the targeted groups of tenants. The outcome would not be the replacement of the bottom quality end of the PRS with new flats but rather a question of whether relatively well off young mobile tenants would prefer the current rental offer they face to a probably somewhat more expensive package. Their decisions will be based on trade-offs regarding space standards, amenities and location for the time periods before most of them become homeowners. To date, most have opted for rental accommodation in the existing stock, provided by a small landlord. It will be interesting to see whether new providers can persuade enough of them to switch to their schemes. Yet it is obviously best to let those potential tenants decide themselves rather than to distort the relative costs of the options through tax breaks and other forms of subsidy. This chapter has focused on policy details in the UK because, as was noted at the beginning, policy details and the contexts in which they exist matter much when evaluating policies in this area. However, the general discussion was about the PRS as a concept. Though simple conclusions are often difficult to make in this area, the general drift of the analysis is that markets work well in this area and that regulation tends not to. Moreover, those that often support extensive controls typically argue that the issues are related to a two-way dispute between the interests of the tenant and those of the landlord. Regulation in that viewpoint alters the distribution of gains between the two. However, that perspective is too narrow. Investors (landlords) are
Jones_c13.indd 280
2/7/2012 10:25:49 AM
The Private Rented Sector As a Source of Affordable Housing
281
motivated by returns. Whatever the rules, they will get the opportunity cost returns from their housing investments. What alters are the volume of those investments; who lives in rental accommodation and their housing futures; and, finally, the current and future circumstances of others without access to such regulated accommodation. Therefore, it is a distributional issue across the population and one of the themes of this chapter has been that typically it is not the weakest that gain from tight housing regulation – quite the opposite.
Notes 1. 2. 3. 4. 5. 6. 7.
8. 9. 10.
Jones_c13.indd 281
Department of Work and Pensions Family Resources Survey 2007–8. English Housing Survey 2007–8. English Housing Survey 2007–8. Department for Communities and Local Government. HM Treasury Government response to the consultation on investment in the PRS, Sept 2010. Household forecasts based on 2008 projections by Department for Communities and Local Government. Income and substitution effects are not well-calibrated in the PRS. Demand functions for tenants do not exist in the same way as for owner occupiers, because of data limitations, so forecasting exercises about future renter demand cannot be very precise. Council of Mortgage Lenders data. English House Condition Survey 2006 Private Landlords Survey, Department for Communities and Local Government. There are no detailed studies of private landlord cost functions.
2/7/2012 10:25:49 AM
14 Conclusions: The Challenges Ahead Colin Jones
The credit crunch was the first global housing market shock, and in its wake there has been a growing international awareness of shared issues about the housing economy in terms of its role in the macroeconomy, and its distributional impacts on the population. As the book demonstrates, its impact is far from over. In many parts of the world the shock waves continue to impinge on housing markets through the constraints on mortgage finance, on the capacity of households to buy a home or move house, and hence on house prices and the viability of housebuilding. Nevertheless the ramifications have unwound long enough to reveal significant issues facing households and policy makers. The subsequent analysis has not just highlighted these short-term consequences but also demonstrated some common longterm themes or questions about the direction housing markets have been flowing.
The credit crunch For many countries the exogenous shock caused by the international banking crisis following the subprime problems brought to an end abruptly a housing market boom. Some countries such as the USA and Denmark were already experiencing a downturn or at least the seeds of one. For others
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_c14.indd 282
2/7/2012 10:25:36 AM
Conclusions: The Challenges Ahead
283
such as Australia and China it was only a modest and transient dampener on a continuing house price boom. Other countries such as Japan and Germany where real house prices have been in long-term decline the picture is more inconsistent with impacts in specific sectors. The credit crunch too revealed key dimensions to the housing market. First, the extent, influence and importance of global capital markets since the removal of exchange controls in different countries from the mid 1970s. This globalisation enabled banking to gradually become more and more sophisticated. There was a growth of wholesale funding and the expansion of mortgage finance especially from the 1990s in many countries. The associated worldwide relaxation of constraints on banks allowed them to move into new countries and the increasingly unfettered competition brought an internationalisation of mortgage products – often changing traditional national approaches to financing house purchase. It also enabled for the first time the use of housing equity release to pay off unsecured lending or fund consumption. The international banking liquidity crisis also revealed the frailty of the system. The global capital markets underestimated the potential for contagion through the interconnectedness of the financial system. It meant that banks were susceptible to risk from weakness in any one part. Banks created a complex web of international and interbank sales and purchases of bonds, many secured by residential mortgages. On the one hand the bonds provided an income to the bank, on the other the issuer was a recognised international financial institution so they could be sold quickly and easily. Ownership of these bonds was seen as liquid assets, and indeed there was an established market track record. In this way banks could address any liquidity problem that might occur. The fallacy of this proposition can be seen by the experience of British banks in the credit crunch. Prior to deregulation in 1981 banks in the UK were legally required to hold 12.5 per cent of their assets as designated liquid assets known as ‘reserve assets’, primarily short-term government bonds. Notwithstanding the relaxation of these regulations, holding liquidity to ensure credibility is a basic tenet of banking. However, by the mid noughties the cash holdings of British banks were at all time low, but there was a delusion of liquidity. Banks thought ownership of these mortgage-backed bonds issued with AAA ratings could easily be translated into cash. Unfortunately this liquidity proved a mirage as the value of these bonds was found to be no longer realisable as the market in these bonds suddenly froze and remained closed without any clear light at the end of the tunnel (Congdon, 2009). In addition the subsequent need to recycle maturing bonds (with no buyers) and the Bank of England’s refusal to offer short-term cash created, first, a liquidity problem for individual banks and then the collapse of the banking system. Banks were required to repay the capital on maturity of these bonds and in the absence of the ability to issue new ones sought
Jones_c14.indd 283
2/7/2012 10:25:36 AM
284
Challenges of the Housing Economy
new capital by the issue of shares. Eventually it was the UK government who bought up this equity capital and effectively nationalised some of the largest banks in the country. The consequences for the housing market (and for the economy) were a dramatic reversal of the availability of loan finance as the position turned from feast to famine. The subsequent impacts on many housing markets around the world demonstrated the power and importance of mortgage finance. In the lead-up to the credit crunch its influence on the housing market in western economies had been seen as supporting the house price booms, even exacerbating upward trends by increasing the personal purchasing power of individual households. This was very much an enabling role, one of four pillars, together with economic growth, low interest rates and increasing numbers of households, underpinning the boom. The short-term role of mortgage finance was arguably forgotten. While, historically there are examples of episodes when mortgage rationing has had a severe downward pressure on house prices these predate the deregulation of the banks, for example the UK in the mid 1970s. Another reason is that the exact amount of mortgage finance available is not easy to measure and therefore difficult/impossible to incorporate in forecasting models of the housing market. It is implicitly assumed in these models that the level of mortgage finance is demand determined. Some thought in the future will need to address this issue in economic models by incorporating potential independent variables such as a rationing on/off binary dummy or average loan-to-value ratios. The credit crunch and the mortgage lending drought represent the immediate context for this book together with the housing market trends over the past two decades. The various contributions have revealed that despite growing economic globalisation, and the development of world capital markets that has supported the expansion of wholesale funding and similar mortgage products available worldwide, each country has a different story. These differences can be seen in terms of the performance of the economy, historical house price cycles, the institutional and public policy context reflected in different taxes, local town planning, etc. and in the response to the international financial crisis.
Commonalities and challenges Synthesis of the different chapters draws out a number of commonalities and challenges: 1. the role of speculation 2. the consequence of the credit crunch for the housing market: a fall in house prices across many countries – although price levels have now stabilised
Jones_c14.indd 284
2/7/2012 10:25:36 AM
Conclusions: The Challenges Ahead
285
3. the long-term accessibility of homeownership and its affordability for young adults 4. the price adjustment following the financial crisis in housing markets – this has been relatively weak when contrasted with housing construction activity that has collapsed 5. the post credit crunch world with tight public sector finances, which means that new models of funding are required for social housing, blurring the differences with private rented accommodation 6. private landlordism, which had flourished in the housing boom – although there were doubts about its future growth These are now considered in more detail by drawing on the detail of the different chapters.
Speculation in the housing market The experience of Las Vegas demonstrated the potential significance of speculation in the housing market and its role in the housing boom. Las Vegas during the noughties was obviously an extreme example where a rapidly expanding city created one of the fastest growing local housing markets in the USA, within a wider national housing boom. As chapter 3 reports, these conditions were ripe for speculation and the specifics of the city and its housing market provide unique insights. During the frenzy of the peak house price inflation at the beginning of the decade the speculative component of market activity became a significant proportion of total transactions. More than 40 per cent of all home transactions were a part of this speculative ‘flipping’, either buying or selling at this point in the cycle. Flipping activity as set out in chapter 3 relates specifically to the USA but clearly speculation in different forms is ubiquitous around the world. The Chinese market discussed in chapter 5, it is argued, is driven partly by these speculative motives pushing house prices to multiples of average income well beyond the pockets of most households. One example is the growth of highly geared buy-to-let landlords buying up small flats in the provincial city centres of Britain during the last housing boom. This segment of the housing market was the fastest growing part of the UK housing market over the noughties, encouraged and supported by planning policies (see chapter 10). A major motivation was the anticipated capital growth rather than the rental income. As prices of these flats accelerated, the submarket saw some buyers deliberately choosing to buy, to leave their properties empty, at least in the short term to benefit from the expected capital gains (Cobbold, 2007). It is interesting to note that chapter 3 finds that the impact of speculative motives within the market varies significantly within the house price cycle and only becomes substantial over a relatively brief period leading up to its
Jones_c14.indd 285
2/7/2012 10:25:36 AM
286
Challenges of the Housing Economy
peak. However, markets particularly subject to speculation appear to be the most volatile. When the market turned in Las Vegas and when the supply of housing outstripped demand, prices fell more than any other city in the USA, almost 60 per cent from peak to trough (chapter 2). Similarly in the UK, city centre flats were the first to experience the downturn and probably the most badly affected housing submarket in the bust. In the downturn, markets particularly affected by speculative investors are driven down by investors selling, sometimes forced to sell for example because of remortgaging problems in the UK or as in the case of Las Vegas by foreclosures. Foreclosures following the credit crunch in our case study countries have only been extensive in the USA, totalling almost 832,000 in 2010 (chapter 2). This appears to reflect the more extensive issue of subprime mortgages (and the rise of unemployment in the recession). This phenomenon contributed significantly to falling prices in the downturn in the USA. In other case studies, countries where the housing market has been adversely affected by the fallout from the financial crisis governments have intervened to avert foreclosures. This has been achieved by constraining banks’ ability to move to repossession by introducing forced time lags in the process. It is impossible to assess the impact on the fall in house prices of foreclosures. Certainly in the UK and Denmark, the market appears to have been less adversely affected than the recession at the end of the 1980s/beginning of the 1990s when there was a large rise in foreclosures. However, other intervening factors may have been at work too; for example, interest rates are much lower than during the previous recession.
Credit crunch and house prices Chapter 1 provides a broad overview of the differential international experience of housing markets to the exogenous shock of the financial crisis based on OECD statistics. The individual countries examined in this book reveal more detail to flesh out this broad aggregate statistical analysis. Referring to the USA, where the initial housing market problems emerged with subprime lending, chapter 2 stresses the widely diverse experiences of localities across the USA. In some of the main cities prices fell by more than a third, even as much as 60 per cent, from peak to trough. Prices in general fell by less in the smaller cities and there is a sense of a rolling housing recession probably through a process of spatial arbitrage. The stable housing markets in Texas stand aside from the boom and bust experienced in the rest of the USA despite strong consistent economic growth, possibly because of relaxed planning constraints. Despite the USA leading into the global housing market abyss the housing crisis is still not over, albeit in recovery after almost five years. In Denmark and the UK the initial dip in the housing markets lasted two years and around 15 months, respectively, and prices fell around 20 per cent.
Jones_c14.indd 286
2/7/2012 10:25:36 AM
Conclusions: The Challenges Ahead
287
In both cases prices bottomed out at the beginning of 2009. However, the subsequent recovery has in each case been muted, clouded by potential storms on the horizon caused by the fiscal crisis and anticipated deflationary macroeconomic policies. These expectations have sapped households’ confidence, already concerned with the scale of personal debt, and prices fell again marginally at the end of 2009 before stabilising. Transactions are at a very low level reflecting a range of factors including the reluctance of householders to sell when prices are low (only partly because of negative equity considerations) and the difficulties of finding a buyer. Spanish house prices (and transactions) have fallen away in similar fashion and to a broadly equivalent degree if over a longer (unfinished) period. The house price booms in Australia and China were little affected by the credit crunch. After a short-lived downturn in Australia house prices returned to double-digit inflation supported by a 4 per cent lowering of interest rates, the government’s response to the international crisis. The Chinese market has limited data but prices have risen continuously by several fold in the past decade, especially in the main cities. There was a modest pause in this upward trend in 2007 but the continuing boom has made the government so concerned with the scale of the house price inflation that they have raised interest rates and minimum loan-to-value ratios requirements. Two of the case studies did not experience the international house price boom but had contrasting fortunes following the banking collapse. The German housing market was hardly touched by the credit crunch because the mortgagors had traditionally been subject to low loan-to-value ratios, and credit availability was not further restricted. However, house prices while stable had been on a downward trajectory for some time prior to the financial crisis. House prices in Japan have also followed a long-term negative trend over the past 20 years. Although house prices had begun to stabilise in the latter half of the noughties they fell by 10 per cent in 2009 following the financial crisis. These market outcomes to the international financial problems are not, of course, totally representative but are simply examples. The differences partly reflect the strength of the impact of banking collapses – all those countries worst affected suffered bank failures. It is important to remember that the general fall in interest rates around the world moderated the fall in house prices in those countries worst affected but also contributed to house price inflation of those countries with limited exposure to the banking crisis.
Access to homeownership The tightening of mortgage credit following the banking liquidity debacle has in the short term had most effect on would-be first-time purchasers, some of whom have had their plans at best delayed and potentially thwarted.
Jones_c14.indd 287
2/7/2012 10:25:36 AM
288
Challenges of the Housing Economy
But it has also drawn even more attention to the longstanding issue of homeownership rates falling among young adults. A starting point is the pricing out of young households from homeownership by the increase in real house prices. As discussed above, the long rise in real house prices in many countries is associated with the increasing availability of credit and the flexibility of new mortgage products. Denmark is a useful example. Prior to 1996, the traditional fixed interest rate repayment mortgage was the norm but in that year variable rate mortgages began to be offered. In the noughties they eventually dominated the market with more than two-thirds of all outstanding Danish mortgages in 2010. Similarly in the USA the standard 30-year mortgage with a fixed interest rate and a 20 per cent down payment introduced in the 1930s was usurped in the 1990s by a variety of mortgage instruments with adjustable mortgage rates and lower (even zero) down payments. This introduction of these new products, as noted earlier, was in part the consequence of the growth of international banks. It can also be argued that these developments were designed to extend the market and improve the access to and affordability of owner-occupation by the introduction of variable interest rate/interest-only mortgages and the reduction of entry barriers by lower deposits. For this reason it was welcomed and even encouraged by policy makers around the world who generally viewed increasing the owner occupation rate as a positive step. The problem is that the long-term outcomes of this process around the world have had the reverse effect. In many of our case studies the level of owner occupation was under threat as young adults struggle with the cost of homeownership given the rise in real prices before the credit crunch. Owner occupation levels have fallen in the USA and England, and in other countries it is stable. Chapter 7 finds that while the level of homeownership in Denmark has been stable for decades at just under a half of all households this reflects an ageing population. In absolute and proportional terms the number of Danish owner occupiers under 30 years of age has fallen significantly and there has also been a modest fall in the 30–39 years age group over the 20 years, 1987–2007. The story is the same in Australia where the level of homeownership has not changed for 40 years. Meanwhile real house price increases over this period, in conjunction with economic and social trends, have created affordability constraints for first-home buyers. The result is that homeownership rates among young first-home buyers have also declined. The beneficiaries of flexible mortgage finance have not been the young but the older generations who have benefitted from the rise in house prices, first by simply being existing owners, but also through the high debt gearing available in a rising market. The innovations of equity release and remortgaging also offered positive opportunities. One explanation is that this has
Jones_c14.indd 288
2/7/2012 10:25:36 AM
Conclusions: The Challenges Ahead
289
occurred through the interaction of rising demand fuelled by the readily available mortgage finance (for some) combined with inelastic supply. This certainly would fit with the arguments set out in chapter 10 for the UK. However, this cannot be a complete explanation of the falling rates of homeownership in younger generations. Even in Germany, which has excess supply, falling prices and long-term tight credit conditions, the level of homeownership has been dropping for those aged under 40 years of age. In Japan, falling house prices and affordability are related to rising unemployment, and the risk of the latter that has created a barrier to homeownership for young adults.
Importance of demographic trends A number of the contributions emphasise the importance of demographic trends in shaping and driving housing markets in terms of age structure and population change. Irrespective of macroeconomic conditions the constant dynamic of the changing population demographic means changing housing requirements and provides the motor for the housing market. It is instructive to consider some examples from the contributions. Chapter 9 draws attention to the importance of immigration and the age structure of the population on the Spanish market over the past two decades. These changes transformed the country’s housing market over this period. In contrast chapter 7 reports that sentiment in Germany is increasingly influenced by demographic expectations, suggesting a weakening housing demand with a shrinking and ageing population. Even though this is balanced by growing household numbers, these trends are seen as of increasing importance in influencing the market in the coming years. China has the highest long-term rate of population growth even after implementing the one child maximum policy. The current rate of population growth is 5 per cent per decade. The age distribution in China is overwhelmingly dominated by two age groups, 35–39 and 40–44 years, representing 19 per cent of the population. Despite the rapid growth in population, chapter 5 argues that such has been the supply response that there is a numerical surplus of housing. So while demographic trends are the longterm underpinning of the housing market the supply response is crucial.
Supply response to the credit crunch The evidence presented in chapters 7 and 10 on Spain and the UK respectively explains the dramatic collapse of new housebuilding following the financial crisis. The different experiences of these countries prior to the credit crunch emphasises the power of the mortgage famine. Spain with record housebuilding and the UK with low levels of housebuilding were
Jones_c14.indd 289
2/7/2012 10:25:36 AM
290
Challenges of the Housing Economy
both impacted in the same way and new supply numbers dived to all-time lows. It appears that the most severe market response to the banking problems has been not via prices but in terms of new building construction. This impact is partly because developers had internal financial problems stemming from an overhang of new housing unsold, with demand falling away, which had led to cash flow problems, and the high asset values of land bought in the boom had to be written down in their financial accounts. In addition, banks were no longer able or willing in the short term to fund new development. The main policy question is not the cutback in supply of new housing for sale in the short-term as there were few buyers, and falling prices increased affordability. But even here there are difficulties in the UK as it had sought to link its affordable/social housing to the output of the private sector via planning agreements. The collapse of new housebuilding for sale has had an equivalent effect on non-market housing at a time when demand is rising with the weak macroeconomic performance. Four years on the continuing very low levels of housebuilding appear set for the medium term in both Spain and the UK and will eventually proffer development constraints.
The future The impact of the credit crunch directly impinged on the owner-occupied sector through falling prices and new housing completions, but its policy ramifications are much wider. As policy makers turned to the challenge to ensure that these problems should not happen again, discussions extended not just to the funding and taxation of homeownership but also to the balance of renting and owning in the housing system. Chapter 2 sets out these issues for the USA but it is clear there are no easy answers when fiscal budgets and economies are in a weak and parlous state. One of the key problems is that phasing out the tax advantages of homeownership to moderate speculation in the future is difficult at a time when household budgets are strained. The best time to do this is when the property market is on an upward curve: for example, the UK phased out tax relief on mortgage interest for owner occupiers in the latter half of the 1990s. However, the bad news was that this change seemed to have very little ameliorating effect of the house boom that accelerated shortly afterwards. The Chinese solution to the affordability problem caused by the housing boom is to focus on a large scale programme of affordable housing, but in most countries this is not an option. The medium-term fiscal austerity that has spread across much of the world following the financial crisis has occurred at a time when there was already a debate about the role of the state in the provision of social/affordable housing. Chapter 11 reports that
Jones_c14.indd 290
2/7/2012 10:25:36 AM
Conclusions: The Challenges Ahead
291
there is a diverse approach to this issue across Europe, but there is a recent trend toward public–private partnership. Fiscal cutbacks have strengthened the arguments for targeting public resources and a more market-oriented approach to the provision of social/affordable housing. In the UK, the need to reduce public spending at a time when personal housing difficulties have been magnified has been the catalyst for rethinking the financing of social/affordable housing. Chapter 11 presents an analysis of how the problem has been formulated by the Scottish government as it seeks value for money within its dwindling public purse. It reveals a catalogue of issues that require to be resolved as policy seeks to draw more private sector resources into the provision of such housing with the associated uncertainties of success. These encompass the willingness and capacity of the private sector, the skills base of social landlords and ultimately how the new ‘products’ will be received in terms of demand. There are also issues about whether the introduction of rental housing at intermediate rent levels (part publicly funded) above those currently applicable to affordable housing will attract households freeing up ‘space’ for lower-income households. In other words besides many management practicalities there are both financial and housing filtering models that need to work for these innovative approaches to work. In a sense the ‘ideal’ solution from the public sector perspective in countries seeking public expenditure cuts is for the private rented sector (PRS) to house low-income households on a long-term basis without subsidy (or at least only income-related subsidy). Chapter 13 examines the potential expansion of the PRS in the UK where the sector has grown dramatically over the past decade, partly in response to the affordability problems caused by rising real house prices. The findings of this chapter, which draws also on international experience, is that the PRS works comparatively well at what it does but that there are limits to what the sector can achieve, and it is no solution to an absolute housing shortage. It is wary about its potential for expansion from its present role and is doubtful about policy developments designed to make it more attractive for long-term tenants. The potential for institutional investment is seen as limited in a sector that is dominated by individual investors. Overall the watchword of chapter 13 is caution, with concerns about whether public subsidy of the sector can be justified, given the uncertainties of the outcomes and the potential spillover effects of change on other parts of the housing system. There are therefore many challenges but also question marks about the future of housing markets. They are bound inextricably to the state of national economies and this adds to the uncertainties at the time of writing with major concerns about the future of currencies, the continuing weak economic growth in many countries in response to the credit crunch and the possibility of severe fiscal rectitude for the foreseeable future. Across the
Jones_c14.indd 291
2/7/2012 10:25:37 AM
292
Challenges of the Housing Economy
world a major casualty of the economic downturn has been the economic prospects of young adults with high youth unemployment prevalent. This leads on to the potential for long-term market decline along the lines of the lost two decades of Japan with low housebuilding and long-term falling prices. Chapter 5 argues that the most effective way forward is through the reflation of the economies. Housing policies, for example those designed to stimulate homeownership, are ineffective in a depressed economy when the most likely way to encourage younger tenants to own their houses is remove the risks stemming from the labour market. Curbing deflation, it argues, is the most successful housing policy.
Concluding remarks The credit crunch had a severe impact on many housing markets around the world emphasising the interrelationships between national economies and the importance of global capital markets. Freely available mortgage finance in the housing boom was made possible through wholesale funding on a global scale. Globalisation is here to stay, as the genie cannot be put back in the bottle, and it is likely to mean a continuing convergence of mortgage finance products across countries. But the prospects for the availability of mortgage finance are less certain, as greater constraints on banks’ liquidity and equity capital are introduced as a result of Basel III (see glossary). Looking back at the ‘good times’ of the past two decades, freely flowing credit facilitated by the deregulation of the banks did not produce the expected results. Part of the reason in many countries was the weak supply response. The growth of flexible mortgage products and the availability of mortgage finance have been to the detriment of young generations. The almost universal rise in real house prices simply priced new households out of the market, while the older generations reaped the rewards. For the future, a key issue is that of homeownership. Owner occupation has reached an upper bound in many countries and is actually falling away in the USA and possibly England. In other countries, homeownership levels are stable only because their population is ageing and the elderly tend to be owner occupiers. Everywhere it seems homeownership rates of young households are falling. There is a clear issue of intergenerational equity. But it is also a wider problem of insufficient housing that imposes constraints on the lives of low-income households too, if not more so, as they are squeezed by the middle income groups who can outbid them. The immediate problem is that the international financial crisis has made the position worse for young adults. In many western countries the increased deposits required by banks in the post credit crunch era are resulting in greater barriers to homeownership for young people. New housing supply
Jones_c14.indd 292
2/7/2012 10:25:37 AM
Conclusions: The Challenges Ahead
293
has fallen to record lows. It is unclear to what extent this is a short-term cyclical problem and, if so, how long it will last. However, the evidence from the book is that there are fundamental difficulties for housing markets. Given the scale of the difficulties it is hoped that the policy response could become a pivotal point in the housing market like the rethink of Keynesianism in the 1970s. A reappraisal of the conventional wisdom is required if the aspirations of the younger generations and the poor are to be met.
Jones_c14.indd 293
2/7/2012 10:25:37 AM
References Abelson P, Joyeux R, Milunovich G and Chung D (2005) Explaining house prices in Australia: 1970 to 2003, Economic Record, 81(255), S96–S103. Acharya V V, Cooley T F, Richardson M P and Walter I (2010) Regulating Wall Street: The Dodd-Frank Act and the New Architecture of Global Finance. Wiley, New York. Acharya V V, Richardson M P, van Nieuwerburgh S and White L J (2011) Guaranteed to Fail: Fannie Mae, Freddie Mac and the Debacle of Mortgage Finance, Princeton University Press, Princeton, NJ. Adams Z and Fuss R (2010) Macroeconomic determinants of international housing markets, Journal of Housing Economics, 19, 1, 38–50. Ahearne A G, Ammer J, Doyle B M, Kole L S and Martin R F (2005) Monetary Policy and House Prices: A Cross-Country Study, Central Bank of Chile Working Paper No. 344. Alliance and Leicester (2007) Parents fork out more than £21K for children to fly the nest, http://www.alliance-leicester-group.co.uk/html/ media/non-indexed/ release. asp? txtTable= PressReleases&txtCode=PR2208071. Accessed March 2008. Almeida H, Murillo C and Crocker L (2006) The financial accelerator: Evidence from international housing markets, Review of Finance, 10, 3, 321–352. Amromin G, Huang J, Sialm C, and Zhong E (2011) Complex Mortgages, mimeo, http://ssrn. com/abstract=1718485. Accessed June 2011. Anderlini J (2010) China revives property tax to avert bubble, Financial Times, Dec.10, http:// www.ft.com/intl/cms/s/0/dfd473ec-e5a7-11de-b5d7-00144feab49a.html#axzz1Oh7hc39p. Accessed June 2011. André C (2010) A Bird’s Eye View of OECD Housing Markets, OECD Economics Department Working Papers, No. 746, OECD, Paris. Andrew M (2006) Housing Tenure Choices by the Young, CML Housing Finance, 07, 1–12. Andrew M and Meen G (2003) Housing transactions and the changing decisions of young households in Britain: The microeconomic evidence, Real Estate Economics, 31, 1, 117–138. Aoki K, Proudman J and Vlieghe G (2001) Why house prices matter, Bank of England Quarterly Bulletin: Winter 2001. Arestis P and Karakitsos E (2010) Subprime Mortgage Market and Current Financial Crisis, chapter in P Arestis, P Mooslechner and K Wagner (eds), Housing Market Challenges in Europe and the United States, Palgrave Macmillan, Basingstoke. Arnott R (1987) Economic theory and housing, chapter in E S Mills, Handbook of Regional and Urban Economics, vol. II, chapter 24, 959–988, North-Holland, Amsterdam. Arnott R (1995) Time for revisionism on rent control? Journal of Economic Perspectives 9, 1, 99–120. Australian Bureau of Statistics (ABS) (2010) Year Book 2009–10, ABS Cat. No. 1301.0. http:// abs.gov.au. Accessed November 2010. Australian Government (2002) (IGR1) Intergenerational Report 2002/03, 2002/03 Budget Paper No. 5, Canberra, www.treasury.gov.au/igr/. Accessed May 2002. Australian Government (2007) (IGR2) Intergenerational Report 2007, Canberra. www.treasury. gov.au/igr/. Accessed July 2010.
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_bref.indd 294
2/7/2012 10:18:53 AM
References
295
Australian Government (2010) (IGR3) Australia to 2050: future challenges, Intergenerational Report 2010. http://www.treasury.gov.au/igr/. Accessed February 2010. Aziz J and Cui L (2007) Explaining China ‘s Low Consumption: The Neglected Role of Household Income, IMF Working Paper WP/07/181, http://www.scribd.com/doc/500177/ Chinese-Consumption-Household-Income-IMF-Working-Paper. Ball M (1999) Chasing a snail: innovation and housebuilding firms’ strategies, Housing Studies, 14, 9–22. Ball M (2006) Buy to Let. The revolution 10 years on. Assessment and Prospects, Association of Residential Letting Agents, Amersham, Buckinghamshire. Ball M (2007) Large-scale Investor Opportunities in Residential Property: an Overview. Research Findings. Investment Property Forum, London. Ball M (2008) The Modern Housing Market – Origins and Prospects, National Federation of Property Professionals, Warwick, Warwickshire. Ball M (2009) European Housing Review 2009, RICS, London Ball M (2010) European Housing Review 2010, RICS, London. Ball M (2011) RICS 2011 European Housing Review, RICS, London. http://www.rics.org/site/ download_feed.aspx?fileID=8882&fileExtension=PDF Accessed June 2011. Ball M and Sunderland D (2001) An Economic History of London 1800–1914, Routledge, London. Ball M and Glascock J (2005) Property Investment Funds for the UK: Potential impact on the private rental market, Council of Mortgage Lenders, London. Ball M, Blanchette R, Anupam N and Wyatt P (2011) Housing markets and independence in older age. Expanding the opportunities, University of Reading, Reading. Bank of England (2010) Trends in Lending June 2010, Bank of England, London. Barlow J (1999) From craft production to mass customisation. Innovation requirements for the UK housebuilding industry, Housing Studies, 14, 23–42. Barker K (2003) Review of Housing supply – Interim report: Analysis. HMSO, London. http:// www.barkerreview.org.uk/. Accessed June 2008. Barker K (2004) Review of Housing supply: Delivering Stability – Securing Our Future Needs, final report – recommendations. HMSO, London. Basu K and Emerson P M (2000) The economics of tenancy rent control The Economic Journal 110(466), 939–962. Baudewyns D (2007) An analysis of the housing market in Greater Brussels, chapter in A Andersson, L Pettersson and U Stromquist (eds) European Metropolitan Housing Markets, Springer. Baum S and Wulff M (2003) Housing Aspirations of Australian Households, AHURI Final Report. Australian Housing and Urban Research Institute, Melbourne. http://www.ahuri. edu.au. Accessed March 2003. Benito A (2006) The down-payment constraint and UK housing market: Does the theory fit the facts? Journal of Housing Economics 15, 1, 1–20. Bernanke B, Gertler M and Gilchrist S (1999), The financial accelerator in a quantitative business cycle framework, chapter in J Taylor and M Woodford (eds), The Handbook of Macroeconomics, Vol. 1, North Holland, Amsterdam. Blackaby B (2000) Understanding Local Housing Markets: Their Role in Local Housing Strategies, CIH, Coventry and CML, London. Blackley D M (1999) The long-run elasticity of new housing supply in the United States: Empirical evidence for 1950 to 1994, Journal of Real Estate Finance and Economics, 18, 1, 25–42. Bloxham P, Kent C and Robson M (2010) Asset Prices,Credit Growth, Monetary and other Policies: an Australian case study, Research Discussion Paper, 2010–06, Reserve Bank of Australia. At http://www.rba.gov.au. Accessed October 2010.
Jones_bref.indd 295
2/7/2012 10:18:53 AM
296
References
Bodman P and Crosby M (2004) Can macroeconomic factors explain high house prices in Australia?, Australian Property Journal, 38, 174–9. Bogdon A S and Can A (1997) Indicators of local housing affordability: comparative and spatial approaches, Real Estate Economics, 25, 1, 43–80. Bolton R E and Westlund H (2003) Local social capital and entrepreneurship, Small Business Economics, 21, 2, 77–113. Bostic R, Gabriel S and Painter G (2009) Housing wealth, financial wealth, and consumption: New evidence from micro data, Regional Science and Urban Economics, 39, 79–89. Bourassa S and Hendershott P (1995) Australian capital city real house prices, 1979–1993, Australian Economic Review 95, 3, 16–26. Bourassa S C and Hoesli M (2008) Why do the Swiss rent? The Journal of Real Estate Finance and Economics 40, 3, 286–309. Bourassa S, Greig A and Troy P (1995) The limits of housing policy: home ownership in Australia, Housing Studies, 10, 1, 83–104. Bourassa S, Hoesli M and Scognamiglio D (2009) Housing finance, prices, and tenure in Switzerland, Swiss Finance Institute Research Paper No. 09–16. Bramley G (1993) Land-use planning and the housing market in Britain—The impact on housebuilding and house prices, Environment and Planning A, 25, 1021–1052. Bramley G (1997) Housing policy: a case of terminal decline? Policy and Politics, 25, 387–407. Bramley G (2003) Planning regulation and housing supply in a market system, chapter in A O’Sullivan and K Gibb, Housing Economics and Public Policy, Blackwell Publishing, Oxford. Bramley G (2007) The sudden discovery of housing supply as a key policy challenge, Housing Studies, 22, 2, 221–242. Bramley G, Karley N and Watkins D (2006) Local Housing Need and Affordability Model for Scotland – Update (2005 Based). Report 72. Communities Scotland, Edinburgh. Bramley G, Pawson H, Morgan J, Wilcox S and Williams P (2010) A Study into the Capacity of Registered Social Landlords and Local Authorities to Build Housing across Scotland. Scottish Government Social Research, Edinburgh. Brash D (2009) How Texas averted the great recession. http://www.houstongrowth.org/files/ HowTexasAvertedRecessionfinal.pdf. Brindley T, Rydin Y and Stoker G (1996) Remaking Planning, Routledge, London. Bureau of Economic Analysis (2011) http://www.bea.gov/regional/reis/drill.cfm. Campbell H, Ellis H, Henneberry J and Gladwell C (2000) Planning obligations, planning practice, and land-use outcomes, Environment and Planning B: Planning and Design, 27, 759–775. Calman K (chair) (2009) Serving Scotland Better: Scotland and the United Kingdom in the 21st Century. Commission on Scottish Devolution Final Report, UK Government, London. Campbell J Y and Cocco J F (2007) How do house prices affect consumption? Evidence from micro data, Journal of Monetary Economics, 54, 591–621. Carruthers J I and Mulligan G F (forthcoming) The ‘plane of living’ and the recent evolution of housing values in the United States, Journal of Economic Geography, forthcoming. Case K E and Quigley J M (2008) How housing booms unwind: Income effects, wealth effects, and feedbacks through financial markets, European Journal of Housing Policy, 8, 2, 161–180. Case K E and Quigley J M (2010) How housing busts end: Home prices, user cost, and rigidities during down cycles, chapter in S J Smith and B A Searle (eds) (2010) The Blackwell Companion to the Economics of Housing: The Housing Wealth of Nations. Blackwell Publishing, Oxford, 459–480. Case K E, Quigley J M and Shiller R J (2005) Comparing wealth effects: The stock market versus the housing market, Advances in Macroeconomics, 5, 1, Article 1.
Jones_bref.indd 296
2/7/2012 10:18:53 AM
References
297
CECODHAS (2011) Housing for All http://www.housingeurope.eu/ China Economic Review (2011) China scraps national property price index, Feb. 17, http://www. chinaeconomicreview.com/industry-focus/latest news/article/2011-02 17/China_scraps_ national_property_price_index.html. Accessed June 2011. Chow K K, Yiu M S, Leung C K Y and Tam D C (2008) Does the DiPasquale-Wheaton Model Explain the House Price Dynamics in China Cities?, HKIMR Working Paper No. 21/2008, Hong Kong Institute for Monetary Research, Hong Kong. Claessens S, Kose M and Terrones M (2008) What Happens during Recessions, Crunches and Busts? IMF Working Paper 08/274. Washington: International Monetary Fund. http://www. imf.org/external/pubs/ft/wp/2008/wp08274.pdf. Accessed January 2011. Clark R (2007) London property market: Flipping easy way to a profit, Financial Times, Sep. 24, http://www.telegraph.co.uk/property/3359394/London-property-market-Flipping-easy-wayto-a-profit.html. Accessed June 2011. CLG, Department for Communities and Local Government (2006a) Delivering Affordable Housing, CLG, London. CLG, Department for Communities and Local Government (2006b) Planning Policy Statement 1, CLG, London. CLG, Department for Communities and Local Government (2007) Homes for the Future, More Affordable, More Sustainable, CLG, London. CLG, Department for Communities and Local Government (2009) Land Use Change Statistics (England) 2008 – provisional estimates, CLG, London. CLG, Department for Communities and Local Government (2010a) House Building: June Quarter 2010 (England), CLG, London. CLG, Department for Communities and Local Government (2010b) Evaluation of the Mixed Communities Demonstration Projects. Final Report, CLG, London. CLG, Department for Communities and Local Government (2011a) Draft National Planning Policy framework: impact assessment, CLG, London. CLG, Department for Communities and Local Government (2011b) New Homes Bonus: Final Scheme Design, CLG, London. CLG, Department for Communities and Local Government (2011c) Social Housing, CLG, London. http://www.communities.gov.uk/housing/socialhousing/ CML, Council of Mortgage Lenders (2011) Annual Report 2010, CML, London. COAG Reform Council (2010) National Affordable Housing Agreement: Baseline performance report for 2008–09, COAG Reform Council, Sydney. http://www.coag.gov.au/crc/docs/ national_affordable_housing_agreement_report_exec_summary.pdf. Accessed June 2010. Cobbold C (2007) What is the Extent of Buy to Leave in England?, CLG Housing Markets and Planning Expert Panel Report. http://www.henley.reading.ac.uk/nmsruntime/saveasdialog. asp?lID=20285&sID=77833. Accessed August 2008. Colliers (2011) Research and Forecast Report: Greater China Office and Residential Market, 1Q, Colliers International. http://www.colliersinternational.com/Content/Repositories/ Base/Markets/HongKong/English/Market_Report/PDFs/GC-Report-1Q-2011.pdf. Accessed June 2011. Congdon P (2009) The Northern Rock affair: An analysis of the ‘teaser rate’ strategy, chapter in F Bruni and D T Llewellyn (eds) The Failure of Northern Rock: A mult-dimensional case study, SUERF, Vienna. Coulson N E (1999) Housing inventory and completion, Journal of Real Estate Finance and Economics, 18, 1, 89–105. Countryside Alliance (2011) The Critical shortfall in Affordable Rural Housing in Britain, Countryside Alliance, London. http://www.countryside-alliance.org.uk/ca/file/Shortfall%20 in%20Affordable%20Rural%20Housing%20Report.pdf Accessed August, 2011.
Jones_bref.indd 297
2/7/2012 10:18:53 AM
298
References
Crook T and Kemp P (1999) Financial Institutions and Private Rented Housing, Joseph Rowntree Foundation, York. Crook A D H and Whitehead C M E (2002) Social housing and planning gain: Is this an appropriate way of providing affordable housing? Environment and Planning A, 34, 1259–1279. Crook A D H, Monk S, Rowley S and Whitehead C M E (2006) Planning gain and the supply of new affordable housing in England: Understanding the numbers. Town Planning Review, 77, 3, 353–373. Crook A D H, Henneberry J M, Rowley S, Smith R and Watkins C A (2008) Valuing Planning Obligations in England: update study of 2005-06, Department for Communities and Local Government, London. Crosby J (2008) Mortgage Finance: final report and recommendations, HM Treasury, London. Crosby N, Hughes C and Murdoch S (2006) Flexible property leasing and the small business tenant, Journal of Property Research, 23, 2, 163–88. Cullingworth J B (1997) British land-use planning: a failure to cope with change? Urban Studies, 34, 5–6, 945–960. Czischke D and Pittini A (2007) Housing Europe 2007: Review of Social, Co-operative and Public Housing in the 27 EU Member States CECODHAS, Brussels. Danmarks Nationalbank (2008) Monetary Review. 3rd Quarter 2008, Danmarks Nationalbank, Copenhagen. http://www.nationalbanken.dk/C1256BE9004F6416/side/Monetary_Review_ 2008_3_Quarter/$file/mon3_2008_web.pdf Accessed June 2011. Danmarks Nationalbank (2009a) Danish Government Borrowing and Debt 2008, Danmarks Nationalbank, Copenhagen. http://www.nationalbanken.dk/C1256BE9004F6416/side/Danish_ Government_Borrowing_and_Debt_2008/$file/SLOG-UK-2008_web.pdf Accessed June 2011. Danmarks Nationalbank (2009b) Financial Stability. 1st Half 2009, Danmarks Nationalbank, Copenhagen. http://www.nationalbanken.dk/C1256BE9004F6416/side/38AE3FF57766C053 C12575CB002F81F4/$file/Financial_stability_1st_2009Eweb.pdf Accessed June 2011. Danmarks Nationalbank (2010) Statistics on the balance sheets and flows of the MFI sector, Danmarks Nationalbank, Copenhagen. http://www.nationalbanken.dk/C1256BE9004F6416/ side/FiK20101222TT/$file/Fik20101222TT.pdf Accessed June 2011. Danmarks Nationalbank (2011) Financial Stability. 2011, May, Danmarks Nationalbank, Copenhagen. http://www.nationalbanken.dk/C1256BE9004F6416/side/Financial_stability_ 2011/$file/UK_2011_FS_web.pdf Accessed June 2011. Das J (2011) Will localism dirty our green and pleasant land?, Guardian, 1 August. http://www. guardian.co.uk/local-government-network/2011/aug/01/localism-green-land?INTCMP= SRCH Accessed August, 2011. Daunton M (1983) House and Home in the Victorian City: Working-Class Housing 1850–1914, Edward Arnold, London. Demary M (2009) The Link between Output, Inflation, Monetary Policy and Housing Price Dynamics, Munich Personal RePEc Archive, Paper No. 15978, http://mpra.ub.uni-muenchen. de/15978/ Demary M, Gans P, Meng R, Schmitz Veltin A, Voigtländer M and Westerheide P (2009): Wirtschaftsfaktor Immobilien. Die Immobilienmärkte aus gesamtwirtschaftlicher Perspektive, Hrsg.: Deutscher Verband für Wohnungswesen Städtebau und Raumordnung e.V./ Gesellschaft für Immobilienwirtschaftliche Forschung e.V. In: Zeitschrift für Immobilienökonomie, Sonderausgabe 2009, Berlin. Depken C A II, Hollans H, and Swidler S (2009) An empirical analysis of residential property flipping. Journal of Real Estate Finance and Economics, 39, 3, 248–263. Depken C A II, Hollans H, and Swidler S (2011) Flips, flops, and foreclosures: anatomy of a real estate bubble. Journal of Financial Economic Policy, 3, 1, 49–65.
Jones_bref.indd 298
2/7/2012 10:18:54 AM
References
299
Deutsch E, Tiwari P and Moriizumi Y (2006) The slowdown in the timing of housing purchases in Japan in the 1990s, Journal of Housing Economics, 15, 230–256. Diaz-Serrano L (2005) On the negative relationship between labour income uncertainty and homeownership: Risk-aversion vs. credit constraints,” Journal of Housing Economics, 14, 109–126. Dimson E, Marsh P, and Staunton M (2003) Global evidence on the equity risk premium, Journal of Applied Corporate Finance, 15, 4, 26–38. DiPasquale D (1999) Why don’t we know more about housing supply?, Journal of Real Estate Finance and Economics, 18, 9–23. DiPasquale D and Wheaton W C (1992) The markets for real estate assets and space: A conceptual framework, AREUEA Journal, 20, 1, 181–197. DiPasquale D and Wheaton W (1994) Housing market dynamics and the future of housing prices, Journal of Urban Economics, 35, 1–28. DiPasquale D and Wheaton W (1996) Urban Economics and Real Estate Markets, Englewood Cliffs, New Jersey. Disney R, Henley A and Stears G (2002) Housing costs, house price shocks and savings behaviour among older people, Regional Science and Urban Economics, 32, 607–625. Disney R, Henley A and Jevons D (2003) House Price Shocks, Negative Equity and Household Consumption in the UK, mimeo. Droste C and Knorr-Siedow T (2011) Social housing in Germany – changing modes for a changing society, chapter in N Houard (ed) Loger l’Europe: Le logement social dans tous ses Etats, French Documentation, Paris. Droste C, Levrier C and Wassenberg F (2008) Urban regeneration in European social housing areas, chapter in K Scanlon and C Whitehead (eds) Social Housing in Europe II, LSE, London. Early D (2000) Rent control, rental housing supply, and the distribution of tenant benefits, Journal of Urban Economics, 48, 2, 185–204. Eiglsperger M (2010) Residential property price statistics for the Euro area and the European Union, chapter in P Arestis, P Mooslechner and K Wagner (eds), Housing Market Challenges in Europe and the United States, Palgrave Macmillan, Basingstoke. Eisenbeis R A and Kaufman G G (2009) Lessons from the demise of the UK’s Northern Rock and the US’s Countrywide and Indymac, chapter in F Bruni and D T Llewellyn (eds) The Failure of Northern Rock: A mult-dimensional case study, SUERF, Vienna. EIU: Economist Intelligence Unit (2011) The Sustainability of China’s Housing Boom, Economist Intelligence Unit Report. http://macrobusiness.com.au/2011/04/eiu-report-onchinas-housing-bubble/ Accessed June 2011. Ellis L (2010) Recent Developments in the Housing Market and its Financing, Speech to the Financial Review Residential Property Conference 2010, Sydney, 18 May 2010. http://www. rba.gov.au. Accessed May 2010. Ellis L (2006) Housing and Housing Finance: The View from Australia and Beyond, Reserve Bank of Australia Discussion Paper, 2006–12. http://www.rba.gov.au. Accessed January 2007. Esping-Andersen G (1990) The Three Worlds of Welfare Capitalism, Polity Press, Cambridge. Empirica (2007) Die Generationen über 50 – Wohnsituation, Potenziale und Perspekiven. Srudie im Auftrag der LBS. Berlin. Engelhardt G V (1994) House prices and the decision to save for down payments. Journal of Urban Economics, 36, 2, 209–237. Engelhardt G V (1996) House prices and home owner saving behavior. Regional Science and Urban Economics, 26, 3, 313–336. ESRI, Economic and Social Research Institute (2001) Latest Permanent tsb/ESRI House Price Index 18/01/2011 http://www.esri.ie/irish_economy/permanent_tsbesri_house_p/ Accessed April 2011.
Jones_bref.indd 299
2/7/2012 10:18:54 AM
300
References
EMF, European Mortgage Federation (2010) Hypostat 2009 European Mortgage Federation, Brussels. http://www.hypo.org/Content/Default.asp Accessed June 2011. Evans A. W. (2003) Shouting very loudly: economics, planning and politics, Town Planning Review, 74, 195–212. Evans A W and Hartwich O M (2005) Unaffordable Housing: Fables and Myths, Policy Exchange, London. Fergus J T (1999) Where, when, and by how much does abnormal weather affect housing construction, Journal of Real Estate Finance and Economics, 18, 1, 63–87. Ferrari E (2008) Do planners need to understand housing markets? Paper presented to ACSP/ AESOP Conference, Chicago, June. Find-a-Property (2011) http://www.findaproperty.com/ First American CoreLogic (2011) www.corelogic.com Fingleton B (2008) Housing supply, demand and affordability, Urban Studies, 45, 8, 1545–1563. Fischel W A (2001), The Homevoter Hypothesis: How Home Values Influence Local Government Taxation, School Finance, and Land Use Policies, Harvard University Press, Cambridge, Massachusetts. Flavin M and Yamashita T (2002) Owner-occupied housing and the composition of the household portfolio, American Economic Review, 92, 1, 345–363. Florida R (2002) The Rise of the Creative Class, Basic Books, New York. Frankel A, Gyntelberg J, Kjeldson K and Persson M (2004) The Danish mortgage market, BIS Quarterly Review, March, 95–109. Fry R, Martin V and Voukelatos N (2010) Overvaluation in Australian housing and equity markets: wealth effects of monetary policy, Economic Record, 86(275), 465–485. FSA, Financial Services Authority (2009) Mortgage Market Review, DP09/3, FSA, London. Gabriel S A, Quigley J M and Rosenthal L A (2009) The mortgage meltdown, the economy, and public policy, The B.E. Journal of Economic Analysis and Policy, 9, 3 (Symposium), Article 1. Galster G (1997) Comparing demand-side and supply-side housing policies: sub-market and spatial perspectives, Housing Studies, 12, 4, 561–577. Genesove D and Mayer C J (2001) Loss aversion and seller behavior: Evidence from the housing market, Quarterly Journal of Economics, 80, 2, pp. 287–312. Gibb K (2002) Trends and change in social housing finance and provision within the European Union, Housing Studies, 17, 2, 325–336. Gibb K and Leishman C (2011) Delivering Affordable Housing in Troubled Times: Scotland National Report for the Joseph Rowntree Foundation Housing and Neighbourhoods Monitor. Policy Press, Bristol. Gibb K and O’Sullivan A (2008) Explaining the Relative Development Costs of Social Housing in Scotland. Scottish Federation of Housing Associations, Glasgow. Girouard N, Kennedy M and André C (2007) Has the Rise in Debt made Households more Vulnerable? Economics Department Working Paper No. 535, OECD, ECO/WKP (2006)63. Girouard N, Kennedy M, van den Noord P and André C (2006) Recent House Price Developments: The Role of Fundamentals, OECD Economics Department Working Paper No. 475, Organization for Economic Cooperation and Development, Paris. Gitelman E and Otto G (2010) Supply Elasticity in the Sydney Housing Market, mimeo. http:// members.optusnet.com.au/∼g.otto/Housing_Supply_Dec_10.pdf. Accessed January 2011. Glaeser E L (2011) Triumph of the City, The Penguin Press, New York. Glaeser E L and Gyourko J (2003) The impact of building restrictions on housing affordability, Federal Reserve Bank of New York Economic Policy Review, 9, 2, 21–39. Glaeser E, Gyourko J and Saks R (2005a) Why have housing prices gone up?, American Economic Review, 95, 2, 329–333.
Jones_bref.indd 300
2/7/2012 10:18:54 AM
References
301
Glaeser E L, Gyourko J and Saks R E (2005b) Urban Growth and Housing Supply, HIER Discussion Paper No. 2,062, SSRN papers series No.658343. Glascock J and Turnbull G (1994) On the supply of landlord labor in small real estate rental firms, Journal of Real Estate Finance and Economics, 8, 1, 21–33. Goodman A C (2005) The other side of eight mile: Suburban population and housing supply, Real Estate Economics, 33, 3, 539–569. Goodman A C and Thibodeau T G (1998) Housing market segmentation, Journal of Housing Economics, 7, 121–143. Green R K and Malpezzi S (2003) A Primer on US Housing Markets and Housing Policy, AREUEA Monograph Series, 3, Washington. Green R K and Wachter S M (2007) The Housing Finance Revolution, Federal Reserve Bank of Kansas City, Paper presented at “Housing, Housing Finance, and Monetary Policy,” a Symposium sponsored by the Federal Reserve Bank of Kansas City, at Jackson Hole, Wyoming, August 30-September 1, 2007. http://www.kc.frb.org/Publicat/Sympos/2007/ PDF/Green_Wachter_0415.pdf Accessed June 2011. Green R, Malpezzi S and Mayo S (2005) Metropolitan-specific estimates of the price elasticity of supply of housing, and their sources, American Economic Review, 95, 2, 334–339. Guirguis H S, Giannikos C I and Anderson R I (2005) The US housing market: Asset pricing forecasts using time varying coefficients, Journal of Real Estate Finance and Economics, 30, 1, 33–53. Guiso L and Jappelli T (2002) Private transfers, borrowing constraints and the timing of homeownership, Journal of Money, Credit, and Banking, 34, 315–339. Gyourko J (2003) Access to home ownership in the United States: The impact of changing perspectives on constraints to tenure choice, chapter in A O’Sullivan and K. Gibb (eds) Housing Economics and Public Policy, Blackwell, Oxford. Gyourko J, Mayer C and Sinai T (2006) Superstar Cities, NBER Working Paper 12355, NBER, Cambridge, Massachusetts. http://www.nber.org/papers/w12355. Accessed August 2006. Haffner M, Elsinga M and Hoekstra J (2008). Rent regulation: The balance between private landlords and tenants in six European countries, European Journal of Housing Policy, 8, 2, 217–233. Halifax (2010) FTB Mortgage Affordability improves to 12 Year High, Lloyds Bank. London. Halifax (2011) Historic House Price Data Set, Lloyds Bank. London. Hall D and Gibb K (2010) Increasing Supply within the Social Rented Sector: A Discussion Paper for the Joseph Rowntree Foundation Housing Market Task Force. Joseph Rowntree Foundation, York. Hanushek E A and Quigley J M (1979) The dynamics of the housing market: A stock adjustment model of housing consumption, Journal of Urban Economics, 6, 1, 90–111. Harloe M (1995), The People’s Home? Social Rented Housing in Europe and America, Blackwell, Oxford. Haurin D R (1991) Income variability, homeownership, and housing demand, Journal of Housing Economics, 1, 60–74. Haurin D R, Hendershott P and Wachter S (1995) Wealth Accumulation and Housing Choices of Young Households: An Explanatory Investigation, NBER working paper, No.5070. National Bureau of Economic Research, Cambridge, Massachusetts. Hayashi F and Prescott E C (2002) The 1990s in Japan: A lost decade, Review of Economic Dynamics, 5, 206–235. HBF, House Builders Federation (2011) Housing Market Report No 226, April 2011, HBF, London. HCA, Housing and Communities Agency (2010) Meeting the Challenge: Market analysis. Findings of the Housing Finance Group, HCA, London. Hegedus J (2007) Social housing in transition countries’ chapter in C Whitehead C and K Scanlon (eds) Social Housing in Europe I, LSE, London.
Jones_bref.indd 301
2/7/2012 10:18:54 AM
302
References
Hellebrandt T and Kawar S (2009) The economics and estimation of negative equity. Bank of England Quarterly Bulletin 2009Q2. http://www.bankofengland.co.uk/publications/ quarterlybulletin/qb0902.pdf Accessed June 2011. Hills J (2007) Ends and Means: The Future Roles of Social Housing in England, Department for Communities and Local Government, London. Hills J, Hubert F, Tomann H and Whitehead C (1990) Shifting subsidies from bricks and mortar to people: Experiences in Britain and West Germany, Housing Studies, 5, 3, 147–167. Himmelberg C, Mayer C and Sinai T (2005) Assessing high house prices: bubbles, fundamentals and misperceptions, Journal of Economic Perspectives, 19, 4, 67–92. HM Treasury (2010) Budget 2010: Securing the Recovery. Economic and Fiscal Strategy Report and Financial Statement and Budget Report, HM Treasury, London. Hori M and Shimizutani S (2004) Asset holding and consumption: Evidence from Japanese panel data in the 1990s, Seoul Journal of Economics, 17, 153–179. Horioka C Y (2006) The causes of Japan’s ‘lost decade’: The role of household consumption, Japan and the World Economy, 18, 378–400. Hou Y (2009) Urban Housing Markets in China, Report 88, Department of Real Estate and Construction Management, Royal Institute of Technology (KTH), Stockholm, Sweden. House of Commons Select Committee on Affordability and Housing Supply (2007) Affordability and the Supply of Housing, Third Report of Session 2005–06, Volume 1: Report and Minutes, Stationery Office, London. House of Commons – Treasury Committee (2008), The Run on the Rock, Fifth Report of Session 2007–08 Volume I: Report and Minutes, Stationery Office, London http://www.suerf. org/download/studies/study20091.pdf Hu J, Su L, Jin S and Jiang W (2006) The rise in house prices in China: Bubbles or fundamentals?, Economics Bulletin, 3, 7, 1–8. Hulse K, Burke T, Ralston L and Stone W (2010) The Benefits and Risks of Home Ownership for Low-Moderate Income Households, AHURI Final Report No. 154. Australian Housing and Urban Research Institute, Melbourne. http://www.ahuri.edu.au. Accessed October 2010. Hulse K, Jones C and Pawson H (2010) Tenurial “competition”, tenure dynamics and the private rented sector: An international reappraisal, Journal of European Real Estate Research, 3, 2, 138–156. IMF, International Monetary Fund (2008) The Changing Housing Cycle and the Implications for Monetary Policy, International Monetary Fund World Economic Outlook, April, Chapter 3. http://www.imf.org/external/pubs/ft/weo/2008/01/pdf/c3.pdf. Accessed July 2010. IMF, International Monetary Fund (2009), Global Prospects and Policies, International Monetary Fund World Economic Outlook, October, Chapter 1. At http://www.imf.org/ external/pubs/ft/weo/2009/02/pdf/c1.pdf. Accessed July 2010. IPSOS MORI (2009) Attitudes to Housing – Findings from IPSOS MORI Public Affairs Monitor Omnibus Survey (England), Department for Communities and Local Government, London. http://www.communities.gov.uk/documents/housing/pdf/1298556.pdf Accessed July 2011. Jaffee D (2008) The U.S. Subprime Mortgage Crisis: Issues Raised and Lessons Learned. Presentation on April 11, 2008 at the Workshop on Fiscal and Monetary Policies and Growth Commission on Growth and Development. http://www.growthcommission.org/storage/ cgdev/documents/jaffee.pdf. Accessed June 2011. Jappelli T and Pagano M (1994) Saving, growth, and liquidity constraints, The Quarterly Journal of Economics, 109, 1, 83–109. Jenkins S (2011) This localism bill will sacrifice our countryside to market forces, Guardian, 28 July. http://www.guardian.co.uk/commentisfree/2011/jul/28/localism-bill-sacrifice-countrysidemarket Jones C (1984) The future of building societies, National Westminster Bank Quarterly Review, May.
Jones_bref.indd 302
2/7/2012 10:18:54 AM
References
303
Jones C (2007) Private investment in rented housing and the role of REITS, European Journal of Housing Policy, 7, 4, 383–400. Jones C (2009) Right to Buy, chapter in P Malpass and R Rowlands (eds) Housing Markets and Policy, Routledge. Jones C (2010) Residential real estate investment trusts in the UK, chapter in A O’Sullivan (ed) Investing in Affordable Housing – A Radical Rethink? Volume 2, Chartered Institute of Housing, Edinburgh. Jones C and Murie A (2006) The Right to Buy, Analysis and Evaluation of a Housing Policy, Blackwell, Oxford. Jones C and Watkins C (1999) Planning for housing, chapter in P Allmendinger and M Chapman (eds) Planning Beyond 2000, John Wiley, Chichester. Jones C and Watkins C (2009) Housing Markets and Planning Policy, Wiley-Blackwell, Oxford. Jones C, Coombes M and Wong C (2012, forthcoming) A system of national tiered housing market areas and spatial planning, Environment and Planning B. Jones L D (1995) Testing the central prediction of housing tenure transition models,” Journal of Urban Economics, 38, 50–73. Kasparova D and White M (2001) The responsiveness of house prices to macroeconomic forces: A cross-country comparison. European Journal of Housing Policy, 1, 3, 385–416. Kemeny, J (1995) From Public Housing to the Social Market, Routledge, London. Kemeny J (2005), The really big trade-off between home ownership and welfare: Castles’ evaluation of the 1980 thesis, and a reformulation 25 years on, Housing, Theory and Society, 22, 2, 59–75. Kennedy S (2010) Housing Supply and Affordability, speech given at the Council of Capital City Lord Mayors’ Towards a National Urban Policy Summit, 27 May 2010. http://www. treasury.gov.au. Accessed January 2011. Kholodilin K A, Menz J-O and Siliverstovs B (2009) What drives housing prices down? Evidence from an international panel, Journal of Economics and Statistics, 230, 1, 59–76. http://ideas.repec.org/a/jns/jbstat/v230y2010i1p59-76.html Accessed November 2011. Knorr-Siedow T (2008) Innovations from below? A new concept of social housing, chapter in K Scanlon and C Whitehead C (eds) Social Housing in Europe II, London School of Economics, London. Krol D and Svorny S (2005) The effect of rent control on commute times, Journal of Urban Economics, 58, 3, 421–436. Kröncke T-A, Schindler F and Westerheide P (2011) Finanzierungsstrategien wohnungswirtschaftlicher Akteure unter veränderten Rahmenbedingungen auf den Finanzierungsmärkten, BMVBSOnline-Publikation Bd. 4/11. Kuijs L (2005) Investment and Saving in China, Working Paper #WPS3633, East Asia Pacific Poverty Reduction and Economic Management, The World Bank Office, Beijing, China. Kurzrock B-M, Mokinski F, Schindler F and Westerheide P (2011), Do Residential Property Companies Systematically Adjust Their Capital Structure? The Case of Germany, ZEW Discussion Paper No. 11–014, Mannheim. La Cava G and Simon J (2003) A tale of two surveys: Household debt and financial constraints in Australia, Reserve Bank of Australia Research Discussion Paper 2003–08. http://www.rba. gov.au. Accessed 22 July 2003. La Cava G and Simon J (2005) Household debt and financial constraints in Australia, Australian Economic Review, 38, 1, 40–60. Lamont O and Stein J C (1999) Leverage and house-price dynamics in US Cities. The RAND Journal of Economics, 30, 3, 498–514. Law Commission (2007) Encouraging Responsible Letting: A Consultation Paper, Consultation Paper No 181, Law Commission, London. http://www.justice.gov.uk/lawcommission/docs/ cp181_Housing_Encouraging_Responsible_Letting_Consultation.pdf Accessed July 2011.
Jones_bref.indd 303
2/7/2012 10:18:54 AM
304
References
Leamer E E (2007) Housing is the Business Cycle, Paper presented at “Housing, Housing Finance, and Monetary Policy,” a Symposium sponsored by the Federal Reserve Bank of Kansas City, at Jackson Hole, Wyoming, August 30-September 1, 2007. http://www. kansascityfed.org/publicat/sympos/2007/PDF/Leamer_0415.pdf Accessed July 2011. Leishman, C and Rowley, S (2012, in press) Affordable housing, chapter in W Clark, D Clapham and K Gibb (eds) Handbook of Housing Studies. Sage, London. Levin L (1998) Are assets fungible? Testing the behavioral theory of life-cycle savings, Journal of Economic Behavior and Organization, 36, 59–83. Lévy-Vroelant C, Reinprecht C, Wassenberg F (2008) Learning from history: changes and path dependency in the social housing sector in Austria, France and the Netherlands (1889–2008), chapter in K Scanlon K and C Whitehead (eds) (2008) Social Housing in Europe II, London School of Economics, London. Li L H (2002) Impacts of land use rights reform on urban development in China, Review of Urban & Regional Development Studies, 11, 3, 193–205. Llewellyn D T (2009) The Northern Rock crisis: A multi-dimensional process, chapter in F Bruni and D T Llewellyn (eds) The Failure of Northern Rock: A mult-dimensional case study, SUERF, Vienna. Lloyds Bank Group (2010) The UK Housing Market over the past 50 Years, Lloyds Bank Group, London. http://www.lloydsbankinggroup.com/media/pdfs/research/2010/50_Years_of_ Housing_UK.pdf Accessed July 2011. Loikkanen H and Lonnqvist H (2007) Metropolitan housing markets - the case of Helsinki, chapter in A Andersson, L Pettersson and U Stromquist (eds) European Metropolitan Housing Markets, Springer. Lou J, Palomar J, Randolph Jr P A and Zhang X (2006) Low and Moderate Income Housing Policy in the People’s Republic of China, Report submitted by the Peking University Center for Real Estate Law, International Housing Coalition (IHC), World Urban Forum III, June, 19–23, 2006, Vancouver, Canada. Loungani P (2010) Housing prices: more room to fall?, Finance and Development, March, 47, 1, 16–19. Lunde J (2005a) Structural changes in the Danish market for owner-occupation, chapter in P Boelhouwer, J Doling J and M Elsinga (eds) Home Ownership: Getting In, Getting From, Getting Out, Delft University Press. Delft, 117–144, Lunde J (2005b) The owner occupiers’ capital structure during a house price boom. – Does negative equity exist as a permanent feature in the Danish housing market. Working Paper. WP 2005-3. Department of Finance, Copenhagen Business School, Copenhagen. Lunde J (2009) Financial soundness indicators for owner occupiers, Housing Studies, 24, 1, 47–66. Lux M (2003) Housing Policy: An End or a Beginning? Open Society Institute, Budapest. Macfarlane I (2003) Do Australian households borrow too much, Reserve Bank of Australia Bulletin, April, 7–16. http://www.rba.gov.au/PublicationsAndResearch. Accessed June 2004. Maclennan D and Williams R (1990) Affordable Housing in Britain and America. Joseph Rowntree Foundation, York. Malpass P (1999) Housing policy: does it have a future, Policy and Politics, 27, 217–228. Malpass P (2008a) Housing and the new welfare state: Wobbly pillar or cornerstone?, Housing Studies, 23, 1, 1–19. Malpass P (2008b) Histories of social housing: a comparative approach, chapter in K Scanlon and C Whitehead (eds) Social Housing in Europe II, London School of Economics, London. Malpezzi S and Maclennan, D (2001) The long-run price elasticity of supply of new residential construction in the United States and the United Kingdom”, Journal of Housing Economics, 10, 278–306. Malpezzi S and Vandell K (2002) Does the low-income housing tax credit increase the supply of housing?, Journal of Housing Economics, 11, 360–380.
Jones_bref.indd 304
2/7/2012 10:18:54 AM
References
305
Mason J (1977) The Supply curve for housing, Real Estate Economics, 7, 3, 362–377. Mayer C J. and Somerville C T (1997) Unifying Empirical and Theoretical Models of Housing Supply, Federal Reserve Bank of Boston Working Paper, Boston. http://ideas.repec.org/p/fip/ fedbwp/96-12.html, Accessed November 2011. Mayer C J and Somerville C T (2000) Residential construction: Using the urban growth model to estimate housing supply, Journal of Urban Economics, 48, 85–109. Mayo S and Sheppard S (2001) Housing supply and the effects of stochastic development control, Journal of Housing Economics, 10, 109–128. Mayes D. (1979) The Property Boom: The Effects of Building Society Behaviour on House Prices, Martin Robertson, Oxford. McFarlane A (2003). Rent stabilization and the long-run supply of housing, Regional Science and Urban Economics, 33, 3, 305–333. McMaster R, U’ren G, Carnie J, Strang G and Cooper S (2008) An Assessment of the Value of Planning Agreements in Scotland, Scottish Government, Edinburgh. Meen G (2001) Modelling Spatial Housing Markets, Kluwer Academic Press. Meen G (2002a) The times series behaviour of house prices: A transatlantic divide, Journal of Housing Economics, 11, 1–23. Meen G (2002b) An Overview of Issues in Housing Modelling and Policy, University of Reading Papers, Reading. Meen G (2003) Housing cycles and efficiency, Scottish Journal of Political Economy, 47, 2, 114–140. Mercer Oliver Wyman (2003) Study on the Financial Integration of European Mortgage Markets, European Mortgage Federation, Brussels. http://www.nykredit.dk/investorcom/ ressourcer/dokumenter/pdf/Wymanreport.pdf Accessed June 2011. Monk S (1999) The Use of Price in Planning for Housing: a Literature Review. DP105, Department of Land Economy, University of Cambridge, Cambridge. Monk S and Whitehead C (2010) Making Housing More Affordable: The role of intermediate tenures, Blackwell, Oxford. More A, Pawson H and Scott S (2005) Management Effectiveness, Theme 5, ODPM, London. Moriizumi Y (2003) Targeted saving by renters for housing purchase in Japan, Journal of Urban Economics, 53, 494–509. Moriizumi Y and Naoi M (2011) Unemployment risk and the timing of homeownership in Japan, Regional Science and Urban Economics, 41, 227–235. Muellbauer J N (2007) Housing, Credit and Consumer Expenditure, Paper presented at “Housing, Housing Finance, and Monetary Policy,” a Symposium sponsored by the Federal Reserve Bank of Kansas City, at Jackson Hole, Wyoming, August 30-September 1, 2007. http://www.kansascityfed.org/publicat/sympos/2007/PDF/Muellbauer_0415.pdf Accessed June 2011. Muellbauer J and Murphy A (2008) Housing markets and the economy: the assessment, Oxford Review of Economic Policy, 24, 1, 1–33. Mullins D and Pawson H (2010) After Council Housing: Britain’s New Social Landlords, Palgrave, Basingstoke. Munro M, Pawson H and Monk S (2005) Widening Choice, Theme 4, Evaluation of English Housing Policy 1995–2000, Office of Deputy Prime Minister, London. Nagy J (1997) Do vacancy decontrol provisions undo rent control?, Journal of Urban Economics, 42, 1, 64–78. NAR, National Association of Realtors (2011). See http://www.realtor.org. NHSC, National Housing Supply Council (2009) State of Supply Report 2008. http://www. nhsc.org.au/. Accessed April 2009. NHSC, National Housing Supply Council (2010) 2nd State of Supply Report. http://www.nhsc. org.au/ Accessed April 2010.
Jones_bref.indd 305
2/7/2012 10:18:54 AM
306
References
NHPAU, National Housing and Planning Advice Unit (2007) Affordability Matters, Department for Communities and Local Government, London. Nationwide (2011) First Time Buyer Affordability Measure, Nationwide, London. Newhaven Research (2008) Delivering Affordable Housing through the Planning System in Scotland, Chartered Institute of Housing, Edinburgh. Norris M and Winston N (2011) Transforming Irish home ownership through credit deregulation, boom and crunch, International Journal of Housing Policy, 11, 1, 1–21. Northern Rock (2011) Company Information, http://companyinfo.northernrock.co.uk/investor Relations/corporateProfile/historicOverview.asp O’Sullivan, A (2003) Economics and housing planning, chapter in A O’Sullivan and K Gibb (eds) Housing Economics and Public Policy, Blackwell, Oxford. O’Sullivan A and Gibb K (2008) The Credit Crunch and the Scottish Housing System. Chartered Institute of Housing (Scotland), Glasgow. OECD (2011) Non-financial accounts by sectors, Financial accounts and Financial Balance sheets, OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, Accessed August 2011. ODPM, Office of the Deputy Prime Minister (2003) Sustainable Communities: Building for the Future, ODPM, London. ODPM, Office of the Deputy Prime Minister (2006) Delivering Affordable Housing,ODPM, London. OECD (2009) “DENMARK”. OECD Economic Surveys. 2009/19. OECD, Paris. OECD (2010) Working Party No. 1 on Macroeconomic and Structural Policy Analysis. The Role of Structural Policies in Housing Markets. Annex 6: Housing Policies in OECD and Candidate for Accession Countries. Survey-based Data and Implications, ECO/CPE/ WP1(2010)14/ANN6, OECD, Paris. OECD (2011a) Economic Outlook No 89, OECD, Paris. OECD (2011b) Economic Surveys France 2011, OECD, Paris. OFT, Office of Fair Trading (2008) Homebuilding in the UK: A market study, Office of Fair Trading, London. Olsen S A (2010) Analyzing Danish residential investments around recessions. An empirical investigation from 1947 to 2009. Master Thesis. Cand. merc. FIR. (In Danish). Ogawa K and Wan J (2007) Household debt and consumption: A quantitative analysis based on household micro data for Japan, Journal of Housing Economics, 16, 127–142. Onado M (2009) Northern Rock: Just the tip of the iceberg, chapter in F Bruni and D T Llewellyn (eds) The Failure of Northern Rock: A mult-dimensional case study, SUERF, Vienna. Otto G (2007) The growth of house prices in Australian capital cities: What do economic fundamentals explain?, Australian Economic Review, 40, 3, 225–238. Oxley M (2009) Financing Affordable Social Housing in Europe, UN-HABITAT, Nairobi. Oxley M, Lishman R, Brown T, Haffner M and Hoekstra J (2011) Promoting Investment in Private Rented Housing Supply, International Policy Comparisons, Department for Communities and Local Government, London. http://www.communities.gov.uk/documents/ housing/pdf/1759530.pdf Accessed July 2011. Panell B (2003) Denmark: Probably the best housing finance system in the world? CLM Housing Finance. Winter, 27–34. Paragon (2011) The UK Private Rented Sector and Buy-to-Let Market 2011, Paragon, London. Paris C and Blackaby F (1979) Not Much Improvement, Heinemann, London. Park M J (2011) Overview of the Housing Recession, University of Southern California, Department of Economics, Los Angeles, CA. Parkinson M, Ball M, Blake N and Key T (2009) The Credit Crunch and Regeneration: Impact and Implications. Department for Communities and Local Government, London. Pawson H and Wilcox S (2010) UK Housing Review 2010–2011. Chartered Institute of Housing, Coventry.
Jones_bref.indd 306
2/7/2012 10:18:54 AM
References
307
Poterba J M (1991) House price dynamics: The role of tax policy and demography, Brookings Papers on Economic Activity, 2, 143–203. Poterba J (2001) Demographic structure and asset returns, Review of Economics and Statistics, 83, 4, 565–584. Prescott J (2000) Statement on PPG 3 by the Deputy Prime Minister, Department of the Environment, Transport and Regions, London. Priemus H and Dieleman F (2002) Social housing policy in the European Union: Past, present and perspectives, Urban Studies, 39, 2, 191–200. Priemus H and Boelhouwer P (1999) Social housing finance, trends and opportunities, Urban Studies, 36, 4, 633–646. Priemus H (2010) Housing finance reform in the making: The case of the Netherlands, Housing Studies, 25, 5, 755–764. Prior A (2005) UK planning reform: a Regulationist interpretation, Planning Theory and Practice 6, 4, 465–484. Productivity Commission (2004) First Home Ownership, Report No. 28. http://www.pc.gov.au Accessed June 2004. Productivity Commission (2005) Economic Implications of an Ageing Australia. http://www. pc.gov.au. Accessed May 2005. Productivity Commission (2011) Caring for Older Australians, Draft Inquiry Report. http:// www.pc.gov.au Accessed January 2011. Pryce G (1999) Construction elasticities and land availability: A two-stage least-squares model of housing supply using the variable elasticity approach, Urban Studies, 36, 13, 2283–2304. Qin X and Ren R (2008) Comparable household saving rates for China and the United States, Review of Income and Wealth, 54, 4, 656–670. Quigley J M (1997) The Economics of Housing, Vol. 1 and 2. Edward Elgar Publishing, Massachusetts. Rady S and Ortalo-Magné F (1998) Housing market fluctuations in a life-cycle economy with credit constraints. Discussion Paper Series. LSE Financial Markets Group. London. Rady S and Ortalo-Magné F (2002) Home ownership: low household mobility, volatile housing prices, high income dispersion. CESifo working paper, Center for Economic Studies, Munich. http://www.cesifo-group.de/portal/pls/portal/docs/1/1189990.PDF Accessed July 2011. Rady S and Ortalo-Magné F ( 2004) Housing transactions and macroeconomic fluctuations. Journal of Housing Economics, 13, 4, 287–303. Rady S and Ortalo-Magné F (2006) Housing market dynamics. The Review of Economic Studies, 73, 2, 459–485. Realty Trac (2011) See http://www. Realty Trac.com Reinhart C M and Reinhart V R (2010) After the Fall. Paper presented at Economic Policy Symposium, Federal Reserve Bank of Kansas City, Jackson Hole. August 2010. http://www. kansascityfed.org/publicat/sympos/2010/reinhart-paper.pdf Accessed July 2011. Reinhart C M and Rogoff K S (2009) This Time is Different. Eight Centuries of Financial Folly, Princeton University Press. Princeton and Oxford. Reserve Bank of Australia (2005) Financial Stability Review, March 2005. http://www.rba.gov. au/PublicationsAndResearch/. Accessed June 2007. Richards A (2009) Housing market developments, Reserve Bank of Australia Bulletin, October 2009, 20–26. http://www.rba.gov. Accessed October 2009. Richards A (2008) Some observations on the cost of housing in Australia, Reserve Bank of Australia Bulletin, April 2008, 26–36. http://www.rba.gov. Accessed November 2009. RICS, Royal Institution of Chartered Surveyors (2011) Valuation of Land for Affordable Housing, RICS, London. Robst J, Deitz R and McGoldrick K (1999) Income variability, uncertainty and housing tenure choice, Regional Science and Urban Economics, 29, 219–229.
Jones_bref.indd 307
2/7/2012 10:18:54 AM
308
References
Roistacher E (1992) Rent regulation in New York City: Simulating decontrol options, Journal of Housing Economics, 2, 2, 107–138. Rothenberg J (1967) Economic Evaluation of Urban Renewal, Brookings Institute, Washington. Rouwendal J (2009) Housing wealth and household portfolios in an ageing society, De Economist, 157, 1, 1–48. Rowlands R, Murie A and Tice A (2006) More than Tenure Mix: Developer and Purchaser Attitudes to New Housing Estates, Chartered Institute of Housing/Joseph Rowntree Foundation, York. Rugg J and Rhodes D (2008) The PRS: Its Contribution and Potential, University of York, Centre for Housing Policy, York. http://www.york.ac.uk/inst/chp/publications/PDF/prsreviewweb.pdf Savageau D (2007) Places Rated Almanac, Places Rated Books, Washington. Savills (2007) Student Housing Report, Savills, London. Scanlon K (2010) International experience, chapter in S Monk and C Whitehead (eds) Making Housing More Affordable: The role of intermediate tenures, Blackwell, Oxford. Scanlon K and Whitehead C (2004) International Trends in Housing Tenure and Mortgage Finance, Council of Mortgage Lenders, London. http://www.cml.org.uk/cml/publications/ research?keyword=&key_area=0&date=0&page=8 Accessed April 2011. Scanlon K and Whitehead C (eds) (2011) Social Housing in Europe II, London School of Economics, London. Scanlon K, Lunde J and Whitehead C (2008) Mortgage product innovation in advanced economies: More choice, more risk, European Journal of Housing Policy 8, 2, 109–131. Scanlon K, Lunde J and Whitehead C (2011) Responding to the housing and financial crises: mortgage lending, mortgage products and government policies, International Journal of Housing Policy, 11, 1, 23–49. School of Planning and Housing (2001) The Role of the Planning System in the Provision of Housing, Scottish Executive, Edinburgh. Scottish Government (2007) Firm Foundations, Scottish Government, Edinburgh. Scottish Government (2010) Housing: Fresh Thinking, New Ideas, Scottish Government, Edinburgh. Scottish Government (2011) Homes Fit for the 21st Century: The Scottish Government’s Strategy and Action Plan for Housing in the Next Decade, 2011–2020, Scottish Government, Edinburgh. Sheiner L (1995) Housing prices and the savings of renters. Journal of Urban Economics, 38, 1, 94–125. Shiller R (2006) The difference in saving rates between China and the US, Economist’s View August 25. http://economistsview.typepad.com/economistsview/2006/08/differences_ in_.htm Accessed June 2011. Shiller R (2007a) Understanding Recent Trends in House Prices and Home Ownership, Cowles Foundation Discussion Paper No. 1630. http://cowles.econ.yale.edu/. Accessed May 2008. Reprinted in Housing, Housing Finance and Monetary Policy, Jackson Hole Conference Series, Federal Reserve Bank of Kansas City, 2008, 85–123. Shiller R J (2007b) Historic Turning Points in Real Estate, Cowles Foundation Discussion Paper No. 1610. Yale University, New Haven. http://cowles.econ.yale.edu/P/cd/d16a/d1610.pdf Accessed July 2011. Shiller R J (2009) Unlearned lessons from the housing bubble, The Economists’ Voice, 6, 7, Article 6. http://www.bepress.com/ev/vol6/iss7/art6. Accessed 2009. Shiratsuka S (2003) Asset price bubble in Japan in the 1980s: Lessons for financial and macroeconomic stability, IMES Discussion Paper Series, No.2003-E-15. http://www.imes.boj.or.jp/ english/publication/edps/fedps_index.html. Somerville C T (1999) Residential construction costs and the supply of new housing: Endogeneity and bias in construction cost indexes, Journal of Real Estate Finance and Economics, 18, 1, 43–62.
Jones_bref.indd 308
2/7/2012 10:18:54 AM
References
309
Stephens M, Burns N and McKay L (2002) Social market or safety net British social rented housing in a European context, Policy Press, Bristol. Stephens M, Elsinga M and Knorr-Siedow T (2008) The privatisation of social housing: Three different pathways, chapter in K Scanlon and C Whitehead (eds) Social Housing in Europe II, London School of Economics, London. Stephens M, Whitehead C and Munro M (2005) Lessons from the Past, Challenges for the Future, Evaluation of English Housing Policy 1995–2000, Office of Deputy Prime Minister, London. Stephens M and Fitzpatrick S (2008) The Future of Social Housing, Shelter, London. Stone M E (1993) Shelter Poverty: New ideas on Housing Affordability, Philadelphia: Temple University Press. Stone M E (2006a) What is housing affordability? The case for the residual income approach, Housing Policy Debate, 17, 1, 151–184. Stone M E (2006b) A housing affordability standard for the UK, Housing Studies, 21, 4, 453–476. Swank J, Kakes J and Tiemann A F (2002) The housing ladder, taxation, and borrowing constraints, Research Paper, Research Centre for European Policy (OCFEB), Erasmus University, Rotterdam. Taltavull P (2000) Activity development leadership and geographic areas in Spain. Journal of Property Investment and Finance, 18, 3, 371–397. Thalmann P (2010) Presentation at ENHR Housing Economics Conference, London, 2010. Tiwari P, Deutsch E and Moriizumi Y (2007) Housing finance arrangements, wealth positioning and housing consumption in Japan: An analysis of built-for-sale homeowners, Journal of Real Estate Finance and Economics, 34, 347–367. Topel R and Rosen S (1988) Housing investment in the United States, Journal of Political Economy, 96, 718–740. Tosics I (2006) Spatial restructuring in post-socialist Budapest, chapter in S Tsenkova and Z Nedovic´-Budic´ (eds) The Urban Mosaic of Post-Socialist Europe, Space, Institutions and Policy, Physica-Verlag, Heidleberg. Treasury (2008) The Architecture of Australia’s Tax and Transfer System. http://taxreview. treasury.gov.au/content/Content.aspx?doc=html/pubs_reports.htm. Accessed December 2008. Treasury (2009) Australia’s Future Tax System: Final Report to the Treasurer. http://taxreview. treasury.gov.au/content/Content.aspx?doc=html/pubs_reports.htm. Accessed May 2010. Tsatsaronis K and Zhu H (2004) What drives housing price dynamics: Cross-country evidence, BIS Quarterly Review, March 2004. www.bis.org/publ/qtrpdf/r_qt0403f.pdf. Accessed January 2005. Turner A (2009) The Turner Review: A regulatory response to the global banking crisis, Financial Services Agency, London. Turner B and Whitehead C (1993) Housing Finance in the 1990s, Research Report SB: 56, National Swedish Institute for Building Research, Gavle. Turner B and Whitehead C (2002) Reducing housing subsidy: Swedish housing policy in an international context, Urban Studies, 39, 2, 201–217. Tumbarello P and Wang S (2010) What Drives House Prices in Australia? A Cross Country Approach, IMF Working Paper wp/10/291, International Monetary Fund, Washington. http:// www.imf.org/external/pubs/cat/longres.cfm?sk=24507. Accessed December 2010. Urban Task Force (1999) Towards an Urban Renaissance. E & FN Spon, London. US Treasury-HUD (2011) Reforming America’s Housing Finance Market. A Report to Congress, February. Department of Housing and Urban Development, Washington DC. http://portal.hud.gov/hudportal/documents/huddoc?id=housingfinmarketreform.pdf Accessed July 2011. Van Kempen R, Dekker K, Hall S and Tosics I (2005) Restructuring Large Housing Estates, Policy Press, Bristol.
Jones_bref.indd 309
2/7/2012 10:18:54 AM
310
References
Venti S F and Wise D A (1990), But they don’t want to reduce housing equity, chapter in D.A. Wise (ed), Issues in the Economics of Aging, University of Chicago Press, Chicago, 13–29. Venti S F and Wise D A (2000) Aging and Housing Equity, NBER Working Paper 7882, NBER, Cambridge, Massachusetts. Voigtlaender, M (2009) Why is the German home ownership rate so low?, Housing Studies, 24, 3, 355–372. Wallace A (2010) Public Attitudes to Housing Joseph Rowntree Foundation, York. Ward J H (1963) Hierarchical grouping to optimise an objective function, Journal of the American Statistical Association, 58, 236–244. WCED, World Commission on Environmental Development (Brundtland Report) (1987) Our Common Future, Oxford University Press, Oxford. Wellings F (2006) British Housebuilders: History and Analysis, Blackwell, Oxford. Westerheide P (2009) The Impact of Low-Income Home Owners on the Volatility of Housing Markets, Paper presented to annual conference of European Real Estate Society, Stockholm, Sweden, June. Westerheide P (2011) Down Payment Constraints and Saving Behaviour of Prospective Home Owners. A Comparison of Germany and UK Paper presented to American Real Estate Society Conference, Seattle, April. Wheaton W and Nechayev G (2008) The 1998–2005 housing “bubble” and the current “correction”: What’s different this time? Journal of Real Estate Research, 30, 1, 1–26. Whitehead C (1974) The UK Housing Market, Saxon House, Farnborough. Whitehead C M E (1991) From need to affordability: An analysis of UK housing objectives, Urban Studies, 28, 871–88. Whitehead C (1999) The provision of finance for social housing: The UK experience, Urban Studies, 36, 4 657–672. Whitehead C (2003a) The economics of social housing, chapter in A O’Sullivan, and K Gibb (eds) Housing Economics and Public Policy, Blackwell, Oxford. Whitehead, C (2003b) Restructuring social housing systems, chapter in R Forrest and J Lee (eds) Housing and Social Change: east-west perspectives, Routledge, London. Whitehead C M E (2007) Planning policies and affordable housing: England as a successful case study, Housing Studies, 22, 1, 25–44. Whitehead C (2008) Financing social housing, chapter in K Scanlon and C Whitehead (eds) Social Housing in Europe II, London School of Economics, London. Whitehead C and Scanlon K (2007) Social housing in Europe, chapter in C Whitehead and K Scanlon (eds) Social Housing in Europe I, London School of Economics, London. Wilcox S, Fitzpatrick S, Stephens M, Pleace N, Wallace A and Rhodes D (2010) The Impact of Devolution: Housing and Homelessness. Joseph Rowntree Foundation, York. Wolswijk G (2010) Fiscal aspects of housing in Europe, chapter in P Arestis, P Mooslechner and K Wagner (eds), Housing Market Challenges in Europe and the United States, Palgrave Macmillan, Basingstoke. Wood G, Berry M, Nygaard C and Taylor E (2008) Community mix, affordable housing and metropolitan planning strategy in Melbourne, Built Environment, 34, 3, 273–290. Wood G, Ong R and Stewart M (2010) Housing taxes and the supply of private rental housing, chapter in M Stewart (ed) Housing and Tax Policy, Conference Series No 26, Australian Tax Research Foundation, 219–239. Wu J, Gyourko J and Deng Y (2010) Evaluating Conditions in Major Chinese Housing Markets, NBER Working Paper 16189, http://www.nber.org/papers/w16189.pdf?new_window=1 Yates J (2010) Tax expenditures and housing, chapter in M Stewart (ed), Housing and Tax Policy, Conference Series No 26, Australian Tax Research Foundation, 39–91. Yates J and Bradbury B (2010) Home ownership as a (crumbling) fourth pillar of social insurance in Australia, Journal of Housing and the Built Environment, 25, 2, 193–211.
Jones_bref.indd 310
2/7/2012 10:18:55 AM
References
311
Yates J and Gabriel M (2006) Housing affordability in Australia, AHURI National Research Venture 3: Housing Affordability for Lower Income Australians, Research Paper 3. http:// www.ahuri.edu.au/nrv/nrv3/NRV3_Assoc_docs.html. Accessed August 2007. Yates J and Whitehead C (1998) In defence of greater agnosticism: A response to Galster’s “ ‘Comparing demand-side and supply-side housing policies’ ”, Housing Studies, 13, 3, 415–423. Yates J and Wulff M (2000) W(h)ither low cost private rental housing?, Urban Policy and Research, 18, 1, 45–64. Yates J and Wulff M (2005) Market provision of affordable rental housing: Lessons from recent trends in Australia, Urban Policy and Research, 23, 1, 5–19. Yates J, Kendig H and Phillips B with Milligan V and Tanton R (2008) Sustaining fair shares: the Australian housing system and intergenerational sustainability, Australian Housing and Urban Research Institute, Final Report No. 111. http://www.ahuri.edu.au. Accessed 5 March 2008. Young R (2001) Housing associations: the new kid on the block, chapter in C Jones and P Robson (eds) The Health of Scottish Housing, Ashgate, Aldershot. Zhang X Q (1997) Chinese housing policy 1949–1978: The development of a welfare system, Planning Perspectives, 12, 433–455. Zhang X Q (2001) Risk and uncertainty in the Chinese housing market, Journal of Real Estate Literature, 9, 2, 161–172. Økonomi- og Erhvervsministeriet (2010) “Boligmarkedet og boligejernes økonomi.” Økonomisk Tema, 9. April.
Jones_bref.indd 311
2/7/2012 10:18:55 AM
Index Aberdeen, 247, 251 Aberdeenshire, 247 Abilene, 27 Affordability, 4, 19, 21, 62, 92, 93. 97, 105, 115–19, 131, 134, 136, 138, 147, 148, 151–2, 162, 175–8, 185, 197, 202, 203, 206, 208, 210, 213, 226, 231, 235–7, 239, 249, 255, 258, 262, 278, 285, 288–91 Affordable housing, 22, 92, 93, 157, 195, 204, 207–9, 212–16, 219, 222, 224, 235–54, 274, 278, 280, 290–291 Alaska, 31 Alliance and Leicester, 4 Andalucia, 182, 184 Angus, 247 Aragon, 182–4, 190 Arizona, 26, 33, 45 Asian financial crisis, 90 Asturias, 182–4, 190 Atlanta, 33 Austin, 34 Australia, 8, 14–18, 20, 108–27, 139, 227, 265, 274, 283, 287–8 Austria, 12, 227–31, 268 Baden-Weurttemberg, 155 Baltimore, 32, 34 Bank of England, 198, 199, 283 Bank of Japan, 60, 62, 65, 88 Bank Trelleborg, 128 Banks, 2, 9 deregulation, 7, 11, 111, 113, 122, 219, 221, 223, 226, 283, 284, 292 Barclays Bank, 199 Barker, 7, 112, 186, 204, 206, 235, 241 Basel III, 124, 292 Bavaria, 155 Beijing, 90, 97–8, 102–3, 106 Belgium, 8, 15, 22, 228, 268, 275 Berlin, 155
Birmingham (USA), 31 Boston, 32–3 Bradford and Bingley, 199, 200 Brandenburg, 155 Bremen, 155 Brussels, 268 Building societies, 11, 16, 163, 198 Bush administration, 43 Business/economic cycle, 1, 6, 10, 18, 106, 123, 174, 178 California, 26, 32–3, 35, 44–5 Canada, 4, 5, 15, 37, 41, 139 Canary Islands, 183–4, 190 Case-Shiller price index, 27–30, 32–4, 36, 53, 55–6 Castilla La Mancha, 182–4, 190, 192 Cataluna, 182–4, 193 Ceuta, 181 Charlotte, 32–3 Chengdu, 106 Chicago, 29, 33 China, 4, 16, 18, 20, 56, 90–107, 283, 287, 289 urban land reform, 92–3 Chongqing, 90, 98–9, 102–3 Clackmannanshire, 247 Cleveland, 31, 33 Clinton administration, 26, 37, 43 Columbus, 34 Consumer spending, 9 Cooperatives, 134, 154, 156, 157, 164, 268 Covered bonds, 11–12, 163, 199 Credit crunch, 1–6, 10–14, 16–24, 58–60, 66, 81, 90, 105–6, 108, 115–16, 124, 127–9, 135, 154, 156, 158, 162, 164–5, 170–186, 197–201, 212, 214, 229, 232–3, 236, 237, 242, 260, 263–4, 282–92 Czech Republic, 139, 227, 228
Challenges of the Housing Economy: An International Perspective, First Edition. Edited by Colin Jones, Michael White and Neil Dunse. © 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
Jones_bindex.indd 312
2/9/2012 10:53:00 AM
Index
Dallas, 33, 274 Demographic trends, 7, 20, 24, 56, 91, 105, 109, 110–112, 120, 121, 124, 156, 162, 168, 171, 184, 185, 213, 215, 229, 257, 263–5, 289 Denmark, 4–6, 8, 12, 14–18, 41, 128–52, 224, 227, 228, 230, 268, 282, 286, 288 Denver, 33 Detroit, 31, 33 Dresden, 223 Dundee, 247, 251 Edinburgh, 247, 251 East Ayrshire, 247 East Dunbartonshire, 247 East Europe transition economies/countries, 221, 223, 225, 227, 231, 233 East Lothian, 247, 251 East Midlands, 209–11 East of England, 209–11 Elderly, 20–21, 64, 67, 74, 82, 88, 152, 169, 213, 217, 230, 240, 292 Equity withdrawal, 9, 26, 36, 39, 41, 59, 78, 88, 125, 145–7, 213, 283, 288 European Union, 8, 16, 40, 131, 175, 221 Eurozone, 131, 173, 175, 176 Extremadura, 182–4 Falkirk, 247, 251 Family life cycle, 3, 75–8, 116, 126, 134 Fannie Mae, 11, 25, 37, 39–42, 44 FHA, 39–40, 42 Fife, 247 Financial accelerator, 9, 153, 165 Financial Crisis see Credit crunch Financial institutions, 11, 36, 58, 60, 141, 172, 174, 177, 244, 253, 274, 275, 278 Finland, 4–6, 8, 14, 15, 139, 227, 228, 265, 275 First time buyers/purchasers, 3–4, 17, 19–20, 57, 69, 81, 83, 93, 111, 114, 116–19, 126, 135, 142, 165, 195, 200, 202–3, 236, 250, 252, 256, 259, 263–5, 287–8 Fiscal constraints, 17, 22, 226, 236, 241, 291 Flip(ping), 19, 50–51, 285 Florida, 26, 33, 35, 45 Foreclosures, 12–13, 14, 18–19, 22, 26, 34, 35–7, 38, 41–2, 47–58, 66, 130, 139, 147, 151, 202, 226, 255, 286
Jones_bindex.indd 313
313
Fort Worth, 34 France, 4, 5, 8, 12, 14, 15, 22, 139, 158–61, 218, 223, 227–31, 233 Freddie Mac, 25, 39–42, 44 Fresno, 32 Gearing, 7, 288 Generation differences, 3–4, 18, 20–23, 109, 117–19, 121–2, 126–7, 130, 134, 152, 156, 162, 195, 215, 234, 262, 267, 288–9, 292–3 German reunification, 154–5, 160–161 Germany, 4, 6, 8, 12, 15, 18, 21, 24, 139, 153–69, 218, 223, 227–8, 230–231, 267–8, 274, 283, 289 Government Housing Loan Corporation, 60–61, 65–6 Grainger Trust, 276 Grampian, 246 Greece, 8, 15, 17, 227 Guangdong, 99, 102 Guangzhou, 90, 98, 103, 106 Halifax Bank price indices, 16, 200–201 Hamburg, 155 Hessen, 155 Highland, 247, 251 Home ownership rates see Owner occupation rates House price to income ratio, 3, 20, 59, 63, 71, 96–9, 106, 114, 132, 160, 162, 175, 202, 258, 263 House price trends, 1–22, 27–39, 47–56, 58–68, 71–3, 81–2, 91–2, 94, 96–9, 108, 113–18, 120–121, 130–133, 151–2, 158–60, 170, 173–6, 178, 185, 197, 200–202, 286–7 Housebuilding, 7, 10, 14, 17, 21–2, 62, 95–7, 131–2, 170–174, 176–9, 182–6, 188, 195–7, 202–15, 217, 219, 225, 236–9, 241–8, 251, 270, 282, 289–90, 292 Housing associations, 196–7, 208–9, 223, 230, 236, 238–45, 247–9, 252–3 Housing need(s), 42, 56, 93, 98, 109, 119, 125, 126, 140, 169, 177, 185–6, 195, 205–6, 215, 217–18, 220, 222, 224, 228, 231, 232, 235, 237–8, 245, 247, 249–50, 254, 259, 274 Hungary, 227, 228, 230
2/9/2012 10:53:00 AM
314
Index
Iceland, 8, 14, 138 IMF, 2, 108, 120, 295, 297, 302, 309 Indianopolis, 34 Inflation, 91, 101, 103, 158, 175 Interest rates, 2, 6–10, 12, 16–17, 21, 24, 26, 61, 68–71, 73, 80, 84, 88, 93–4, 98, 100–102, 104–6, 110–11, 114–17, 119–21, 123–4, 129, 131, 133, 135–7, 139, 147–52, 162–5, 172, 175, 178, 180, 185, 188–90, 218, 228, 232, 236, 284, 286–7 Intermediate housing, 204, 229, 235, 243–5, 249–53, 258, 291 Ireland, 2–6, 12, 14–17, 138, 139, 153, 158, 159, 227, 228, 230, 265 Italy, 4, 5, 12, 15, 17, 139, 158, 159, 161, 227, 268 Jacksonville, 34, 36 Japan, 4–6, 14, 15, 18, 19, 24, 58–89, 139, 283, 287, 289, 292 Kansas City, 34 Key workers, 205, 229, 276 Kiel, 223 Korea, 15 Labour market, 19, 58–60, 69, 71, 74, 81, 82, 85, 88, 112, 117, 162, 173, 186, 216, 217, 266–7, 292 Lanarkshire, 246 Land price/value, 62–6, 74, 81, 88, 92–3, 105, 112, 122, 188, 201, 204, 208, 212, 218, 224 Las Vegas, 9, 33, 35, 47–57, 285–6 Lehman Brothers, 13, 128, 199 Loan to income ratio, 114–15 Loan to value ratio, 7, 13, 53, 124, 135, 163, 165–6, 175–6, 197, 200, 214, 284, 287 London, 56, 208–12, 241, 263, 266, 269 Los Angeles, 27, 29, 32–3 Lothians, 246–7, 251 Louisville, 27, 34 Lower Saxony, 155 Luxembourg, 227–8 Maarstricht Treaty, 131 Madrid, 182–4, 190, 193 Market failure, 225, 231, 277, 280 Marriage, 71, 78 Mecklenburg-Vorpornmem, 155
Jones_bindex.indd 314
Melbourne, 112, 116 Melilla, 181 Miami, 27, 32–3 Michigan, 45 Midlothian, 247 Migration, 7, 111, 121, 172, 177, 184–5, 263, 289 Milwaukee, 27, 34 Minneapolis, 33 Mississippi, 31 Mixed communities, 208, 224–6, 230, 246 Monetary policy, 20, 60, 65, 90, 92, 94, 98–106, 123, 164 Moray, 247 Mortgage backed securities, 11–13, 24, 40–41, 58, 135, 198–200, 214, 283–4, 292 Mortgage finance availability, 7, 14, 17, 21, 82, 113, 119, 165, 171, 175, 177, 180, 185, 197, 201, 214, 223, 263, 284, 287–8, 292 Mortgages fixed rate, 38, 135, 136, 139, 148, 149, 151, 163, 197, 201, 288 interest only, 26, 38, 136–9, 163, 288 variable rate, 26, 37–8, 41, 53, 80, 114, 136–9, 148–51, 163, 197, 256, 288 Murcia, 182–4, 190 Nashville, 34 National Association of Realtors, 27–36 National Housing Trust, 248, 251–3 Nationwide BS price indices, 16, 202 Negative equity, 3, 14, 17, 36, 141, 145–8, 152, 287 Netherlands, 4–5, 6, 8, 12, 15, 17, 136, 139, 148, 158–9, 161, 163, 218, 222–4, 227–30, 233, 275 New Orleans, 32, 34 New York, 29, 31–3, 271 New Zealand, 5, 14–15, 17 Nickei index, 62 Nimbyism, 204, 213 Ningxia, 99, 102 North Ayrshire, 247 North East (England), 209–11 North Lanarkshire, 247 North West (England), 209–11 Northern Rock, 198–9 Northrhine-Westphalia, 155 Norway, 4, 6, 15, 17, 22, 139
2/9/2012 10:53:01 AM
Index
Obama administration, 25, 43–4 OECD, 4–5, 14–15, 17, 100, 108, 110–111, 120, 129, 131–2, 138–9, 148, 157–9, 223, 228–9, 231, 233, 286, 294, 300, 306 Ohio, 31 Orkney Islands, 247 Orlando, 34, 36 Owner occupation rates, 8, 19–20, 37, 43, 64, 66–7, 71, 74, 81–2, 108–10, 117–19, 125–7, 133–4, 152, 154–6, 162, 169, 196, 262, 268, 275, 288, 292 Paragon, 200 Pennsylvania, 31 Pension funds see Financial institutions Perth, 247 Philadelphia, 34 Phoenix, 27, 32–3, 35 Planning agreements, 44, 195, 207–9, 237, 239–40, 242, 290 constraints see supply constraints policies, 18, 21–2, 24, 35, 44, 112, 171, 177, 195–6, 204–15, 284, 285 Poland, 227–8 Portland, 33 Portugal, 8, 14, 227 Poverty, 118–19 Private rented sector, 3, 7, 8, 23, 196–7, 199–200, 202, 210, 212, 217, 220, 250, 253, 255–81, 291 buy to let, 199, 200, 210, 212, 256, 257, 272, 273 deregulation, 257, 263 Promotion/sustainability of home ownership, 20, 36–7, 108–27 Public sector cutbacks see Fiscal constraints Rabo bank, 275 Raleigh, 34 REITs, 274–8 Rent control/regulation, 7, 23, 256, 261–2, 265–71, 276, 278–9 Reserve Bank of Australia, 112–13, 115, 120 Retirement, 9, 74, 82, 118, 213 Rhineland-Palantinade, 155 Right to Buy see Social housing privatisation Roskilde Bank, 128 Royal Bank of Scotland, 199
Jones_bindex.indd 315
315
Saarland, 155 Sacramento, 34, 36 St Louis, 27, 34 Salt Lake City, 34 San Antonio, 31, 34 San Diego, 32–3, 274 San Francisco, 32–3 San Jose, 34, 36 Saxony, 155 Saxony-Anhalt, 155 Schelswig-Holstein, 155 Scotland, 208, 235–54, 279 Seattle, 33 Securitisation see Mortgage backed securities Security of tenure, 23, 256, 265–9, 275, 278–9 Shanghai, 20, 90, 94, 97–9, 102–3, 106 Shared equity housing, 204, 235–6, 239, 242–4, 248, 250, 253 Shared ownership see Shared equity housing Shenzhen, 92, 106 Shetland Islands, 247 Slovenia, 227 Social housing, 7–8, 19, 22–3, 110, 154, 157, 168, 196–7, 204–5, 210, 213, 216–34, 236–9, 242, 244, 246, 249–50, 259, 262–6, 274–5, 285, 290 privatisation, 12, 154, 196–7, 205, 220–221, 226, 230, 241, 247, 275 Social mix see Mixed communities South Ayrshire, 247 South East (England), 208–12, 273, 278 South Lanarkshire, 247 South West (England), 209–12 Spain, 4–6, 12, 14–15, 17–18, 21, 136, 139, 153, 158–61, 170–194, 224, 227–8, 289–90 Stirling, 247, 251 Subprime lending/mortgages, 10, 12–13, 40, 53, 55–6, 58–60, 65, 67, 81, 128, 135, 163, 165, 198–9, 201, 282, 286 Subsidies, 8, 22, 25, 37, 43–4, 111, 129, 197, 218–24, 228, 230–233, 237, 258, 259, 279 Supply constraints, 7, 18, 22, 35, 112–13, 185, 188, 208–13, 259 elasticities, 21, 112, 170–194, 212, 259, 268–9 Sweden, 4, 8, 14–15, 17, 139, 222, 227–8, 230, 268
2/9/2012 10:53:01 AM
316
Index
Switzerland, 5–6, 14–15, 268, 275 Sydbank, 128 Sydney, 112 Tampa, 33 Tax relief on mortgage interest, 19, 37, 197, 290 Texas, 27, 33–5, 43, 286 Tibet, 99, 102 Todd-Frank Act, 45 Tokyo, 62, 73, 89 Trouble Asset Relief Program, 49, 57 Tucson, 34 UK, 3–9, 11–17, 22–3, 37, 74, 80, 88, 123, 138–9, 157–61, 165, 186, 195–215, 218, 223, 224, 227–30, 233, 235–54, 283–6, 289–91 Unemployment, 19, 26, 34–6, 42, 53, 55–6, 58–9, 65, 67–73, 81–5, 88, 110, 174, 186, 260, 267, 286, 289, 292
Jones_bindex.indd 316
Urban regeneration, 207, 225, 231, 273, 278 Urumqi, 90, 98, 103 USA, 1, 4–6, 8–19, 25–57, 71, 78, 90, 95, 100, 123, 135, 139, 153, 160–161, 198, 223, 227, 265, 267, 271, 274, 282, 285–6, 288, 290, 292 Valencia, 182–4, 194 Washington DC, 32–3, 36 Welfare payments/state, 7, 22, 91, 217–21, 227, 231, 233 West Lothian, 247 West Midlands (England), 209–11 West Virginia, 31 Wholesale funding see Mortgage backed securities Yorkshire and Humberside, 209–11 Zurich, 267
2/9/2012 10:53:01 AM