Housing Market Challenges in Europe and the United States
This page intentionally left blank
Housing Market Challenges in Europe and the United States Any Solutions Available? Edited by
Philip Arestis Peter Mooslechner and Karin Wagner
Editorial and selection matter © Philip Arestis, Peter Mooslechner and Karin Wagner 2009 Individual chapters © Contributors 2009 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6–10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2009 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN 978–0–230–22903–7 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 18 17 16 15 14 13 12 11 10 09 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne
Contents List of Figures
vii
List of Tables
ix
Notes on the Contributors
xi
1 Introduction: Housing Market Challenges in Europe and the United States Philip Arestis, Peter Mooslechner and Karin Wagner 2 Housing Markets in Europe and in the USA: What Are the Relevant Issues Today? Peter Mooslechner and Karin Wagner 3 Subprime Mortgage Market and Current Financial Crisis Philip Arestis and Elias Karakitsos 4 Determinants of Homeownership Rates: Housing Finance and the Role of the State Elisabeth Springler and Karin Wagner 5 The Rental Housing Market Dieter Gstach Housing Markets, Business Cycles and Economic Policies Christophe André and Nathalie Girouard
7
European Rental Markets: Regulation or Liberalization? The Spanish Case Montserrat Pareja-Eastaway and María Teresa Sánchez-Martínez Fiscal Aspects of Housing in Europe Guido Wolswijk
9 Towards a New Housing System in Transitional Countries: The Case of Hungary József Hegedüs 10
15 40
60 85
6
8
1
House Price and Other Housing Market Data: A User’s Perspective Anthony Murphy
v
109
131 158
178
203
vi
Contents
11 Residential Property Price Statistics for the Euro Area and the European Union Martin Eiglsperger
221
12 Housing and Financial Wealth in Austria: What Can Survey Data Tell Us for the Analysis of Financial Stability Issues? Pirmin Fessler, Peter Mooslechner, Martin Schürz and Karin Wagner
239
Index
264
List of Figures 2.1 2.2 2.3 4.1 4.2 4.3 5.1a 5.1b 5.2a 5.2b 5.3a 5.3b 5.4 5.5 5.6a 5.6b 6.1 6.2 7.1 7.2 7.3 7.4 8.1 9.1 9.2 9.3 9.4 9.5 9.6
Real housing investment Mortgage debt outstanding and disposable income Share of variable rate lending in new loans for house purchase and mortgage debt to GDP in 2007 Annual changes in homeownership rates in percentage points, 1997–2006 Selected economies: Supply side allowances to GDP ratio in per cent in 2001 Development of subsidized housing supply and housing permissions and completions in Austria, 1996–2003 Yearly growth rates of real rents in per cent Yearly growth rates of real rents in per cent Yearly growth rates of real rents in per cent Yearly growth rates of real rents in per cent Distribution of disposable household income Distribution of disposable household income Rented market shares by household size Rented market shares by age group Distribution of rent-to-income ratio in per cent Distribution of rent-to-income ratio in per cent OECD real house prices and the business cycle Marginal propensities to consume out of housing wealth and mortgage market indicators Rented housing in several European countries Contracts and rents in Barcelona, 1984–2007 Regulated and unregulated housing, 1962–2007 ‘Disqualified’ regulated housing Housing taxes, 2006 Macroeconomic trends: GDP, CPI, and interest on housing (%) New construction and building permits between 1989 and 2008 House price changes Housing loans, 1989–2008 Housing subsidies, 1998–2007, as per cent of the GDP Household borrowing, 1996–2008 vii
21 27 29 63 69 70 93 93 94 94 99 99 101 102 103 103 111 116 134 144 146 147 164 181 183 183 187 188 190
viii
List of Figures
10.1 Channels of transmission of the mortgage and housing crisis 10.2 Various measures of UK house prices 10.3 Various measures of Greater London house prices 10.4 Estimated long-run effects of changes in the credit conditions index 10.5 Simulated real US house prices 11.1 Residential property prices for euro area countries, annual percentage changes 11.2 Residential property prices for non-euro area EU countries, annual percentage changes 11.3 Residential property prices for the euro area at annual and semi-annual frequency, annual percentage changes 12.1 Stock and/or mutual fund share holdings by gross financial wealth decile 12.2 Which sources do you rely on when you seek information on financial issues?
203 209 209 215 216 224 225 234 252 257
List of tables 2.1 Housing market indicators 2.2 Timing of maximum correlation 4.1 Introduction and usage of securitization for housing finance in European economies and the USA 4.2 Fixed effects model for ownership rate 4.3 Fixed effects model for ownership rate (incl. interest rates) A4.1 Cost of financing – time series A4.2 Fixed effects model with instrumental variables 6.1 House prices in real terms and relative to rents and income 6.2 Short-term and long-term impact of financial and housing wealth on consumption 6.3 Long-term impact of housing equity withdrawal on consumption 7.1 Housing tenure 7.2 Tenure in Spain 7.3 Cost of rented dwellings according to the year of signature 8.1 Main tax categories affecting housing/mortgages in the euro area, 2007 8.2 Overall housing subsidy measures, 1999 A8.1 Tax on imputed rents A8.2 Mortgage interest payments tax deductibility A8.3 Tax on housing capital gains A8.4 Indirect taxes 9.1 Basic indicators of housing conditions in Hungary, 1970–2000 9.2 Housing allowance, 2000–2007 10.1 Simple mix adjusted house price examples 10.2 Start dates of some European house price data 10.3 Various measures of quarterly ln house prices for the UK
ix
17 22
67 74 76 79 80 112 115 116 135 144 150 161 165 171 172 173 174 182 196 205 206 210
x
List of Tables
10.4 Data for modeling aggregate house prices A11.1 Overview table of residential property prices in EU countries, annual percentage changes 12.1 Ownership rates by socio-economic characteristics 12.2 Logit-regression: determinants of homeownership
217 236 243 248
Notes on the Contributors Christophe André is an economist in the Economics Department of the Organisation for Economic Cooperation and Development (OECD). He has contributed to several editions of the OECD Economic Outlook and has been involved in macroeconomic modelling and forecasting, as well as in research in areas ranging from housing to monetary and fiscal policy. Philip Arestis is University Director of Research, Cambridge Centre for Economics and Public Policy, Department of Land Economy, University of Cambridge, UK; Distinguished Adjunct Professor of Economics, Department of Economics, University of Utah, USA; Senior Scholar, Levy Economics Institute, New York, USA; Visiting Professor, Leeds Business School, University of Leeds; Department of Finance and Management Studies, School of Oriental and African Studies (SOAS), University of London, UK. Martin Eiglsperger is Senior Economist Statistician in the Euro Area Accounts and Economic Data Division of the ECB’s Directorate General Statistics. He held previous positions at the Deutsche Bundesbank, the German Federal Statistical Office and the Department of Business Administration and Economics of the University of Bamberg. He received a Ph.D. degree in economics from the University of Bamberg. Pirmin Fessler is an economist at the Austrian central bank. He studied economics at the University of Vienna. His current research focuses on wealth inequality, intergenerational transfers and private households’ portfolio choice. He is member of the Household Finance and Consumption Network of the European Central Bank. Nathalie Girouard is adviser to the OECD Secretary-General. Before assuming her responsibilities in the cabinet, Nathalie was part of the team editing the OECD Economic Outlook. Her fields of research include consumption behaviour, housing and mortgage markets and their effects on the wider economy. She has published several working papers and OECD documents.
xi
xii
Notes on the Contributors
Dieter Gstach is associate professor of economics at Vienna University of Economics and Business. He is also a member of the Research Institute for Spatial and Real Estate Economics. His current research focuses on the macroeconomic role of the housing market. Jószef Hegedüs is a founding member of the Metropolitan Research Institute, which was established in 1989 in Budapest, Hungary. He has been co-organizer of the East European Working Group of the European Network for Housing Research since 1989, and a member of the Housing Policy Council, a high-level advisory group in housing policy matters in Hungary, since 1996. He has been Affiliated Professor at Corvinus University since 2007. Elias Karakitsos is Director of Guildhall Asset Management; chairman of Global Economic Research; and Associate Member of the Centre for Economic and Public Policy, University of Cambridge. He was Professor at Imperial College, Head of Economics for ten years and has acted as an adviser to governments and financial institutions, including Citibank, Oppenheimer, Allianz, Credit Agricole and Standard Chartered. He is the author of five books/monographs, 80 papers in learned journals and more than 300 reports on financial markets. Peter Mooslechner is Director of the Economic Analysis and Research Department of the Oesterreichische Nationalbank, Vienna. He teaches economics and economic policy at the Vienna University of Economics and Business Administration and is a member of the Monetary Policy Committee of the ECB. His publications cover macroeconomics, monetary and fiscal policy, financial markets and Eastern European issues. Anthony Murphy is Economics Fellow at Hertford College, University of Oxford. He is an applied econometrician and works on housing, savings and labour markets, as well as empirical finance. Montserrat Pareja-Eastaway is Associate Professor of Economics at the University of Barcelona, Spain. Housing affordability, tenure and housing policy measures are, among others, key aspects of her research. She has been the Spanish partner responsible for several EU-funded projects and is a member of the ENHR Co-ordination Committee.
Notes on the Contributors xiii
María Teresa Sánchez-Martínez is Associate Professor of Economics at the University of Granada, Spain. Her research interests are housing finance, distributive aspects of public expenditure on housing and housing policies from a comparative perspective. She has published numerous articles in specialized journals in the field of housing economy. Martin Schürz is Head of the Monetary Unit of the Austrian central bank’s Economic Analysis Division. He teaches at the Vienna University of Economics and Business Administration and at the University of Applied Sciences (bfi), Vienna. He has published several books and his research interests are in the field of wealth inequality. Elisabeth Springler is Assistant Professor at the Vienna University of Economics and Business Administration, Austria. She has a strong interest in Post-Keynesian Economics. Her current research interests are monetary economics, financial structures, regulation of financial systems and institutions, as well as housing economics. Karin Wagner works as an economist at the Oesterreichische Nationalbank. Since 1997 she has worked in the Economic Analysis Division of the Economic Analysis and Research Department. Before this, she did various research projects. Her research interests include various aspects of the Austrian economy from a macroeconomic perspective, especially housing market developments and, from a microeconomic perspective, wealth survey topics. Guido Wolswijk is Principal Economist at the European Central Bank, focusing on fiscal policies in the euro area. Before that, he worked for the Dutch central bank and at ING Bank. His research interests include fiscal policy, monetary policy and housing market developments, including the interaction between these topics.
This page intentionally left blank
1 Introduction: Housing Market Challenges in Europe and the United States Philip Arestis, Peter Mooslechner and Karin Wagner
The Oesterreichische Nationalbank (Central Bank of Austria) in September 2008 organized a conference entitled ‘Housing Market Challenges in Europe and the United States – Any Solutions Available?’ All the papers presented at the conference have been substantially revised since September 2008 to account for more recent developments in the housing markets in both Europe and the United States (USA). The result is the current book with a slightly different title, comprising 12 chapters. In what follows in this Introduction we attempt to put the contents of the book in context and at the same time summarize the contributions. Peter Mooslechner and Karin Wagner, in Chapter 2, open the discussion with their contribution entitled, ‘Housing Markets in Europe and in the USA: What are the Relevant Issues Today?’. In a significant number of recessions, the housing sector has preceded if not caused the economic downturn – a stylized fact also underlined by the current financial turmoil and economic crisis. The chapter analyzes some of the key aspects of the economic relevance of housing markets in the present context. First, it considers the effects that the housing sector and housing finance have on the macro economy. Since the housing sector accounts for a considerable part of a country’s welfare, wealth and GDP, it significantly shapes a country’s long-term economic development. In this respect it is important to pin down the underlying drivers of house prices, and to assess the economic implications of house price volatility. The differences in house price developments across countries may reflect country-specific factors like demographic differences, institutional regulations or cyclical positions; but changes in the structure of housing finance driven by globalization and liberalization are likely to play an important role as well. Above all, current macroeconomic analysis highlights the role of housing wealth as a household’s principal asset 1
2
Housing Market Challenges in Europe and the United States
and the role of mortgage debt as the largest liability in households’ balance sheets. Besides possible substantial wealth effects on consumption, changes in house prices can generate significant spillover effects for the macro economy as a whole, an obvious aspect of the current financial crisis. Second, fiscal and regulation aspects are important. Specific institutional settings together with the tax treatment of housing expenses have direct implications for a household’s income and wealth position. Again, there are fairly strong national differences in the tax deductibility of interest rates, the amount of allowances and subsidies given to households, and the instruments used for this purpose. Third, housing finance patterns have a direct effect on housing markets and on households’ wealth position. In this respect, lately there has been a decrease in the spread between interest rates on loans and the cost of funding, which may be due to increased competition. In parallel, a tendency to move away from traditional methods of financing that relied on specialized intermediaries or government programs towards more market-based systems of housing finance has become visible. At the same time, the funding of housing loans has changed significantly. In particular, there has been a boom in securitization, through which house financing has become international. As a result, fluctuations spill over much more easily from one ‘national’ market to others, indeed to international capital markets. This is a particular challenge for supervisors because of national differences in supervisory structures and practices. Finally, the chapter looks at the relationship between monetary policy and housing markets. Cross-country differences concerning the structure of mortgage loans provide for a heterogeneous transmission process. In the USA long-term fixed-rate mortgages prevail, while other countries mostly use variablerate loans, which lead to a faster transmission of monetary policy shocks. Despite marked changes in the overall transmission of monetary policy, the housing market is still one of the main channels of monetary policy. Philip Arestis and Elias Karakitsos proceed in Chapter 3 to the theme of ‘Subprime Mortgage Market and Current Financial Crisis’. The thesis of this contribution is that the current financial crisis is the result of three forces: Financial liberalization, financial innovation (what we now know as the subprime mortgage market) and easy monetary policy in the USA, the UK and other countries. The first feature is the financial liberalization policies initiated by governments both in the developed and developing world since the 1970s. The second feature is an important financial innovation that emerged following the financial liberalization experience. The financial innovation in question is based on the issue of financial structured products, such as Collateralized Debt Obligations
Philip Arestis, Peter Mooslechner, Karin Wagner 3
(CDOs) that played a key role in the swelling of the subprime market. Other forms of asset backed securities were also issued related to commercial real estate, auto loans and student loans, whereas Credit Default Swaps (CDSs) were issued to insure investors against the risk of default of the issuer. The third feature springs from the type of new economic policies pursued by a significant number of central banks around the world. This new policy is entirely focused on monetary policy, and the emphasis on frequent interest rate changes as a means of controlling inflation. In the USA, Alan Greenspan injected liquidity and cut interest rates following the Asian–Russian crisis of 1997–98, which was only partially drained later on. Afraid of deflation in the aftermath of the burst of the internet bubble, Greenspan cut interest rates from 6.5 per cent to 1 per cent and injected huge liquidity. More important, he was late and slow in draining that liquidity and reversing the rate cuts from the middle of 2004. Ben Bernanke imitated Greenspan and injected further liquidity following the ongoing credit crisis that erupted in the summer of 2007. This liquidity financed the last and most pronounced phase of the commodity bubble in the first half of 2008 which pushed the price of oil to US$ 147 per barrel. The commodity bubble was the last one in the current cycle, as it affected CPI-inflation. Whereas central banks are loath in hiking rates to curb asset price inflation, a surge in CPI-inflation falls squarely into their realm. The surge in commodity prices forced some central banks, like the ECB, to tighten monetary policy, whereas it delayed others, like the Fed and the Bank of England, from the urgently needed rate cuts, thus contributing to the downturn in the autumn of 2008. The acceleration of the economic downturn in the third quarter of 2008 burst the commodity bubble and demolished the myth of decoupling between developing and developed countries. The impact of these three types of development has been the creation of enormous liquidity and household debt in the major economies but which, in the USA and UK in particular, has reached unsustainable magnitudes and produced the current crisis. This contribution relies on these three features for an explanation of the origins of the current crisis. But ultimately the focus of this contribution is on the creation and subsequent developments in the subprime mortgage market. Elisabeth Springler and Karin Wagner turn their attention in Chapter 4 to homeownership in an attempt to examine the ‘Determinants of Homeownership Rates: Housing Finance and the Role of the State’. They argue that the current financial and real economic crises started off in the USA when the housing bubble burst and led to a global economic recession affecting both industrialized and emerging markets. Given similar
4
Housing Market Challenges in Europe and the United States
developments in the housing markets in numerous European economies compared to the pre-housing-crises experience observed in the USA, the fear has arisen recently that a similar experience is inevitable in Europe. A number of contributions have avoided analyzing the impact of institutions and the role of the state in implementing economic policies to promote the housing sector. Additionally, studies of housing bubbles, with only a few exceptions, neglect completely the impact of homeownership rates. Although – and similar to the development of house prices – homeownership rates also increased in most economies recently. Combining these two missing elements in current economic research, this chapter focuses on the determination of homeownership rates in the European economies and in the USA. Special attention is paid to the role of the state in providing tax exemptions and other indirect subsidies. These incentives of the state can additionally be promoted by measures in the financial sector. In this context, especially the liberalization of mortgage markets that led to a decrease in interest rates can be seen as a further incentive for the promotion of homeownership rates. The underlying hypotheses for the determination of homeownership rates in the light of the current crises are the following: First of all, tax incentives have a positive impact on homeownership rates. Second, interest rates have an inverse impact on homeownership rates, which lead to the conclusion that the current liberalization of European mortgage markets and the ongoing innovations in the field of housing finance have a positive impact on homeownership rates. If these interrelations can be detected, further light can be shed on the explanation of the housing crises, by connecting this evidence with the existing literature on house bubbles. As interest rates serve as an important explanatory variable of house price developments, they also turn out to be a crucial factor of housing bubbles; consequently, a clear causal interrelation between economic policies of the state and house prices can be detected. Economic policies become a direct factor for homeownership rates and gain indirectly increasing importance in the determination of housing bubbles. Taking the observed negative effects of the housing crises in the USA for individual households into account, the need to rethink economic policy emerges. Not only do housing crises not seem to be determined by market forces, they are directly fostered by the policies of the public sector owing to the increase in homeownership rates. To empirically test for the described interrelations between homeownership rates, the role of the state and housing finance products, a time series model of mortgage debt and cost of financing an owner-occupied home is applied to European economies and to the United States.
Philip Arestis, Peter Mooslechner, Karin Wagner 5
Dieter Gstach in Chapter 5, ‘The Rental Housing Market’, analyzes the current and potential role of rental housing in the macroeconomic housing literature. Recent data on rental housing are contrasted with the handling of rental housing in the literature. Particularly, results from 2006 micro data are used to provide relevant statistics regarding the role of rental housing. It is argued that the neglect of rental markets is a serious gap in macroeconomic analysis involving housing markets. In housing studies rental housing has always been an important topic. But, when it comes to the economic role of housing within more general setups, matters are very different. In this more scattered literature the rental housing sector is largely neglected. This observation even applies to special journal issues dedicated especially to the macroeconomics of housing, such as volume 13/4 of the Journal of Housing Economics or volume 24/1 of the Oxford Review of Economic Policy. This leads to the question of whether it is indeed justified to proceed with as-if-economics and simply assume the rental market away. Two variations of this question emerge: Can we really ignore the possibility that rising house prices via rising rents exert a negative long-run impact upon consumption? Is the widespread institution of rent indexing, which directly affects the consumer price index, really irrelevant for price stability? As, for example, the above-mentioned micro data suggest, renting households on average have significantly lower disposable incomes than owneroccupiers. The ratio is roughly 3:2 between owner incomes and incomes from renting with rising tendency. Data also show that households that rent typically spend 25 per cent of disposable income on rent with rising tendency again. These figures together with average shares from renting of 30 per cent in industrialized countries, amount to almost 6 per cent of aggregate consumption taking the form of rental payments (for equal average consumption propensities). Consequently, 6 per cent would also be the weight of rents in national price indexes (typically they are higher, for various reasons). So, the story about the housing market and the macro economy should be augmented by considerations of a rent channel. While it may take some time for this channel to become effective, it points in the opposite direction of the collateral effect. A major reason behind this effect could be different consumption propensities of tenants and landlords, which should come as no surprise given the significantly different incomes of homeowners and renters. Furthermore, as various policy measures influence rents directly, the latter may also play a macroeconomic role in the short run. Such measures include not only the most obvious case of direct rent and tenancy regulations, but many more, such as tax treatment of rental income, subsidization of ‘second
6
Housing Market Challenges in Europe and the United States
homes’ and, more generally, the various measures affecting the relative price between rental and owner-occupied housing. Many deficiencies of housing related statistics are well documented in the literature and have been pointed out in various contributions to this book. This is not true for the notorious lack of reliable and internationally comparable rental rate data. But these are required to test whether indeed rental housing is irrelevant for macroeconomic analysis as the relevant strand of literature on housing implies. Christophe André and Nathalie Girouard in Chapter 6, entitled ‘Housing Markets, Business Cycles and Economic Policies’, argue that from the mid-1990s to 2006–07, the vast majority of OECD countries experienced an exceptional expansion of their housing markets, both in terms of magnitude and duration. Moreover, deviating from historical patterns, the latest housing upswing has been disconnected from the business cycle, as the economic downturn of the early years of the century was not accompanied by a slowdown of housing markets. Housing has contributed to the expansion of economic activity by enhancing the effect of interest rate cuts on economic growth: Residential investment has been strong and wealth effects have supported private consumption, especially in English-speaking countries. Estimates of long-run marginal propensities to consume out of housing wealth are in the range of 0.05 and 0.08 in Australia, Canada, the Netherlands, the United Kingdom and the United States, while they are much lower in other continental European countries and Japan. Econometric investigations point to a transmission of housing wealth to private consumption through the refinancing of mortgages and home equity loans (more generally housing equity withdrawal). This extraction of liquidity from the housing market has been strongest in countries with the most innovative mortgage markets, which offered a wide range of products and allowed broad access to financing; for example through high loan-to-value ratios, second mortgages, equity release products, alternative interest rate adjustment and repayment structures or subprime loans. Financial innovations have contributed to excessive lending and housing booms, which are at the root of the current financial turmoil. Tax systems favoring home ownership through tax deductibility have sometimes exacerbated the problem. These developments have revived the debate on the role of monetary policy in asset price cycles and on the optimal features of regulation and supervision of financial markets and institutions. For monetary policy to be efficient in containing asset price bubbles, it is generally considered that three conditions need to be met: First, monetary authorities need to detect the development of a bubble in a timely manner; second, a
Philip Arestis, Peter Mooslechner, Karin Wagner 7
modest tightening of monetary policy needs to be able to control speculation; and third, expected gains from avoiding a bubble in terms of medium-term macroeconomic performance must be substantial. Obviously, the severity of the current recession requires preventive action. However, the difficulty in meeting the first two conditions implies that monetary policy is probably too blunt a tool to deal with asset bubbles, even if ‘leaning against the wind’ strategies are sometimes applied. While loose monetary policy has fuelled the housing boom, many excesses in credit expansion would have been prevented by adequate regulation and supervision. In particular, a dramatic growth in the share of assets held outside the traditional banking system, especially in investment banks, structured investment vehicles, conduits and hedge funds, has weakened the control over the financial system. Restoring conditions for a sustainable development of housing markets will imply reinforcing the supervision and regulation of the financial system, in particular controlling the level of leverage, avoiding pro-cyclical provisioning and capital standards. Improving risk management and transparency and ensuring appropriate underwriting standards are further requirements. Montserrat Pareja-Eastaway and María Teresa Sánchez-Martínez focus on the rented markets in Chapter 7, ‘European Rental Markets: Regulation or Liberalization? The Spanish Case’. They argue that housing markets show nowadays a peculiar transformation in Europe and the United States. Not only do they reflect the current economic scenario but also they portray the results of recurrent housing policies over time. The European landscape provides quite a diverse picture in terms of the dominance of one form of tenure from another. Academic literature has described the evolution of European housing systems by means of comparative analysis and the selection of representative case studies. In the majority of European countries, a massive construction of social housing, mainly publicly rented, took place after the Second World War in order to provide shelter for low-income earners. However, the tenure pattern has been developed and transformed by each country according to its own socio-political system, especially so in terms of the interaction of the demand and supply forces and housing policy priorities. Nowadays, the change of the economic scenario directly affects the approach adopted towards the mechanisms used by housing policies in order to achieve their targets. Housing markets are currently unstable and uncertainty characterizes any expectation on the final market outcome. However, in terms of tenure, there are some general trends at the European scale, that is to say, a general increase in homeownership and a fall in size and quality of the rented dwellings. The characteristics of the rented markets
8
Housing Market Challenges in Europe and the United States
and their regulations vary across Europe. In some cases, the percentage of dwellings oriented to this segment has been stable (e.g. Germany); in others, it has fallen dramatically (e.g. Spain). In many cases, rent controls or certain regulations are implemented. Different reasons lie behind the need to regulate or control the market; the question is the extent to which the evolution of the rented markets in terms of rent and size depends upon the attributes of their regulation. In other words, are there any other variables to take into account to analyze and forecast the relative importance of rented markets in Europe? In particular, how do the economic scenario and the specific measures of housing policy affect the rented sector development? The aim of Chapter 7 is two-fold: First, it provides a broad picture of the situation in Europe, covering not only structural facts of the market but also typologies of regulation; and, secondly, using this as a framework, it will go in depth on the peculiarities of the Spanish case. Regulations on the rented sector in Spain have proved to cover a wide spectrum of measures: From the freeze of rents and duration of contracts to a minimum agreement between landlords and tenants going through a period of complete liberalization. However, nowadays the size of the private rented sector scarcely reaches 10 per cent of the housing stock. Other measures may have certainly influenced this situation. Finally, the future challenges for policy measures oriented towards this sector regarding not only the current Spanish situation but also other European markets, are discussed. Guido Wolswijk in Chapter 8, ‘Fiscal Aspects of Housing in Europe’, discusses the role of housing and mortgage markets fiscal instruments. A large degree of volatility in housing markets and mortgage markets may have disruptive effects on the economy and also impair financial stability. National governments have various tax and subsidy measures at their disposal that affect housing decisions. The effectiveness of these instruments has not changed much since the start of EMU as the volume of cross-border activities in housing remains limited. The chapter focuses on the euro area countries given the lack of national monetary policy options for them. This aspect potentially increases the value-added of national fiscal instruments in affecting housing market developments. The first part of the chapter describes the set-up of fiscal housing instruments in euro area countries as in the year 2007. It focuses on the taxation of the imputed rental value of the house, income tax deductibility of mortgage interest payments, capital gains taxes on housing gains, and VAT rates applying to new houses. While revealing some degree of diversity, in most countries the fiscal system overall financially favors owner-occupied housing compared with investment in other assets. This
Philip Arestis, Peter Mooslechner, Karin Wagner 9
preferential treatment is usually motivated by positive external effects of owner-occupied housing; it is also possible, though, that incomedistribution considerations and effects may also play a role. The chapter also mentions some possible drawbacks of a non-neutral government approach to housing markets, including the need to raise taxes on other activities to finance the preferential treatment of housing. The potential of fiscal instruments for preventing or correcting housing market imbalances is discussed in the second part of the chapter. Given that housing markets in the euro area are still predominantly national, fiscal instruments, unlike the euro area monetary policy, can be geared towards country-specific developments in housing and housing finance. The potential roles of such micro policy measures, however, do not always receive sufficient attention. Structural fiscal measures may contribute to reducing volatility on housing markets and in this way contribute to limiting busts that can have disruptive effects on the economy and on financial stability. Options for governments, for instance, include increasing the sensitivity of taxes to house prices (for instance, by updating more regularly the market value of houses, which often acts as tax base of various taxes), and reducing mortgage interest relief which homeowners receive via the tax deductibility of interest payments. Although there appear to be more arguments in favoring downsizing, the implicit subsidy given via mortgage interest tax relief and the capitalization of these benefits in house prices, make introducing or extending limits on interest deductibility politically difficult. Caution is needed in the use of fiscal fine-tuning measures, aimed at ad hoc correction of a housing market imbalance, in normal circumstances. Doubts prevail about possibilities for identifying situations when fiscal authorities should react; also in terms of the ability to provide adequate time and calibrate such actions, which may cause a government measure to have pro-cyclical effects. Precise modalities of fiscal measures, however, need to take into account the exact situation and the specific national housing market characteristics. József Hegedüs, in Chapter 9, ‘Towards a New Housing System in Transitional Countries: The Case of Hungary’, suggests that governments in the region worked under constant fiscal pressure caused by the social and economic costs of the bankrupted socialist economy. As a consequence, the state had to ‘withdraw’ from the housing sector thereby privatizing the housing sector in its entirety. The chapter uses the case of Hungary to show the emergence of the new housing model in the region. After the transitional recession of the 1990s, Hungary launched a new housing policy in 2000 focusing on three areas of the sector: Development of the
10
Housing Market Challenges in Europe and the United States
mortgage market, renovation of the existing housing stock (especially urban, multi-unit housing estates) and the rebirth of the social housing sector. Housing programs went through different stages under the influence of different pressure groups (construction industry, banking sector, local governments) and political forces (party experts group). The mortgage market was developed by competitive private banks, which were ready to service emerging middle-class families. The process was supported by a substantial interest rate subsidy, and as a result the share of outstanding housing loan to GDP increased from 2 per cent to 10 per cent in four years. Because of the high fiscal cost, the Hungarian government cut the interest rate subsidies in 2004, which, however, did not stop the growth of the mortgage market. The typically foreign-owned banks with easy access to the capital market provided low interest rate foreign currency loans (with high exchange risk and interest rate risk). From 2004, the share of the foreign currency loans increased from 20 per cent to 60 per cent, which made the country highly vulnerable to the economic downturn. The social housing program (construction subsidy for local governments to build or buy social rental units) was halted in 2004 because of the huge budgetary burden, and it was replaced with a rent-subsidy program that did not prove to be operational. The rehabilitation of the multi-family, prefabricated housing estates has proved to be the most successful program since 2004, because it is relatively cheap and politically effective (reaching a large number of families). Despite results of the housing programs (increased housing output, growth of the mortgage market, and increasing renovation in the housing estates), signs of the housing market crisis emerged, such as a fiscal pressure on the programs, uneven house price trends, growing inequality in housing conditions, lack of adequate social housing and the regressive subsidy system. The chapter argues that countries in transition followed the same trends as they faced the same housing market challenges after the transition, although differences in institutional solutions in housing finance and housing welfare schemes have become more and more important. The economic crisis of 2008 has reached the region in a transitional stage of development of the housing market. The expected economic hardship caused by the recession will affect the housing market drastically, such as house price decrease, increasing mortgage arrears, and difficulty to have access to mortgage. The housing policy response to the crisis will determine the future housing model in the region. In Chapter 10, ‘House Price and Other Housing Market Data: A User’s Perspective’, Anthony Murphy argues that there are data needs for policy makers and others who want to analyze and model housing markets.
Philip Arestis, Peter Mooslechner, Karin Wagner 11
A range of mix-adjusted and other house price indices are available for some countries, notably the UK and the USA. However, in many other countries house price data are limited. Disaggregated house price data for first time buyers are useful since credit conditions matter a great deal. In both the UK and the USA, one may infer changes in credit conditions from changes in the distribution of loan to value (LTV) and loan to income (LTI) ratios, as well as from surveys of loan officers. Mix-adjusted house price data are clearly preferable to average or median house price data. The construction of these mix-adjusted house price indices involves many choices. For example, one choice is whether mix-adjusted house prices indices based on one type or another of regressions are more accurate than another. Another question is whether the mix-adjusted results be grossed up to the stock of owner-occupied housing or the flow of house purchases. There is no single ideal index and, apart from very short-run forecasting, the choice of mix-adjustment procedure is generally a second order issue. Various UK and USA house price indices display the same medium and long-run trends, since they are cointegrated with each other. For modeling purposes, it does not matter much which UK or USA house price series are used. When modeling European housing markets, the important issue is not the choice of house price index but the shortage of long-run, disaggregated average and/or mixadjusted house prices data, as well as other housing related data. What data are needed depends on the models used. The most basic theory of what determines house prices is just a story of supply and demand, where the supply, the stock of houses, is given in the short run. Then prices depend on the stock of housing and the factors driving demand. In the USA, where rental markets are well developed and with rents generally market determined in contrast to the heavily regulated rental markets of the UK, the most popular model of house prices is the house price to rent model. When applying the inverted housing demand and house price to rent models to European housing markets, one is immediately confronted by large gaps in the house price and rent time series. At the macro level, one would like consistent, medium-run time series data on the housing stock (including vacancy rates) and the main components of household wealth. In addition, the distribution of LTV and LTI data for first time and other borrowers would be very useful for tracking changes in credit conditions. At the micro level, one would like survey measures of house price expectations and lending, as well as more information on housing finance and wealth in household and panel surveys. The chapter deals with both the micro and macro issues that relate to housing data.
12
Housing Market Challenges in Europe and the United States
In Chapter 11, ‘Residential Property Price Statistics for the Euro Area and the European Union’, Martin Eiglsperger presents the residential property price indices for the euro area and EU Member States compiled by the ECB in co-operation with the EU national central banks. Data on dwelling prices used for calculating these indices have been collected from various national sources, mostly from private or government statistical institutes, ministries, registers, real estate agencies and associations and mortgage lenders. Since 2001, the ECB has calculated an aggregate for the euro area by weighting together changes in prices for houses and flats for the euro area countries. Since then, the statistical quality of both the indicators for the countries and the euro area aggregate has improved, but still remain below the standards of other economic statistics and price indicators for the euro area. Dwelling price statistics for non-euro area EU countries have been collected since early 2005. This chapter starts with briefly introducing the work of the ECB and the EU national central banks in this area. This is followed by an outline of the relevance of statistics on residential property prices for ECB analyses, in particular with respect to housing wealth and the channels through which changes in house prices may impact on overall inflation. For these purposes, the ECB requires good quality statistics on house prices both on national levels and for the euro area as a whole. Breakdowns into indicators for urban and non-urban areas and for new and existing houses and flats can help identify different price dynamics. The presentation of the national sources focuses on the statistical features of the dwelling price data currently collected for euro area and EU countries. While for each EU country at least one indicator is available, the collected statistics differ in various statistical dimensions. National residential property price indicators do not only vary in terms of reporting frequency and timeliness at which new data become available. They also vary in the type of price information used in the index and the way the indicators control for the differences in the composition of the sample. For the calculation of an aggregated euro area indicator for residential property prices, the ECB identifies those national indicators which best represent price developments in the respective country and which are most appropriate for identifying pure price changes over time. This is followed by an explanation of how the weights of the euro area residential property price indicator have been derived and which aggregation formula is used for its compilation. While annual data of the euro area dwelling price index are provided from 1981 on, the semi-annual index starts in 1996 when sufficient country data on infra-annual price changes are available, including an interpolation of the annual price index for
Philip Arestis, Peter Mooslechner, Karin Wagner 13
Germany. Finally, the chapter provides an outline of possible future developments in the area of house price statistics. Most promising is an EU-wide project of the European Statistical System from which a substantial contribution to the improvement and the harmonization of EU residential property prices can be expected. Finally, in Chapter 12, Pirmin Fessler, Peter Mooslechner, Martin Schürz and Karin Wagner turn their attention to microeconomic data in ‘Housing and Financial Wealth in Austria: What Can Survey Data Tell Us for the Analysis of Financial Stability Issues?’. Microeconomic data on private households provide increasingly important information for many economic policy issues. This chapter examines how micro data can be used to improve the analysis of financial stability. At the beginning of the chapter the authors provide a brief survey of the recent literature that employs survey data on household finance and consumption with a focus on financial stability. The availability of micro data for understanding the impact of shocks, policies and institutional changes is particularly important in view of the extremely large heterogeneity in economic behavior of households. The recent financial crisis has demonstrated that a relatively small fraction of households can have important effects on market outcomes. The authors use data of two surveys, the 2008 survey on Austrian households’ housing wealth and the 2004 survey on financial wealth, both conducted by Oesterreichische Nationalbank. Housing assets are the most important form of collateral and the value of housing property affects households’ expenditure by improving access to credit for liquidity-constrained households. High-income households tend to have more debt than low-income households; however, the latter are more burdened by their debt. By combining both financial and housing wealth data, the chapter draws a better picture of households’ vulnerability and checks whether households’ debt in Austria represents a risk to financial stability. The dynamics of economic aggregates are determined not only by macroeconomic variables, but also by household-specific factors. This is particularly true for household consumption, savings and balance sheets, which are to a large extent driven by expectations about future individual income and demographic and social characteristics. The Austrian household-level data allows quantifying the size and relevance of the impact of shocks, policies and institutional changes on various groups of individuals. For example, financial integration and financial innovation made it easier for households to borrow against their future income, smooth consumption and diversify their portfolios. The resulting changes in the composition of the assets and their potential implications for wealth distribution, the
14
Housing Market Challenges in Europe and the United States
relative impact of policies on different household groups and the ultimate effect on macroeconomic variables can be judged with the micro data. Micro data allows the authors to analyze this structure, assessing the mismatch between assets and liabilities of households and identifying how many individuals have accumulated too much debt and what risks such over-accumulation poses to their finances and ultimately to the economy. The evidence shows that the structure of portfolios depends on age, wealth and household characteristics. A major factor in increasing stock market participation has been a surge in indirect holdings through financial intermediaries such as mutual funds and retirement accounts. Risk-taking still remains strongly correlated with wealth. Monitoring further changes in portfolio behavior is particularly relevant for an assessment of the impact of financial innovation. The recent financial turmoil has shown that numerous households and lenders underestimated the risks associated with high indebtedness. We would like to thank the authors for their stimulating contributions to the conference, and also the research department of the central bank of Austria for their support and sponsoring of the conference. Taiba Batool and Gemma Papageorgiou at Palgrave Macmillan, their staff, and Rita Schwarz from Editorial Processing at the Oesterreichische Nationalbank have been extremely supportive throughout the life of this project.
2 Housing Markets in Europe and in the USA: What Are the Relevant Issues Today? Peter Mooslechner and Karin Wagner
2.1 Introduction While housing markets and housing finance have undergone remarkable changes over the past decades, both in the USA and in Europe, housing continues to play an important role in the economy. In a significant number of recessions, the housing sector has preceded if not caused the economic downturn – a stylized fact – also underlined by the current financial turmoil and economic crisis. At the same time, there has been a significant increase in indebtedness of private households. In this chapter we take a closer look at some of the fundamental factors behind this trend of increasing mortgage growth. In particular, the chapter analyzes some of the recent changes observed in housing finance and studies key aspects of the economic relevance of housing markets in the present context. The chapter is organized as follows. First, we take a look at the relationship of housing markets to their national but also international macroeconomic surroundings (Section 2.2) and analyze how, for instance, house prices are linked to other macroeconomic variables. Thereafter (Section 2.3), we deal with some fiscal and regulation issues and Section 2.4 focuses on the tremendous changes in housing finance we have faced since the beginning of the 1990s. Central banks’ opportunities to react proactively to asset prices movements are the topics discussed in Section 2.5. We end with some policy conclusions that can be drawn in the light of the current economic crisis which – once again – originates from the housing market. 15
16
Housing Market Challenges in Europe and the United States
2.2 The housing sector, housing finance and the macro economy As the housing sector accounts for a considerable part of a country’s welfare, wealth and GDP, it significantly shapes a country’s long-term development. In this respect, it is important to pin down the underlying drivers of house prices, and to assess their economic implications, as well as the fact that house price developments diverge strongly across countries. In the following section we want to show some issues concerning the interaction of housing markets and the business cycle, especially the impact of house price developments through wealth channels on consumption. The question of how monetary policy mechanism can by interest rates changes affect house prices and wealth will be discussed later (Section 2.5). 2.2.1 Variation in house price developments and ownership rates While house prices have risen to record levels in Spain, UK and in Ireland, Austria and Germany show, if any, just slight increases in house prices (Table 2.1). These differences may reflect country-specific factors such as demographic differences, institutional regulations or cyclical positions; but changes in the structure of housing finance driven by globalization and liberalization are likely to play an important role as well. Above all, current macroeconomic analysis underlines the role of housing wealth as a household’s principal asset and the role of mortgage debt as the largest liability in households’ balance sheets. Housing tenancy structures differ considerably across European and US countries. While it is 43 per cent in Germany, it is 86 per cent in Spain and 75 per cent in Ireland (see Table 2.1). This can be explained by different tax incentives (see Section 2.3) and by differences in the access to mortgage financing. Home ownership rates were rising in almost all OECD-countries from the Second World War until the early 1990s. Against the fact that it is still out of reach for some income groups, owing one’s home has become less exclusive over time. It seems as if ownership rates reached a level and remained there, no additional increase has been seen since the 1990s (Atterhög and Song, 2004). The reached level depends on social attitudes to home ownership, legal and tax systems, etc. (Scanlon and Whitehead, 2004). In general, the late 1980s and early 1990s were periods when many governments cut back on subsidies and liberalization in mortgage markets and new legal arrangements and financial instruments were created. Furthermore, increased uncertainty took place, as labor markets and
Peter Mooslechner and Karin Wagner Table 2.1
17
Housing market indicators Real house prices year-on-year in %
Price-to- Price-torent income ratio ratio
2000–05 2006 2007 2
Level relative to long-term average1
United States 5.6 Germany –3.1 France 9.4 Italy 6.5 United Kingdom 9.8 Denmark 5.7 Finland 4.0 Ireland 7.9 The Netherlands 2.9 Norway 4.5 Spain 12.2 Sweden 6.0 Austria –0.8
4.5 –1.8 10.0 4.1 3.8 19.4 8.4 10.5 2.9 10.7 6.3 10.6 1.4
–0.3 –2.2 4.9 3.1 8.4 2.9 5.5 –1.8 2.6 11.5 2.6 8.6 1.5
123 71 159 127 151 162 146 167 156 158 187 160 –
102 64 138 114 141 143 105 133 158 121 147 120 –
Ownership Typical rates LTV in % in % 2007
68 43 80 70 70 54 58 75 54 77 86 50 57
2007
78 70 91 65 69 80 81 83 101 – 72.5 77 84
Note: House prices deflated by the Consumer Price Index. 1. Long-term average = 100, latest quarter available. 2. Average of available quarters where full year is not yet complete. Source: Girouard et al. (2006), OECD (2008), European Mortgage Federation, ECB (2009), Univerity of Technology Vienna.
governments were more exposed to competition through globalization and privatization. Tenure choice is determined by permanent income, the cost on owneroccupying versus renting (price–rent ratio) and by demographic variables. Households asking for a mortgage are often required equity contributions from borrowers/lending institutions. Thus, besides income, accumulated savings are important as a down-payment. They are an additional borrowing constraint and therefore also determine the timing of home purchase. In the 1990s down-payments were lowered in many countries (Duca and Rosenthal, 1994) – one of the changes in housing finance that increased competition between lenders (for further changes in the financing conditions for households see Section 2.4). Chiuri and Jappelli (2003) find that in countries with relatively high down-payment ratios ownership rates of the young are relatively low. Owning one’s home is typically considered as reaching the peak in the ‘housing career’. As mentioned earlier, differences in the access to mortgage markets and mortgage financing may be also one of the reasons
18
Housing Market Challenges in Europe and the United States
for the different ownership rates across countries. The share of homeowners of British and Dutch households aged 25–29 years is higher than in France, Italy, Germany or Spain (Catte et al., 2004, p. 26). However, countries showing high ownership rates such as Spain or Italy are not among those countries with the highest developed mortgage markets. This is to say, there seem to be other reasons responsible for that – besides tax regulations and mortgage markets. Bequests and other types of wealth transfers are considerably important in home purchases: e.g. more than a third of Italian households report that they got their home as a gift or bequest or that they got financial support when buying their home (Guiso and Jappelli, 1999). Scanlon and Whitehead (2004) conclude that younger households in most European countries tend to enter homeownership at a later period in their life than in former years because of higher house prices and greater access problems. Chiuri and Jappelli (2003), using micro data of 14 OECD countries, show that ownership rates vary significantly across countries. Whereas first homes are typically bought in the 20s in Australia, Finland, Sweden, UK, Canada and US, in other countries such as Austria, Spain, Italy people buy their first homes between 30 and 40 years old. Homeownership rates and, therefore, the size of the rental market, differ across Europe and the US (Table 2.1). The share of rented dwellings has decreased since the 1980s in most European countries (ECB, 2003). Nevertheless, in some countries the proportion of the private rented sector is still high. Additionally, not all homeowners live in their homes. Furthermore, the size of social rented homes and cooperatives play a crucial role and can be highly influenced by (national) policy-makers. According to a survey, German households own about 75 per cent of all residential property – but more than 30 per cent of all housing is rented out by private individuals to other households. 18 per cent of all housing is rented out by private enterprises, including cooperatives. Contrary, in the Netherlands private (and enterprise-owned) rental housing plays a minor role (just 10 per cent of dwellings to be rented), social housing is important (over 30 per cent). In Italy, about a quarter of the total rental housing stock is owned by the public. In France and Finland around 20 per cent is social renting by the government or public enterprises (ECB, 2009). 2.2.2 House price movements as a major cause of wealth effects Why are developments of house prices so important for the economy and how can house price movements influence other macroeconomic
Peter Mooslechner and Karin Wagner
19
variables and economic activity? House price increases or decreases have to be seen in the context of a country’s business cycle as housing markets tend to track the country’s business cycle. There are wealth effects on consumption from changes in house prices. The life cycle model (Ando and Modigliani, 1963) with consumption depending on a household’s lifetime income and wealth suggests that consumers will distribute increases in anticipated wealth over time and the marginal propensity to consume (MPC) out of all wealth – from any form of wealth – should be the same number, something just over the real interest rate. But one has to take into account that for those having a bequest motive the consumption effect derived from housing wealth may be smaller than that derived from other forms of assets. To assess the strength of the links between residential prices and consumption Catte et al. (2004) calculate the longrun marginal propensity to consume out of financial wealth to be in the range from 0.01 in Italy to 0.07 in Japan. They estimate the OECD average to be 0.035 and for the US their estimate is 0.03. In general, wealth effects from the real estate market are not of the same magnitude as wealth effects for the stock market. Consumption is affected differently by the form in which wealth is held for several reasons. An obvious point is that while households get information on changes in their stock market wealth easily by various media (internet, TV, newspapers, etc.), they may be less aware of changes in the value of their housing wealth. A number of empirical studies have analyzed the impact of housing wealth on consumption in the United States (for a survey, see Altissimo et al., 2005). The results point to a marginal propensity to consume between 2 and 9 cents out of each dollar of nominal housing wealth. Case et al. (2005) compare stock market wealth effects versus housing market wealth effects using data for a panel of 14 countries (Canada, the US and 12 European countries) over the period 1975 to 1996. One of the econometric models they apply shows that a 10 per cent increase in housing wealth leads to a growth of consumption by roughly 1.1 per cent. Contrary, a 10 per cent increase in stock market wealth has almost no effect on consumption. Results for a panel of US states tell that a 10 per cent increase in housing wealth and in stock market wealth has the same effect on consumption – both cause an increase of 0.4 per cent. However, all the econometric specification presented supports the conclusion that changes in house prices have a larger effect than changes in stock market prices to influence households’ consumption in Europe and in the US. Muellbauer (2007) suggests that the broader access to credit had important implications on consumer behavior in many countries. Credit
20
Housing Market Challenges in Europe and the United States
market liberalization raised consumption-to-income ratios and reduced consumer savings by increasing the collateral value of housing wealth. According to Muellbauer, this is the main reason for the increase in the size of housing wealth effects. The second reason is that house price changes now have larger effects on consumer consumption than changes in stock prices. Furthermore, house price movements may have led to income effects through the rental market (Giuliodori, 2004). With deregulated renting systems, higher house prices my cause higher rental prices. This income effect is positive for landlords or institutional investors while it is negative for tenants. As mentioned earlier, the size of the rental sector differs across countries. Therefore, the higher the share of owner-occupiers and the lower the rental market’s size in a country, the larger the house price effect will be on income and on consumption decisions of households. 2.2.3 Housing prices and the business cycle In many recessions the housing sector preceded overall economic downturn, even if the housing markets’ share in GDP is too small to cause a recession by itself. Changes in house prices generate not only substantial wealth effects on consumption, but also – as evidenced by the current financial crisis – significant spillover effects, which can cause severe problems for the macro economy as a whole. For example, booms in housing investment were responsible for increased employment, as the construction sector accounted for more than 20 per cent of all employment gains since 2000 in the US, France, Spain, Denmark, Norway, Sweden, Ireland and Greece (OECD, 2007). In the past the volume of residential investment tended to turn prior to house prices in a country’s business cycle. Over recent decades housing investment has shown high growth rates in many countries. Low interest rates have been one of the driving factors as they stimulate the demand for housing, which leads to higher house prices and, in turn, stimulate residential investment. If house prices rise faster than construction costs it makes sense for individuals or construction companies to invest in new dwellings. The extent of this supply effect of increased house prices by residential construction and housing investment can differ depending on the national construction regulations, availability of specialized workers, etc. (Giuliodori, 2004).1 Many countries which had reached a ten-year peak in housing investment growth (Figure 2.1, left panel) had experienced rapid population growth, especially from immigration. While housing investment was on the rise till 2007, in 2008 a downturn started in almost all countries (Figure 2.1, right panel).
Decline since end of 2007
Ten-year peak
year-on year in %
in % of GDP
10
12
5 10
0 5
8
10 15
6
20 4
25 30
2
Ø1998–2008
to 2007q4
Ireland
UK
Spain
Norway
France
Denmark
Italy
Finland
Sweden
Belgium
Austria
Germany
Ireland
UK
Spain
Norway
France
Denmark
Italy
Finland
Sweden
2008
to 2008q4
Real housing investment
Source: OeNB, OECD Economic Outlook 84 database.
10.1057/9780230246980 - Housing Market Challenges in Europe and the United States, Edited by Philip Arestis, Peter Mooslechner and Karin Wagner
21
Figure 2.1
Belgium
Austria
Germany
Netherlands
USA
2007
Netherlands
40
0
USA
35
22
Housing Market Challenges in Europe and the United States
Table 2.2
Timing of maximum correlation Output gap contemporaneous or lagged < 1 year
Output gap lagged 1–2 years
Output gap lagged 3–4 years
Strong
Denmark, Finland, Ireland, United Kingdom
Spain
Average
Japan
Canada, France, Sweden
Australia, Germany, Switzerland
Weak
New Zealand
Norway, United States
Belgium, Italy, Netherlands
Intensity of correlation
Note: Correlations are between de-trended real-house price levels and the output gap. They are calculated for the period 1970–2002, based on semi-annual data. Countries are ranked according to the value of the maximum correlations and of the lags at which these are found. The intensity of correlation is indicated as strong if the maximum correlation coefficient is >0.65, average if 0.50–0.65, weak if <0.50. Source: OECD (2004).
But how closely are house price developments related to output developments over the cycle? Lags between output fluctuations and real house prices vary considerably both from one cycle to another and between countries (Table 2.2). These lags determine the degree to which house price movements are pro- or counter-cyclical (OECD, 2004). 2.2.4 House prices and mortgage markets MPC out of housing wealth on consumption is positively correlated with mortgage markets’ size and structure. In principle, housing is characterized as a local good and it depends on national regulations. Furthermore, institutional arrangements in housing and mortgage markets play an important role not only for overall economic development but also for the propagation of shocks. They influence the magnitude and speed with which monetary policy responses (Section 2.5) to shocks transmitted through economies. And they have an impact on the characteristics of mortgage markets, determining the extent of housing wealth effects on households’ consumption. The influence of housing wealth on consumption depends on the extent to which homeowners are able to borrow against housing assets through mortgage equity withdrawal (by taking advantage of low refinancing rates and increased house values) and on the level of mortgage debt across countries. Thus, these regulatory
Peter Mooslechner and Karin Wagner
23
and institutional features of mortgage markets (the size of required LTVs, valuation methods applied by the banks, etc.) can soften or tighten the transmission effect of house price developments to credit availability for households or investors. Consumption responses to changes in housing wealth are expected to be higher in financial markets with easier access to mortgage financing. Mortgage markets’ variety is an additional factor when discussing house price and wealth effects on consumption. If mortgage markets are competitive, they offer a greater variety of mortgage products. Tsatsaronis and Zhu (2004) found that the impact of interest rates on house prices is stronger and faster with more developed mortgage markets. This in turn facilitates equity withdrawal. Besides the United States, countries with well-developed mortgage markets are the UK, Canada, Australia and the Netherlands (OECD, 2008). In these countries consumption is most strongly correlated with house prices (Catte et al., 2004). In the US and in the UK, weaker house prices and tightening lending standards have already led to substantial falls in housing equity withdrawal (OECD, 2008). The composition of mortgage loans in a market – being predominantly variable-rated of fixed rated – and the size of the respective share prevailing in a country also plays a role when discussing the impact on consumption (see Section 2.5). The current economic crisis underlines the tremendous effects which can be caused by housing markets. Although the US mortgage market is quite small, the developments there stood at the origin of the financial turmoil in summer/fall 2007. How could this happen? Subprime loans are characterized by the fact that these mortgages are given to households with poor credit ratings. Over the past years, increased product variety in mortgage markets enabled liquidity-constraint borrowers to get access to loans. Given their higher default risk, these subprime households were charged higher interest rates and many subprime loans incorporated a ‘teaser’ component (with an artificially low initial rate over the first one or two years) or there were ‘balloon’ loans (with only interest to be paid for a number of years). Fuelled by these elements, the US subprime mortgage market has been growing rapidly since 2003, accounting for 14 per cent of the total value of the total value of the mortgage market in 2006 and about 20 per cent of new mortgages issued in 2006. The securitization market enabled to pool subprime loans in residential mortgage-backed securities (RMBS) and by this housing finance got international (Section 2.4). With slowing house price growth and the end of the fixed-rate period of teaser rates in 2007, the delinquency rate
24
Housing Market Challenges in Europe and the United States
of subprime loans with adjustable rates rose to 21 per cent at the beginning of 2008 compared to 10 per cent at the beginning of 2004 (Source: Mortgage Bankers Association, National Delinquency Survey). Spillover effects into other sectors of financial markets and to investment banking occurred because of links to the RMBS market or credit lines to subprime borrowers. In summer/fall 2008 all standalone investment banks were converted to commercial (depository) banks under the Gramm–Leach–Bliley Act. The historical ‘advantages’ applied by investment banks in the past – the avoidance of safety net regulation and the ability of having access to repos as substitute for deposit funding – were no longer existing and a quite small segment of housing finance in the US caused a world-wide recession.
2.3 Fiscal and regulation aspects in the housing market Regulation of housing markets smoothes fluctuations in housing markets – one of the most important arguments to foster regulation efforts. There are quite different regulations between countries on the tax deductibility of interest rates and on the amount and way allowances and subsidies are given to households. Owner-occupied ratios of euro area countries vary widely. Only in a few countries (Cyprus, Spain, Greece, Ireland, Slovenia) the ownership ratio is above 75 per cent. An extreme outlier is Germany with a ratio of 43 per cent (2007), in Austria, France, Finland and the Netherlands it is in the range between 55 per cent and 60 per cent. An owner-occupied dwelling can theoretically be treated either as consumption good or as an investment good. Regarded as an investment good, it would require that imputed rent is taxed as income and mortgage interest repayments are tax deductible. If it is seen as a consumption good, neither taxation of imputed rent nor mortgage interest tax relief should be in place (Haffner, 2003). Britain, France and Germany are examples of the consumption good approach, while the Netherlands, Belgium and Norway are examples of the investment good approach. Exceptions are Finland, Sweden and Denmark: These countries give mortgage interest tax relief still in spite of the fact that imputed rent taxation was abolished in the early 1990s (Boelhouwer et al., 2004). There exist various policy measures to encourage housing investment and households’ affordability. Institutional regulations, subsidized mortgage loans and mortgage interest rate deductions from income tax are some of the examples of smoothing volatility in housing markets.
Peter Mooslechner and Karin Wagner
25
Examples of individual aid are capital grants or subsidies/allowances to reduce mortgage interest expenses (e.g. deductibility from personal income). Further state interventions in the housing market are the construction of new flats/houses or support to subsidized dwellings. Furthermore, differences within Europe and the US exist concerning the taxation of imputed rent of owner-occupied dwellings. Taxes can be categorized as: • direct taxes related to the underlying property, • direct taxes related to taxpayers’ net household income, or • indirect taxes in the form of one-off measures (transaction costs or
fees for a property transfer or inheritance). 2.3.1 Taxation regulation changed over time Nowadays, only few countries impose a tax on imputed rent for owneroccupied housing with the valuation base being usually lower than the market value. In most European countries and in the US, mortgage interest payments from primary residences are tax-deductable. In many countries the system of tax deductibility changed over time: e.g. in contrast to the 1990s, for a short time there is no more distinction between interests and capital redemptions (Belgium) or fiscal deductibility is no more related to income. This measure helps young people investing earlier in the housing sector (Belgium, Finland, Spain) or the allowances are subject to income ceilings excluding high-income households (as in Germany 1992 till 1996). All in all, the importance of tax deductibility of interest payments has decreased over time. In many countries the effective marginal tax rate at which a mortgage interest tax relief can be claimed was reduced over time (e.g. in Ireland from 47 per cent to 20 per cent). A further possibility to support home ownership consists in loan guarantees by the state, facilitating the provision of mortgage loans for home-buyers by reducing the need for personal guarantees (for example, in Finland). In some countries indirect measures were imposed to curb demand by investors for residential property (e.g. in Ireland via stamp duties). There are differences in the tax treatment of acquisitions of dwellings – while in Germany subsidies for the purchase of used dwellings were curtailed in comparison with those for the purchase of new buildings, some countries (e.g. Belgium) changed from favoring new dwellings in comparison to existing ones and nowadays there is an equal tax treatment independent of the date of building. Most countries abandoned inheritance taxes and wealth taxes over the last decade. Withdrawal of state funding, shifts and reductions in subsidies, changes
26
Housing Market Challenges in Europe and the United States
in regulations and supervision, but also an increasing reliance on market processes are recent trends in the housing subsidy policies. Furthermore, transaction costs play a crucial role when buying property (ECB, 2009). Tax deductibility of mortgage interest payments is often claimed to stimulate owner-occupied housing. Many studies in the field deal with the US tax system, which is known for its generosity towards homeowners. Swank et al. (2002) took a closer look on the tax regulations in the Netherlands, where the preferential tax treatment of homeowners is often said to be comparable with US practice. They found that both starters and movers benefit from mortgage interest deduction for higher income groups. However, such tax favoring tends to facilitate house price bubbles, especially when interest rates are low and LTV-ratios are high. They conclude that the efficiency of implicit tax subsidies to homeowners depends heavily on the price responsiveness of newly built dwellings. Matsaganis and Flevotomou (2007) show the distributional effect of mortgage interest tax relief to be regressive. Replacing mortgage interest tax relief (and, at the same time, housing benefits) by a tenureneutral universal housing transfer would have redistribution effects of resources from rich to poor. The appropriate use of taxation incentives as a policy instrument in generating demand-led residential development is a big challenge for policy-makers. Furthermore, it is evident that fiscal aspects of mortgage financing are predominantly country-specific and play an important role when discussing housing market development in the euro area.
2.4 Effects of housing finance patterns on housing markets and on households’ wealth position 2.4.1 Increased wealth goes in line with indebtedness of households Housing wealth is the important part of the net worth of the household sector, and loans for house purchase are the main category on the liability side. Contrary to financial assets which are held mainly by the better-off parts of the population, housing wealth is more equally distributed across households. Housing wealth can enable credit access for liquidity-constrained households which otherwise would have no access to uncollateralized consumer loans. Housing assets are the most important form of collateral for private households. As such, pronounced price fluctuations in house prices are transmitted directly to households’ net wealth with implications on households’ expenditure and repayment capacity of their debts. In the
Peter Mooslechner and Karin Wagner Euro Area
USA 1992 100
1992 100
400
400
350
350
300
300
250
250
200
200
150
150
100 1993
27
1996
1999
Mortgage debt
2002
2005
100 1993
1996
1999
2002
2005
Disposible income
Figure 2.2 Mortgage debt outstanding and disposable income Source: ECB, OECD, BEA, Thomson Reuters.
late 1980s and early 1990s, most European countries faced strong increases in house prices. In that period, monetary policy intervention to curb inflationary pressures, coupled with the subsequent recession, was followed by a pronounced slowdown in house prices. The envisaged expansion of mortgages went hand in hand with a strong rise in the number of households with a mortgage loan and with a modest rise in the average amount of mortgage per indebted household. Rising disposable incomes of households over the last ten years gave households the possibility to take on higher loan amounts. Larger loan amounts were also facilitated by longer maturities of mortgage loans. While in the1990s the average maturity of mortgage loans was roughly 10 years, some countries lengthened the average maturity of new housing loans during the 1990s to over 20 years. Furthermore, more flexibility in repayment schedules was introduced. In addition, LTVs (loan-to-value ratios) increased in most European countries. Figure 2.2 presents the evolution of the outstanding mortgage debt and of disposable income of private households. Obviously, there is a constant increase in the outstanding mortgage debt and the growth has clearly accelerated since 2000. The divergence between mortgage and income has increased over past years – and even more in the euro area than in the US. Nevertheless, it is important to analyze not only the debt situation of households but also to take a closer look at interest payments as indicator of the vulnerability of households. However, while indebtedness of households is at record levels, the share of interest expenses in
28
Housing Market Challenges in Europe and the United States
households’ disposable income is quite low and much lower than at the beginning of the 1990s. At the end of the 1990s and until the end of 2005, short-term nominal interest rates remained at very low levels by historical standards (with the exception of a short-lived period in between mid-1999 and the end of 2001), contributing to an increasingly broadly-based edging up of households’ indebtedness. Since 2005, households’ interest burden generally increased again. A closer look into the debt position of households is provided by micro data of private households’ surveys. Demographic developments and migration flows also drive mortgage demand; with Spain and Ireland as the countries where demographic factors may have contributed most to the strong increases in house purchases (ECB, 2009). Against the background of increased indebtedness we have lately seen a decrease in the mortgage spread (the spread between interest rates on loans and the cost of funding for Monetary Financial Institutions (MFIs), which may be due to increased competition. And, inter alia, increased competition in the housing (and mortgage) markets over the last years led to the introduction of new mortgage products that enabled borrowers to take a highly geared position. Mortgage loans are primarily used for housing investment. But private households also take on debt for consumer goods or other purposes based on historical or cultural reasons. While in the European countries taking out a mortgage for purposes other than buying a house (instead e.g. using it for consumption purposes) is not a widespread phenomenon, in the US a higher share of mortgages was taken out for non-housing purposes. Despite the liberalization and deregulation process in the 1980s and the 1990s, which influenced the lending practices in mortgage finance markets in Europe and in the US heavily, the household mortgage debt in per cent of GDP differs greatly among Europe and the US (Figure 2.3) as well as concerning the structure of mortgage loans. In the US long-term fixed-rate mortgages prevail, while other countries prefer variable-rated loans. However, the share of variable rate loans varies within Europe also (10 per cent to 99 per cent in 2007, see ECB, 2009). In the US mortgages with an introductory period of low interest are widespread, which may cause problems at a later stage. Overall, Figure 2.3 shows that there is no systematic correlation between the share of variable-rate loans and mortgage debt to GDP ratio. Refinancing practices of banks may play a role regarding the share of fixed versus variable rate loans. In addition, institutional (e.g. by posing higher or lower fees for early repayment, foreclosure procedures) and fiscal regulations (subsidization of owner-occupied housing, see
Peter Mooslechner and Karin Wagner
29
in % 100 90 80 70 60 50 40 30 20 10 0
BE
DE
IE
GR
ES
FR
IT
LU
MT
NL
AT
PT
FI
US
Share of variable-rate lending in new loans for house purchase1 Residential mortgage debt to GDP Ratio
Figure 2.3 Share of variable rate lending in new loans for house purchase and mortgage debt to GDP in 2007 Note: 1. Data for US not available. Source: ECB, European Mortgage Federation, Federal Reserve Board.
Section 2.3) are important as well. In some European countries (e.g. Austria, Slovenia) it was quite popular to hold mortgage loans denominated in foreign currency (mainly in Swiss francs). But such loans bear a couple of additional risks which turned out to become pretty important in the course of the financial crisis.2 2.4.2 Changes in housing finance gave an international dimension to housing markets In parallel to the increased indebtedness of households, a tendency away from traditional methods of financing towards more market-based systems of housing finance has become prominent.3 The funding of housing loans – still largely provided by banks in Europe – has changed significantly. New funding instruments with longer maturities (e.g. issuance of mortgage covered bonds) developed over the past 10 years. In particular, we have seen a boom in securitization: Individual mortgage loans are pooled and used as collateral and are sold to investors. Instruments like mortgagebacked securities or special purpose entities (or special purpose vehicles) were created. Traditionally, financial intermediaries have distributed directly their financial products and funded the mortgages internally by themselves. With the new structure of mortgage markets, borrowers do not directly meet the lender. A mortgage broker is in between distributing
30
Housing Market Challenges in Europe and the United States
the mortgage products to many lenders who securitize them and give them as asset-backed securities to a special purpose vehicle. Within this process of securitization mortgage is increasingly financed by institutional investors as pension and mutual funds from all around the world. Thereby, housing financing has become international. As a result, fluctuations spill over much more easily from one ‘national’ market to others or to – international – capital markets, creating a particular challenge for supervisors and for central banks. 2.4.3 Loosening of credit standards The rising use of credit scoring in household lending and other changes in lending practices has allowed borrowers to get access to mortgage products who previously could not apply for loans. At the same time, there developed a tendency for a loosening of credit standards. But stricter loan conditions (e.g. lower LTVs) in Europe resulted in a much smaller share of subprime loans than in the US. Differences in accounting and supervisory practices may have also been responsible for this and, in addition, mortgage equity withdrawal plays a more prominent role in the US housing market than in Europe. The development of securitization markets has been facilitated by various newly-created legislative and institutional regulations and laws on the national and international level. Rising delinquencies of subprime residential mortgages in the US led in late 2007 and 2008 to substantial losses to holders of securities backed by those mortgages and to sharp increases in credit spreads for those securities. Mortgage companies specializing in subprime products experienced funding shortages and failed. In general, the role of securitization in loosening credit standards in Europe was far less significant than in the US.
2.5 Housing markets and monetary policy Empirical literature finds that house prices are affected by interest rate changes. Economies have repeatedly been subject to asset price bubbles; sharp upturns in asset prices that depart from fundamental values eventually crash. Once again, in the midst of the current global financial crisis, both the fragility of the financial system as well as the pronounced weakness in real activity called for policy action. Against the background of an increased share of alternative funding institutions such as investment banks engaged in banking practices outside the traditional banking (shadow banking) and insurance companies (see Section 2.4), there was a call for rapid actions of (traditional) central banks. In that situation the
Peter Mooslechner and Karin Wagner
31
functioning of traditional monetary transmission policy channels was even more essential. As mortgage markets are internationally linked (see Section 2.4.2), consorted actions of the major banks around the world were necessary. In these days, central banks responded very quickly by easing monetary policy markedly and providing almost unlimited liquidity. Today’s crisis, although not comparable in any respect to any previous one since the Great Depression, is once again characterized by unusual corrections in asset prices, covering a broad range of asset categories from stock markets to real estate. In general, the influences of asset prices on demand and inflation through traditional wealth and cost-of-capital channels (as discussed in Section 2.2) fall directly within the concerns of monetary policy. However, asset price bubbles are not alike. Not all asset price bubbles create risks to the financial system, but the present subprime crisis and falling property prices have put central banks under pressure for economic policy as well as for financial stability reasons. Obviously, previous asset price booms – in particular those affecting housing markets – have triggered many periods of financial instability as well as severe economic downturns.4 Financial history reveals the following typical chain of events. Triggered by exuberant expectations a credit boom starts, which increases the demand for certain assets thereby raising their prices. Rising asset prices make borrowing based on these assets more attractive, thereby increasing demand and hence their values further. This feedback loop generates a bubble relying on expectations of further expected rises in asset prices. At some uncertain point in time the bubble bursts. The following collapse in asset values turns the feedback loop around into a downward spiral: Loans default, lenders restrict credit supply, asset demand declines, fire sales of assets take place – and prices collapse. The resulting devaluations erode the balance sheets of financial institutions, in the worst case distributing the negative effects across a broad range of assets. In the end, this weakens economic activity and increases macroeconomic risk, and, in the extreme, can challenge the functioning of the entire financial system. Although we have learned much about the effects of asset price cycles on the economy out of a wide range of boom and bust episodes of all kinds, once again the particular sequence of the current crisis came as a surprise. This underlines that many open questions remain on the optimal policy reactions associated with such cycles; many of them relate to monetary policy strategy and implementation, especially whether there exists a rationale for a central bank to react to asset price developments.
32
Housing Market Challenges in Europe and the United States
The most recent crisis episode may be interpreted to provide additional evidence that central banks should react proactively to asset prices movements, and do so ‘over and beyond’ what these imply for aggregate demand and inflation.5 Of course, conducting monetary policy in this way is not easy. In addition to the fact that the central bank must form a view of whether a particular asset price increase is dangerous or not, it requires monetary policy to have predictable effects on asset prices. Furthermore, the size of interest rate movements required to prevent a bubble from developing must not be so large as to cause output and inflation to fall substantially below the central bank’s objectives.6 2.5.1 Conflicting policy approaches There are two different strategies for monetary policy to deal with an asset price bubble. A central bank following a traditional monetary policy strategy does not attempt to influence the speculative part of asset prices. For quite some time the central bankers’ traditional answer to the question was ‘no’. Asset prices react to all kinds of shocks and monetary policy should react directly to these shocks and not to asset prices themselves. Price stability will be good for financial stability in general and asset price stability in particular, as price stability helps reduce misperceptions about future returns and reduces the problem of asymmetric information. Therefore, the central bank responds to asset price movements, only to the extent that those have an impact on future output and inflation. However, there are arguments for a more activist policy. Central banks were urged to take extra action by tightening policy, to limit the consequences of the bubble. This might be justified if three conditions were met. First, policy-makers must be able to identify bubbles in a timely fashion. Second, tighter monetary policy must have a high probability to dampen speculative activity. And third, the expected improvement in future economic performance must be sufficiently great. Even if the likelihood of meeting these conditions seems to be low, in the aftermath of the bursting of the housing bubble the severity of the crisis might seem to question this judgment. It is important to demonstrate the benefits of a more activist, asset price oriented monetary policy as there are also implied risks. Identifying the underlying source of an asset price shock along with the difficulty to predict the transmission mechanism from asset prices to the real economy tend to be rather complicated. But if the detrimental effects on the real economy in case of a bust are significantly larger than the positive effects of the boom, the optimal policy for a central bank is to ‘lean against the
Peter Mooslechner and Karin Wagner
33
growing asset price bubble’ in order to contain the bubble and so reduce the costs when it bursts. Whether the uncertainty associated with any attempt to derive the fundamental value of a financial asset is a serious argument against a more pre-emptive monetary policy remains open. Some very simple indicators of bubbles obviously exist. And other concepts on which we base monetary policy – like the output gap – are also difficult to measure. Trying to cope with the problems instead of ignoring asset prices seems to be justified, although the degree of uncertainty for estimating misalignments is likely to be high. How does all this relate to the current financial and economic situation? It remains controversial whether a tighter monetary policy stance from 2003 onwards could have prevented the housing bubble. Greenspan (2002) summarized the then prevailing view, that central banks should follow a ‘strategy of addressing the bubble’s consequences rather than the bubble itself’. Borio and Lowe (2002) conclude that bubbles can be identified in real time if central banks expand their view beyond asset prices to include other variables that signal a threat to financial stability. For example, they find that episodes of lasting rapid credit expansion, booming stock or house prices, and high levels of investment are almost always followed by periods of stress in the financial system.
2.5.2 The challenge of changing financing structures A particular – often neglected – aspect of increasing importance is that the market structure of housing finance (as described Section 2.4) affects the impact monetary policy will have on the economy.7 As mortgage debt forms the dominant part of households’ total debt, conditions in mortgage markets are an essential component of the transmission mechanism. An increase in monetary policy rates will be transmitted to interest rates applied on mortgages (interest rate channel) and will very likely also reduce the supply of loans (credit channel) via a number of factors. But the overall effects on the efficacy of monetary policy transmission remain theoretically ambiguous. They are likely to be asymmetric also, as more debt makes households more vulnerable to potential credit supply constraints. This may result in monetary policy measures having more impact on disposable income and, hence, consumption. At the same time, mortgage loan characteristics and crosscountry differences create a significant heterogeneity in transmission, for example, based on differences in the share of variable rate loans (see Section 2.4).
34
Housing Market Challenges in Europe and the United States
It is difficult to assess the empirical relevance of all changes in housing finance in a comprehensive manner, in particular when taking into account the links between housing and financial markets in a period of crisis. Therefore, it comes as no surprise, that empirical studies show conflicting results: Some appear to confirm a reduced monetary policy effect in the United States due to financial innovations, several point towards a stronger pass-through to market interest rates because of increased competition (see Section 2.2). A number of studies claim that the monetary policy transmission mechanism in the euro area has not changed significantly over the past decades; at the same time monetary policy effects tend to be higher in countries with more developed mortgage markets.8 Reviewing the available empirical evidence, it seems too early to draw firm conclusions on the effects on monetary policy transmission. It seems very likely that during periods of economic growth and positive expectations the ability of monetary policy to moderate the expansion would have been reduced as a result of the greater flexibility of mortgage funding. But even this might be questioned given the fundamental changes the current crisis may impact on the structure of mortgage financing. Against this background, several important challenges arise with respect to the design of monetary policy. Highlighted by the recent financial crisis, the questions of how monetary policy can best be conducted to minimize the risks of a pro-cyclical behavior of credit, ‘leaning against the wind’ of growing financial imbalances; what the optimal monetary policy reaction is in the case of financial distress and busts; and, more generally, how to ensure a symmetric monetary policy reaction to booms and busts can now be addressed. There is not yet a definite answer as to how central banks should best deal with asset prices. At the same time, this issue will become even more important in the future. Therefore, monetary policy cannot expect to escape easily from this fundamental challenge. Because of their perceived severe macroeconomic effects, asset price movements will remain a central concern for monetary policy-makers. This is and will be the case not only because of monetary policy considerations, as beyond the setting of short-term nominal interest rates, a broader view of monetary policy includes regulatory aspects of financial stability. Whenever monetary policy is seen as an instrument too blunt to be used to target asset prices, the combination of monetary policy, financial stability considerations and supervision of financial institutions might be the potentially more successful avenue to follow. An obvious other lesson of the current crisis is that there is an urgent need to improve our understanding of the workings of the financial system,
Peter Mooslechner and Karin Wagner
35
its vulnerabilities, and its links to the real economy. It is of crucial importance to identify more quickly if financial innovation and other factors are leading to a build-up of destabilizing forces, such as rapidly rising asset prices or excessive leverage. In this context, a thorough review of the regulatory structure of financial systems (including hedge funds and rating agencies) is unavoidable. Some of these issues are outside the direct responsibility of monetary policy but within the policy concerns of the broader regulatory framework governing our financial systems. Central banks, therefore, cannot escape from addressing these issues, even if it may be too early to draw sufficient robust lessons for monetary policy in its narrow sense.
2.6 Conclusions This chapter tries to structure some of the most relevant issues regarding housing markets and housing finance; and it discusses some of the macroeconomic effects that can cause or at least are amplified by housing markets. The current crisis underlines the multiplication factors of housing markets and the interdependences with other sectors and markets of the economy. Against the background of rising house prices and rising wealth of private households, mortgage markets have faced quite significant changes over the past decade. Besides the fact that – encouraged by increased globalization of financial and capital markets – housing markets and mortgage financing have become more international, new products were introduced which enabled not only just the well-off private households to get access to their own home. In particular, these new products significantly contributed to new trends in home financing: Formerly, mortgage markets for home financing were traditionally dominated by institutions taking deposits as the main source for loan financing. Residential real estate markets have benefited over the past decade from an increased reliance on marketbased forms of financing and a variety of newly invented mortgage products, namely tradable instruments and securitization of mortgage assets. In particular, securitization changed the mortgage environment significantly for borrowers and lenders and financial liberalization and technological innovations favored this process. Households, in principle, benefited from this new environment in mortgage markets by lower transaction costs, more flexibility on mortgage contracts and increased access to loans for home purchase. The phenomenon of securitization brought new challenges for regulations and accounting practices. What can be seen clearly in the light of the current financial and economic crisis in this respect is the need
36
Housing Market Challenges in Europe and the United States
for better supervisory structures. As many investors have been hit by tremendous losses over the past months it may be expected that complex financial assets will become much less popular and the funding from this channel will become scarce. But even if some mortgage products reduce or even disappear, the current crisis has illustrated the dangerous spillover effects that can result from developments within housing finance and mortgage markets. Against this background and highlighted by the recent financial crisis, several important challenges arise with respect to the design of monetary policy. Topics to be discussed are (i) how monetary policy can be best conducted to minimize the risks of a pro-cyclical behavior of credit, (ii) the appropriate policy strategy of ‘leaning against the wind’ of growing financial imbalances, (iii) the question which monetary policy reaction is the optimal one in case of financial distress and busts; and, more generally, (iv) how such a symmetric monetary policy reaction can be ensured. There is not yet a definite answer how central banks should best deal with asset prices. At the same time, this issue will become even more important in the forthcoming future. Therefore, monetary policy will have to face these fundamental challenges.
Notes 1. Ball (2002) and ECB (2003) review some cross-country papers, indicating that – in the short run – a relatively low price elasticity exits in Europe compared to the US. 2. Foreign currency loans are regularly granted with a single, fixed maturity date. For the purpose of repaying the loan amount at the end of the term, a repayment vehicle is set up. Loans with repayment vehicles entail a higher risk than regular loans since it is always hard to forecast the repayment vehicles’ earnings trends. 3. The greatest share of the stock of mortgages to households is originated by MFIs (in the euro area about 90% (ECB, 2009)). However, non-MFIs such as pension funds and/or insurance companies are also involved in home financing of households and, furthermore, other financial intermediaries and public entities provide mortgage loans too. Initially, in US the investment banks were created as counterparts to commercial banks. They were allowed to take deposits but there was a stronger regulatory control foreseen compared with commercial banks. The variety of funding instruments and the share of alternative funding institutions engaged in households’ home financing increased over the last decade. 4. See Ahearne et al. (2005) and Goodhart and Hofmann (2007). 5. See Borio and Lowe (2002) and Cecchetti et al. (2000). 6. See Bean (2004), Bernanke (2002) and Kohn (2006). 7. Many studies (e.g. CGFS, 2006 and IMF, 2008) have investigated the consequences of recent changes in housing finance for monetary policy.
Peter Mooslechner and Karin Wagner
37
8. See the papers by Peek and Wilcox (2006); Kuttner and Mosser (2002); Muellbauer (2007); Iacoviello and Minetti (2003); Goodhart and Hofmann (2007); de Bondt (2005), Gropp et al. (2007); Weber et al. (2008); Calza et al. (2007) and IMF (2008).
References Ahearne, A.G., J. Ammer, B.M. Doyle, L.S. Kole and R.F. Martin (2005) House Prices and Monetary Policy: A Cross-Country Study, Board of Governors of the Federal Reserve System, International Finance Discussion Papers No. 841. Altissimo, F., E. Georgiou, T. Sastre, M.T. Valderrama, G. Sterne, M. Stocker, M. Weth, K. Whelan and A. Willman (2005) Wealth and Asset Price Effects on Economic Activity, ECB Occasional Paper Series, No. 29, June. Ando, A. and F. Modigliani (1963) The Life-Cycle Hypothesis of Saving: Aggregate Implications and Tests, American Economic Review, No.103, 55–84. Altissimo, F., E. Georgiou, T. Sastre, M.T. Valderrama, G. Sterne, M. Stocker, M. Weth, K. Whelan and A. Willman (2005) Wealth and Asset Price Effects on Economic Activity, ECB Occasional Paper Series, No. 29, June. Atterhög, M. and H.-S. Song (2004) Policies to Increase Assess to Home Ownership for Low Income Groups, ENHR Conference, Cambridge, 2–6 July. Ball, M. (2002), RICS European Housing Review, RICS, London. Bean, C. (2004), Asset Prices, Monetary Policy and Financial Stability: A Central Banker’s View, speech given at the American Economic Association Annual Meeting, San Diego. Bernanke, B.S. (2002) Asset-Price ‘Bubbles’ and Monetary Policy, speech delivered at the New York Chapter of the National Association for Business Economics, New York, October 15. Bernanke, B.S. (2007) Housing, Housing Finance and Monetary Policy, in: Housing, Housing Finance, and Monetary Policy, proceedings of a symposium by the Federal Reserve Bank of Kansas City, 30 August–1 September, Jackson Hole, Wyoming, pp. 1–20. Boelhouwer, P., M. Haffner, P. Neuteboom and P. De Vries (2004) House Prices and Income Tax in the Netherlands: An International Perspective, Housing Studies 19 (3), 415–32. de Bondt, G. (2005) Interest Rate Pass-through: Empirical Results for the Euro Area, German Economic Review, 6(1), 37–78. Borio, C. and P. Lowe (2002) Asset Prices, Financial and Monetary Stability: Exploring the Nexus, BIS Working Paper, No. 114. Calza, C., T. Monacelli and L. Stracca (2007) Mortgage Markets, Collateral Constraints and Monetary Policy: Do Institutional Factors Matter?, Center for Financial Studies, Working Paper Series No. 2007/10. Case, K.E., J.M. Quigley and R.J. Shiller (2005) Comparing Wealth Effects: The Stock Market versus the Housing Market, Advances in Macroeconomics, Volume 5, Issue 1. Catte, P., N. Girouard, R. Price and C. André (2004) Housing Markets, Wealth and the Business Cycle, OECD Economics Department Working Papers No. 394, Paris. Cecchetti, S.G., H. Genberg, J. Lipsky and S. Wadhwani (2000) Asset Prices and Central Bank Policy, Geneva Report on the World Economy 2, CEPR and ICMB.
38
Housing Market Challenges in Europe and the United States
Chiuri M.C. and T. Jappelli (2003) Financial Market Imperfections and Home Ownership: A Comparative Study. European Economic Review, 47, 857–75. Committee on the Global Financial System (2006) Housing Finance in the Global Financial Market, CGFS papers No. 26. Duca J. and S. Rosenthal (1994) Borrowing Constraints and Access to OwnerOccupied Housing, Regional Science and Urban Economics 24 (1994), 3101–22. Dynan, K., and D. Kohn (2007) The Rise in US Households Indebtedness: Causes and Consequences, Federal Reserve Board, Washington, D.C. ECB (2003) Structural Factors in the EU Housing Markets, March 2003. ECB (2009) Housing Finance in the Euro Area, ECB Occasional Paper, No. 101. Ellis, L. (2008) The Housing Meltdown: Why Did It Happen in the United States?, BIS Working Paper, No. 259, September 2008. European Commission (2005) Green Paper on Mortgage Credit in the EU, COM(2005) 327. Girouard, N., M. Kennedy, P. van den Noord and C. André (2006), Recent House Price Developments: The Role of Fundamentals, OECD Economics Department Working Papers, No. 475. Giuliodori M. (2004) Monetary Policy Shocks and the Role of House Prices Across European Countries, DNB Working Paper, No. 15. Guiso, L. and T. Jappelli (1999) Private Transfers, Borrowing Constraints and the Timing of Homeownership, Centre for Studies in Economics and Finance, Working Paper No. 17. Goodhart, C. and B. Hofmann (2007) House Prices, Money, Credit, and the Macroeconomy, Financial Markets Group, London School of Economics, mimeo. Greenspan, A. (2002) Economic Volatility, speech delivered at a symposium by the Federal Reserve Bank of Kansas City, 30 August–1 September, Jackson Hole, Wyoming. Gropp, R., C. Kok Sørensen and J.-D. Lichtenberger (2007) The Dynamics of Bank Spreads and Financial Structure, ECB Working Paper Series, No. 714. Guiso, L. and T. Jappelli (1999) Private Transfers, Borrowing Constraints and the Timing of Homeownership, Centre for Studies in Economics and Finance, Working Paper No. 17. Haffner, M. (2003) Tenure Neutrality: A Financial-Economic Interpretation, Housing, Theory and Society 20(2), 72–85. Iacoviello, M. and R. Minetti (2003) Financial Liberalization and the Sensitivity of House Prices to Monetary Policy: Theory and Evidence, The Manchester School, Vol. 71:1, pp. 20–34. International Monetary Fund (2006) Household Credit Growth in Emerging Market Countries, Global Financial Stability Review, September. International Monetary Fund (2008) The Changing Housing Cycle and the Implications for Monetary Policy, World Economic Outlook, Chapter 3, April. Kohn, D.L. (2006) Monetary Policy and Asset Prices, speech at Monetary Policy: A Journey from Theory to Practice, a European Central Bank Colloquium held in honor of Otmar Issing, Frankfurt. Kok Sørensen, C. and J.-D. Lichtenberger (2007) Mortgage Interest Rate Dispersion in the Euro Area, ECB Working Paper Series, No. 733.
Peter Mooslechner and Karin Wagner
39
Kuttner, K. and P. Mosser (2002) The Monetary Transmission Mechanism: Some Answers and Further Questions, Federal Reserve Bank of New York, Economic Policy Review, May, 15–26. Leamer, E. (2007) Housing Is the Business Cycle, in: Housing, Housing Finance, and Monetary Policy, proceedings of a symposium by the Federal Reserve Bank of Kansas City, 30 August–1 September, Jackson Hole, Wyoming, 149–233. Matsaganis, M. and M. Flevotomou (2007) The Impact of Mortgage Interest Tax Relief in the Netherlands, Sweden, Finland, Italy and Greece, EUROMOD Working Paper No. EM2/07. McCarthy, J. and R.W. Peach (2005) Is There a ‘Bubble’ in the Housing Market Now? NFI Policy Brief No. 2005-PB-01. Mishkin, F.S. (2008) Housing and the Monetary Transmission Mechanism, in: Housing, Housing Finance, and Monetary Policy, proceedings of a symposium by the Federal Reserve Bank of Kansas City, 30 August–1 September, Jackson Hole, Wyoming, 359–413. Mortgage Bankers Association, National Delinquency Survey, www.mbaa.org Muellbauer, J.N. (2007) Housing, Credit and Consumer Expenditure, in: Housing, Housing Finance, and Monetary Policy, proceedings of a symposium by the Federal Reserve Bank of Kansas City, 30 August–1 September, Jackson Hole, Wyoming, 267–334. OECD (2004) Economic Outlook, Volume 2004/1, No. 75, June. OECD (2007) Economic Outlook, Volume 2007/1, No. 81, June. OECD (2008) Economic Outlook, Volume 2008/2, No. 84, December. Peek, J. and J. Wilcox (2006) Housing, Credit Constraints and Macro Stability: the Secondary Mortgage Market and Reduced Cyclicality of Residential Investment, American Economic Review, 96(2), 135–140. Scanlon, K. and C. Whitehead (2004) Housing Tenure and Mortgage Systems: A Survey of 19 Countries, ENHR Conference, Cambridge, July, 2–6. Slacalek, J. (2006) What Drives Personal Consumption? The Role of Housing and Financial Wealth, German Institute for Economic Research DIW Berlin Discussion Paper, No. 647. Swank, J., J. Kakesa and A.F. Tiemana (2002) The Housing Ladder, Taxation, and Borrowing Constraints, DNB Research Report No. 688, June 2002. Tsatsaronis, K. and H. Zhu (2004) What Drives Housing Price Dynamics: Crosscountry Evidence, BIS Quarterly Review, March 2004. Van den Noord, P. (2003) Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence, OECD Economics Department Working Paper, No. 356. Weber, A., R. Gerke and A. Worms (2008) Has the Monetary Transmission Process in the Euro Area Changed? Evidence Based on VAR Estimates, paper presented at the Seventh BIS Annual Conference on Whither Monetary Policy? Monetary Policy Challenges in the Decade Ahead, Bank for International Settlements, 26–27 June. Wheaton, W.C. and G. Nechayev (2008) The 1998–2005 Housing ‘Bubble’ and the Current ‘Correction’: What’s Different This Time?, Journal of Real Estate Research, 30(1), 1–26.
3 Subprime Mortgage Market and Current Financial Crisis Philip Arestis1 and Elias Karakitsos2
3.1 Introduction The prevalent view is that the current credit crisis has its origin in the bust of the housing bubble. But what is missing from this view is that the finance of a bubble is only possible through a corresponding increase in credit – no credit, no bubble. Thus at the heart of the current woes lies the excessive liquidity that was put in place in the last ten years or so.3 This liquidity financed in the first instance the internet bubble, but because there was no deleverage following the burst of this bubble the liquidity went on to finance other bubbles, including housing, private equity and commodities. Thus, the housing bubble is a transformation of the previous internet bubble. The excessive liquidity in the 2000s was the result of three forces: Financial liberalization, financial innovation and easy monetary policy in the US and Japan. In the US, Alan Greenspan injected liquidity and cut interest rates following the Asian–Russian crisis of 1997–98, which was only partially drained later on. Afraid of deflation in the aftermath of the burst of the internet bubble, Greenspan cut interest rates from 6.5 per cent to 1 per cent and injected huge liquidity. More important, he was late and slow in draining that liquidity and reversing the rate cuts from the middle of 2004. Ben Bernanke has imitated Greenspan and injected further liquidity following the ongoing credit crisis that erupted in the summer of 2007. This liquidity financed the last and most pronounced phase of the commodity bubble in the first half of 2008 that pushed the price of oil to $147 per barrel. The commodity bubble was the last one in the current cycle, as it affected CPI-inflation. Whereas central banks are loath in hiking rates to curb asset price inflation, a surge in CPI-inflation falls squarely into their realm. The surge 40
Philip Arestis and Elias Karakitsos 41
in commodity prices forced some central banks, like the ECB, to tighten monetary policy, whereas it delayed others, like the Fed and the Bank of England, from the urgently needed rate cuts, thus contributing to the downturn in the autumn of 2008. The acceleration of the economic downturn in the third quarter of 2008 burst the commodity bubble and demolished the myth of decoupling between developing and developed countries. The Bank of Japan has also contributed to this huge liquidity by printing money aggressively over the period 2001 to 2006 through buying back government bonds from financial institutions. The monetary base increased at nearly 20 per cent per annum in the three years to 2004, in what is called the era of ‘liquidity easing’. But even before that the monetary base was increasing at 7 per cent per annum in 1993–99. This huge liquidity bolstered the yen ‘carry-trade,’ which acquired its own momentum by leading into yen depreciation that further bolstered yen carry-trade. It is also true that financial liberalization, which had been going on since the 1970s, along with financial innovations that emanated from that era, played an equally, if not more, important role than easy monetary policy in creating the huge liquidity of the 2000s. The financial liberalization era allowed financial institutions to initiate a new financial activity, which was based on the discretion of the banks to dispose of their loan portfolio in accordance with risk management. That financial innovation relied heavily on interlinked securities and derivatives, all related to asset-backed securities and subprime mortgages in particular. Subprime mortgages are a financial innovation designed to enable home ownership to risky borrowers. It is, therefore, the contention of this contribution that the origins of the current financial crisis can be explained by three interrelated features that have been going on since the 1970s. The first feature is the financial liberalization policies initiated by governments both in the developed and developing world since that time.4 The second feature is an important financial innovation that emerged following the financial liberalization experience. The financial innovation in question is based on the issue of financial structured products, such as Collateralized Debt Obligations (CDOs) that played a key role in the swelling of the subprime market. Other forms of asset-backed securities were also issued related to commercial real estate, auto loans and student loans, whereas credit default swaps (CDSs) were issued to insure investors against the risk of default of the issuer. The third feature springs from the type of new economic policies pursued by a significant number of central banks around the world, which
42
Housing Market Challenges in Europe and the United States
aspire to the New Consensus in Macroeconomics (see, for example, Arestis, 2007). This new policy is entirely focused on monetary policy at the nearly total demise of fiscal policy, and more importantly from the point of view of this contribution, its emphasis on frequent interest rate changes as a vehicle to controlling inflation. The impact of these three types of development has been the creation of enormous liquidity and household debt in the major economies, but in the US and UK in particular, which has reached unsustainable magnitudes and produced the current crisis. This contribution relies on these three features for a possible explanation of the origins of the current crisis. But the root of the current financial crisis is the creation and subsequent developments in the subprime mortgage market, the focus of this contribution. We begin with a brief discussion of financial liberalization in Section 2. This is followed in Section 3 by an extensive discussion of the financial innovation, the subprime mortgage market, which helped to promote the climate for the financial crisis of August 2007. Section 4 is devoted to the current economic policies as an additional potential source of the current financial crisis. Section 5 attempts a quantitative assessment of the current financial crisis, and Section 6 attempts to derive lessons from the current financial crisis. Section 7 summarizes and concludes.
3.2 Financial liberalization Ever since 1975 there has been a period, which we may label as neoliberalism, or Washington Consensus or globalization consensus. The main characteristic of this period has been financial deregulation and free capital mobility, or more succinctly financial liberalization. This is justified by the ‘efficient markets hypothesis,’ which assumes that all unfettered markets clear continuously thereby making disequilibria, such as bubbles, highly unlikely. Economic policy designed to eliminate bubbles would lead to ‘financial repression,’ a very bad outcome in this view. The principle of financial liberalization is based on the premise that the financial sector of an economy provides real services, whereby financial instruments, markets and institutions arise to ameliorate market frictions: They can mitigate the effects of incomplete information and transaction costs. The early experience of countries, which went through financial liberalization, leads to the conclusion that what happened in the relevant economies was that financial liberalization typically unleashed a massive demand for credit by households and firms that was not offset by a comparable increase in the saving rate. Loan rates rose as households demanded more credit to finance purchases of
Philip Arestis and Elias Karakitsos 43
consumer durables, and banks were very happy to oblige. In terms of bank behavior, banks increased deposit and lending rates to compensate for losses attributable to loan defaults. High real interest rates completely failed to increase savings or boost investment – they actually fell as a proportion of GNP over the period. The only type of savings that did increase was foreign savings, i.e. external debt. This, however, made the ‘liberalized’ economies more vulnerable to oscillations in the international economy, increasing the debt/asset ratio and thus service obligations and promoting the debt crises experienced in the 1980s and 1990s in the main. Financial liberalization thus managed to displace domestic for international markets. Long-term productive investment never materialized either. Instead, short-term speculative activities flourished whereby firms adopted risky financial strategies, thereby causing banking crises and economic collapse. Despite, though, the early troublesome attempts at financial liberalization, and the increasing problems and skepticism surrounding the financial liberalization thesis over the years since its inauguration, it, nevertheless, had a relatively early impact on development policy through the work of the IMF and the World Bank. The latter two institutions, perhaps in their traditional role as promoters of what were claimed to be free market conditions, were keen to encourage financial liberalization policies in developing countries as part of more general reforms or stabilization programs. But the near unanimity of the international agencies on the benefits of financial liberalization has never found support by contributors elsewhere. It would appear actually to be the case that financial liberalization is a very controversial issue. Be that as it may, when events following the implementation of financial liberalization prescriptions did not confirm their theoretical premises, there occurred a revision of the main tenets of the thesis. Gradual financial liberalization is to be preferred. In this gradual process a sequencing of ‘financial liberalization’ is recommended. A further response by the proponents of the financial liberalization thesis has been to argue that where liberalization failed it was because of the existence of implicit or explicit deposit insurance coupled with inadequate banking supervision and macroeconomic instability. These conditions were conducive to excessive risk-taking by the banks, a form of moral hazard, which can lead to ‘too high’ real interest rates, bankruptcies of firms and bank failures. This experience led to recommendations, which included ‘adequate banking supervision,’ aiming to ensure that banks have a well-diversified loan portfolio, ‘macroeconomic stability,’ which refers to low and stable inflation and a sustainable fiscal deficit, and sequencing of financial reforms.
44
Housing Market Challenges in Europe and the United States
These post hoc theoretical revisions were thought sufficient to defend the original thesis of a disappointing empirical record. Despite all these modifications, however, there is no doubt that the proponents of the financial liberalization thesis do not even contemplate abandoning it. No amount of revision has changed the objective of the thesis, which is to pursue the optimal path to financial liberalization, free from any political, i.e. state, intervention. We suggest that it was essentially the financial liberalization era, which promoted the financial innovation that caused the current financial crisis along with the new monetary policy as argued below.
3.3 Financial innovations A new financial development emerged following the financial liberalization era, which has played an equally, if not more, important role than easy monetary policy in creating the huge liquidity and debt of the 2000s. In terms of financial liberalization in the US, the repeal of the US 1933 Glass–Steagall Act in 1999 was an important event. The repeal of that Act allowed the merging of commercial and investment banking and thereby enabling financial institutions to separate loan origination from loan portfolio.5 Banks were no longer obliged to keep their own loan portfolio. It was at the discretion of the banks to dispose of their loan portfolio in accordance with risk management. The repeal of the 1933 Act in 1999 promoted an important financial innovation, which encouraged banks to provide risky loans without applying the three C’s to each borrower – Collateral, Credit history and Character. This was so since banks could easily sell these mortgages or other loans to an underwriter, or act as an underwriter to sell to the public exotic mortgages backed by low quality securities. This led to the unprecedented growth of the subprime market (loans to borrowers with poor credit history or with questionable ability to service their loans in adverse economic conditions) especially in the last three years to 2007.6 Banks set up Structured Investment Vehicles (SIVs) with a simple legal structure (trust or just a limited liability company) that required a very small capital base. This created a ‘shadow-banking’ working in parallel to banking, but outside the regulatory umbrella and sowed the seeds for the current credit crisis. That innovation was heralded as ‘a movement that seems to reconcile socioeconomic equity with the imperatives of profitability in a competitive and turbulent industry’ so that ‘mortgage lending has emerged as the key to revitalizing the inner city, opening access to suburban housing markets, and promoting household wealth accumulation. Prodded
Philip Arestis and Elias Karakitsos 45
by policy makers, the housing finance industry is now racing to tap new markets for homeownership by reaching traditionally undeserved populations of racial and ethnic minorities, recent immigrants, Native Americans, and low- to moderate-income (LMI) households’ (Listokin et al., 2000, p. 19). The new financial innovation was based on the idea that the borrower and the lender can benefit from house price appreciation over short horizons, whereby the mortgage was rolled into another mortgage. The appreciation of housing becomes the basis of refinancing over short periods of time. Borrowers thereby were able to finance and refinance their homes in view of the capital gains as a result of house price appreciation. The appreciation enabled borrowers to turn it into collateral for new mortgages or extracting the equity for consumption. Lenders are also willing to lend to riskier borrowers. When prices of houses rise and the borrowers ‘extract equity’ through refinancing, lenders incorporate high fee prepayments to secure themselves. The main characteristic of a subprime mortgage market is that it is designed to force refinancing over a period of two to three years. Subprime mortgages are, thus, short term, thereby making refinancing important. But there is a prepayment penalty, whereby too early refinancing is undesirable. Most subprime mortgages are adjustable-rate mortgages, in that the interest rate is adjusted at a ‘reset’ date and rate, where the latter is significantly higher than the initial mortgage rate, but affordable (Gorton, 2008, p. 13). There is, thus, the incentive for the borrowers to refinance their mortgage before the ‘reset’ date. But the prepayment penalty makes too early refinancing undesirable.7 In fact, ‘no other consumer loan has the design feature that the borrower’s ability to repay is so sensitively linked to appreciation of an underlying asset’ (Gorton, 2008, p. 19). The subprime mortgage market worked well, precisely as it was supposed to work, over the period 1998 to 2007. And as Gorton (2008, p. 18) reports, the fraction of subprime refinancing, which involved equity extraction is calculated to have been anything between 51.3 per cent to 58.6 per cent over that period. The next question is how the subprime mortgages were financed. The short answer is securitization and, as mentioned in note 6, between 2005 and 2006 the subprime mortgage origination was about $1.2 trillion, 80 per cent of which was securitized (see table on p. 20 of Gorton, 2008). Banks set up trusts or just limited liability companies, what is known as Structured Investment Vehicles (SIVs), which required a very small capital base.8 This created parallel banking outside the regulatory umbrella and sowed the seeds for the current credit crisis. The SIV
46
Housing Market Challenges in Europe and the United States
operations were financed by borrowing from the short end of the capital markets, the rate of which is linked to the inter-bank rate of interest, the LIBOR rate. This short-term capital was then used to buy the risky segment of the loan portfolio of the mother company. The loan portfolio was then re-packaged in the form of Collateralized Debt Obligations (CDO), which was sold to other banks and to the personal sector. In the process and so long as the inter-bank LIBOR rate remained below the rates of CDOs, SIVs made profits. CDOs are financial securities that bundle different kinds of debt. They range from corporate bonds to securities underpinned by mortgages to debt backed by money owed on credit cards, and thereby cut debt into slices. These slices are sold to investors in the form of bonds. While the slices contain the same debt, they differ in terms of which pay the most interest and which are least at risk of losing money. Slices that pay lesser amounts of interest are the last to get wiped out by losses if there are defaults in the debt pooled in the CDO. Slices that pay more feel the pain more quickly. This is the way that some high-risk debts can be packaged to receive investment-graded credit ratings. This is a result of the CDO structure and the diversification gained by bundling different debts. At the same time, CDOs use borrowed money to amplify returns. The popularity of CDOs grew as low interest rates caused investors to embrace products that offered the promise of higher yields. Advocates argue that CDOs allow investors to buy into higher-yielding securities while taking on the same risk as they would with safe, lower-yielding securities. They also insist that CDOs are another tool that allow financial markets to further spread risk so it is not concentrated in financial institutions but shared with the personal sector, thereby reducing systemic risk. But the opponents think CDOs are an example of financial engineering gone haywire. CDOs are ‘more sleight of hand’ than a sound way to generate diversified returns. They are a method for Wall Street to repackage securities as a way to make more money. Indeed, Wall Street has made millions of dollars in fees in recent years by creating CDOs, selling them, servicing them and helping investors trade them. They are vehicles generally used by institutional investors, such as pension funds or hedge funds, not individual investors. As a result, these days banks hold few traditional liquid assets, such as government bonds; they are loaned up with claims of varying quality on the private sector, largely based on residential or commercial property. The housing bubble burst when the yield curve became inverted with long-term interest rates lower than the inter-bank LIBOR rate of interest. This confirms the myopic attitude of financial institutions in
Philip Arestis and Elias Karakitsos 47
making profits and raises the issue of whether management acts in the best interest of shareholders in the long run. The cynics would say that as the remunerations of management are linked to current profits they have an incentive to make risky investments that would hurt in the long run the interests of shareholders. If and when these investments turn sour a new management would be called in to clear up the mess. The old management will walk away with huge profits. The complex structure and highly illiquid nature of the CDO market has complicated the task of credit rating institutions, which erroneously assigned AAA-status to many worthless papers. The overstated credit rating has contributed to the growth of the CDO market in the upswing of the cycle, but also to its downfall in the downswing. This aggravated the losses of financial institutions during the credit crisis. The CDO market, which at the peak hit $10 trillion, injected huge liquidity into the system. This was not reflected in monetary aggregates and, therefore, not monitored by central banks with respect to its implications for financial markets and the economy. The sale of CDOs to international investors made the US housing bubble a global problem and provided the transmission mechanism for the contagion to the world economy and Europe, in particular. The complex interlinking of securities, structures and derivatives resulted in asymmetric information and loss of information, especially so in terms of the risks involved, which were unknown to anyone (Gorton, 2008, p. 45). The chain of interlinked securities does not allow the location of the risk involved to be determined in that its resting place cannot be ascertained. Ultimately, loss of confidence emerged since establishing the underlying mortgages was not possible.9 Interestingly enough, while this interlinking implied spreading the risk around, it resulted in loss of transparency as to where the risks in question would eventually emerge. The banks were so greedy in providing risky loans that in the upswing of the cycle the pace of accumulation was faster than the pace of unloading them from their books. Thus, when the credit crisis started many banks found a higher than desired stock of CDOs in their balance sheets. The losses from CDOs exacerbated the losses of financial institutions. For reasons of reputation, many banks were forced to incorporate the balance sheets of the SIVs into their books. In normal times financial innovations reduce risk and convince central bankers that there is a minimal systemic risk of contagion. This is indeed what happened in the first year of the subprime crisis. Prior to the eruption of the credit crisis in August 2007, central bankers on both sides of the Atlantic had underestimated the systemic risk from the collapse of the subprime market. They claimed in the spring of 2007 that only a
48
Housing Market Challenges in Europe and the United States
few individuals and institutions would be hurt with minimum damage to the economy as a whole. This led the Fed Chairman to keep interest rates high as late as August 2007. But there was a drastic reversal of that policy following the plunge of equity prices and the widening of credit spreads in August 2007. The Fed injected liquidity and cut interest rates aggressively from 5.25 per cent to 1.0 per cent over the period August 2007 to October 2008. The Fed also took extraordinary steps over this period to extend liquidity to brokers and investment banks in addition to commercial banks. In the US, when all programs are put together, the total liquidity injected into the system amounts to $7.4 trillion or 50 per cent of nominal GDP. This huge liquidity poses problems for an orderly deleverage of the financial system in the future unless it is drained after the panic phase of the bubble dissipates. All major central banks have an aversion to bailing out speculators when asset bubbles burst, but ultimately, as custodians of the financial system they have to do exactly that. They justify their actions as stemming from the goal of preventing the burst of the bubble from taking its toll on the economy. The intention may be different, but the result is the same: Speculators, careless investors and banks are bailed out. Thus instead of encouraging deleverage and taking steps to drain the excess liquidity that has been at the root of all problems in the current decade, central banks rushed to act as lender of last resort and prevent the risk from becoming systemic, thereby posing a threat to the whole financial system in the long run. The Fed adopted a risk management approach to the current crisis with the bailout of Bear Stearns in March 2008, which set a precedent for the bailouts of Fannie Mae, Freddie Mac and AIG in September 2008, but the bankruptcy of Lehman Bros fuelled the losses of financial institutions and aggravated the financial crisis. After the collapse of Lehman the US policy-makers have not allowed anyone to fail, with the latest example being the Citigroup bank. The Fed, for reasons of moral hazard, suggested a low price for the takeover of Bear Stearns by JP Morgan, which, however, penalized shareholders and not the management that was responsible for the bad investments. While there is no doubt that the Fed response is right in the short run, it is wrong from a long-term perspective. The prodigious liquidity injected since the outbreak of the crisis came back to haunt us through the last phase of the commodities bubble in the first half of 2008, as it fanned CPI-inflation and called for central banks to act. Some central banks, such as the ECB, hiked rates, while others were prevented from cutting rates at a time that growth was weakening, thus precipitating the downturn in the global economy since the third quarter of 2008.
Philip Arestis and Elias Karakitsos 49
The commodity bubble burst in the summer of 2008, as expectations of decoupling between the growth rate of Brazil, Russia, India and China (the BRIC countries) and the mature economies were dashed, in view of the international contagion of the credit crisis. More recently, middle of September 2008, what began in August 2007 with market turmoil surrounding US subprime mortgages became a financial storm of historic proportions. The US government announced sweeping actions to head off wider market disruptions, including plans to purchase distressed mortgage related securities on a massive scale, as well as a one-year guarantee of money market mutual funds. Consequently, one may restate the problem by suggesting that financial innovations and closer links between banks transformed what started in August 2007 as a liquidity crisis into a solvency issue for the financial sector.10 The credit crisis can be seen as unfolding in three stages. In the first stage credit spreads are widening as banks become unwilling to lend to each other for fear of contagion from potential losses on the assets of the borrowing banks. In the second stage the losses of the financial institutions are unraveling, while in the third stage the ramifications to the economy are felt. Credit spreads have widened since the summer of 2007, although coordinated central bank efforts have succeeded at times in suppressing them. The losses of financial institutions have reached so far nearly $1 trillion, as asset-backed securities have lost around 80 per cent of their value. In this process the systemic risk to the entire financial system heightened to the point of collapse as Fannie Mae and Freddie Mac, which hold or guarantee nearly half of mortgage-backed securities ($5.4 trillion), came to a bankruptcy point and had to be bailed out by the US Treasury. In spite of the bailout of the two giants in the US mortgage market and the near collapse and eventual bailout of AIG, the systemic risk remained high with the bankruptcy of Lehman. The crisis has brought the demise of the investment-bank model and the remaining institutions (Morgan Stanley and Goldman Sachs) are running for cover behind the façade of commercial banks. As noted above, the Citibank bank is the latest victim in this process. The ramifications to the economy are likely to stem from the response of the banks to these losses – tightening of lending standards, higher cost of lending, lower availability of credit, hoarding of money balances. The only certain way that banks will get out of this mess in the long run is through a very steep yield curve in government bonds. The Fed will likely adopt a zero interest rate policy, while the 10-year yield will hover around 3 per cent offering 3 per cent gain in the banking system. The financial crisis will impair growth and reduce the rate of growth of
50
Housing Market Challenges in Europe and the United States
potential output, as even companies with good ideas and profitable new products will be denied credit. But the financial crisis will enable households and companies to curb their debt through time, thus rebuilding their impaired balance sheets. But as asset prices (houses and equities) fall the net wealth of the personal sector will be further eroded, thus forcing the savings ratio up and consumer expenditure down. With consumption falling companies will respond by shedding their labor force, cutting production and curtailing investment expenditure, thus further harming the incomes of households. This is the asset and debt deflation process.
3.4 Current economic policies The major policy implication is that monetary policy has been upgraded in the form of interest rate policy, while fiscal policy has been downgraded. A major objective of policy is ‘maintaining price stability’ (King, 2005, p. 2). King (2005) also argues that ‘Far from being ineffective, a monetary policy aimed at price stability has proved to be the key to successful management of aggregate demand’ (p. 2). However, the experience since the credit crunch of August 2007 does not seem to validate this claim. Be that as it may, this policy is undertaken through Inflation Targeting (IT). Fiscal policy, by contrast, in the last 10 years has been concerned with broadly balancing government expenditure and taxation. Its importance has been effectively downgraded as an active instrument of economic policy. The downgrade of fiscal policy is based on the usual arguments of crowding out of government deficits and thus the ineffectiveness of fiscal policy has relied on an assumption (see, however, Arestis and Sawyer, 2003, for a critique and a different view). An important assumption that permits monetary policy to have the effect as described above and within the NCM theoretical framework is the existence of temporary nominal rigidities in the form of sticky wages, prices and information, or some combination of these frictions. So that, the Central bank by manipulating the nominal rate of interest is able to influence real interest rates and hence real spending in the short run. In the long run, changes in interest rates affect inflation but have no impact on real spending or the level of economic activity, or indeed the level of unemployment; all of which can only be affected by the supply side of the economy. The financial liberalization policies pursued since the 1970s and the financial innovation, both discussed above, have produced excessive liquidity in the system thereby increasing household debt substantially. The
Philip Arestis and Elias Karakitsos 51
excessive liquidity, which became apparent by the early 2000s, was not merely the result of financial innovation, itself promoted by the financial liberalization experience as discussed above. It has also come about from the type of monetary policy following the introduction of the new monetary policy framework, the focus of which, as shown above is frequent manipulation of interest rates. In the US at the time, the Fed Chairman, Alan Greenspan, injected liquidity and cut interest rates following the Asian–Russian crises of 1997 and 1998, which was only partially drained later on. In view of the deflation dangers in the aftermath of the burst of the internet bubble in March 2000, Alan Greenspan cut interest rates in a sequence of steps from 6.5 per cent to 1.0 per cent and injected huge liquidity into the US economy. Moreover, he was late and slow in draining that liquidity and reversing the rate cuts. Ben Bernanke, the new Fed Chairman after Alan Greenspan, imitated his predecessor and injected further liquidity following the ongoing credit crisis that erupted in the summer of 2007. This experience has resulted in a serious build-up of household debt and asset holdings. Looking at debt statistics, we find that between 1998 and 2002 outstanding household debt, including mortgage debt, in the UK was 72.0 per cent of GDP; between 2003 and 2007 it shot to 94.3 per cent of GDP. In the same periods as above, outstanding household debt jumped from 76.7 per cent to GDP to 97.6 per cent of GDP in the case of the US. And in the euro area from 48.5 to 56.6 respectively (see BIS, 2008, p. 29). Clearly, this has made household expenditure more sensitive to short-term interest rate changes. Consequently, the dangers with the current conduct of monetary policy are clear: Frequent changes in interest rates can have serious effects. Low interest rates cause bubbles; high interest rates work through applying economic pressures on vulnerable social groups. Monetary policy, therefore, that depends on manipulating the rate of interest to control inflation cannot prevent the ramifications of the credit crisis. It surely is the case that regulatory and prudential controls have become extremely necessary. Many commentators during the crisis have advocated policies that avoid moral hazard. Central bankers share these concerns, but as custodians of the financial system they have to take action when markets are dysfunctional. In the current crisis they have injected temporary liquidity and provided direct loans to banks in trouble, but at a penal rate. At the beginning of the crisis central banks refrained from lowering rates that would turn the temporary injection of liquidity into a permanent one, thereby avoiding moral hazard issues. But as the crisis deepened the Fed, but not the ECB, cut interest rates and turned temporary liquidity
52
Housing Market Challenges in Europe and the United States
into permanent. This raises the issue of whether merely concentrating on inflation a central bank is rather too monolithic. The Fed’s focus on issues other than housing has given us the overheated housing market this decade, the unraveling of which is threatening to plunge the US into the worst recession in the post-Second World War era. The experience of many countries, including of course the US, shows that successful control of CPI-inflation does not guarantee control of asset price inflation. The thrust of the argument is the ‘paradox of credibility,’ implying that, the more a central bank succeeds in keeping prices stable, the more likely that signs of an overheating economy will show up first in asset bubbles.
3.5 Quantitative effects The model discussed in Arestis and Karakitsos (2004) provides an assessment of the short-term effects of this asset debt deflation process. US relative house prices, which have already fallen by 18 per cent since their peak in July 2006, are likely to be eroded by another 18 per cent by the end of 2009. Even nominal house prices, which have already fallen 12 per cent in the same time period, are likely to fall by another 18 per cent by the end of 2009. The model suggests that the trough of the housing market is likely to be hit towards the end of 2009. A year after house prices peaked equity prices commenced falling, thus putting further downward pressure on the wealth of households. Financial wealth has declined by 9 per cent by the end of June 2008, from its peak in September 2007, and the model suggests that further losses are likely with the benchmark S&P 500 bottoming at around 700 by the end of 2009. This is dragging the US economy into recession through a weakness in consumption. Inflation will dissipate to 1.5 per cent in the next 12 months, while the Fed has already cut the fed funds rate to 1 per cent. However, the risks are on the downside as house prices are likely to overshoot their long-run equilibrium, thus triggering second-round effects in bank losses and the wealth of the personal sector. The precise forecast will depend on the final estimate of the bank losses, which have now reached $1 trillion. In the second quarter of 2008 households reduced for the first time their mortgage debt by more than 3 per cent. The model suggests that mortgage debt will decline by 13 per cent by the end of 2009. The net effect of the decline in house prices and equities was a reduction of personal sector wealth by 10 per cent (June 2008); but it is likely to be slightly bigger by the end of 2009. This is due to the combined effect of falling prices and lower debt. Consumers are likely to retrench as a result of the decline in wealth, thus prompting firms to shed labor.
Philip Arestis and Elias Karakitsos 53
Job losses will mount in the next 12 months and bottom probably at the end of 2009. The combined effect of a fall in net wealth and real disposable income will curb consumption growth to 1 per cent in 2008 and just 0.1 per cent in 2009. Businesses are bound to curtail investment. The model suggests that investment will fall –6 per cent in 2008, but increase less than 1 per cent in 2009. Export growth, the only robust component of aggregate demand so far, will be halved in 2009. The overall effect on GDP is expected to be 1.5 per cent in 2008 and just 0.6 per cent in 2009. CPI-inflation will decline in the course of the next 12 months in response to a widening negative output gap and because of the burst of the commodities bubble, as the theory of decoupling between BRIC countries and the western world has collapsed. The process is likely to involve second-round effects. As house prices and equity prices continue to fall the losses of financial institutions are magnified with further deflationary effects on the economy. The risks are on the downside with house prices likely to overshoot their longrun equilibrium of 30 per cent. In the absence of policy intervention these second-round effects take hold and the asset–debt deflation process deepens. Judging from the experience of past crises, such as Japan in the 1990s, the Great Depression in the 1930s and the railways in the late 1800s, the deflation process takes around 10 years to unwind. However, the Paulson rescue plan should speed up the process of adjustment and the asset–debt deflation process may take two to three years. Two parameters will shape the accuracy of the forecast – the extent of house price drop and the losses of financial institutions.
3.6 Lessons from the current financial crisis The US housing market was the primary cause of the credit crisis. That was helped by the huge liquidity that was put in place by ‘bad’ financial engineering and some mistakes in the conduct of monetary policy, especially in the US. As a result of both forces the global economy is now in the midst of a very serious downturn. The liquidity to which we have just referred has financed a number of bubbles in the last 10 years with a major impact on the economy (internet, housing, and commodities) and a few more (shipping and private equity) with a minor impact on the economy. From a European perspective microeconomic fundamentals and country specific factors have differentiated the countries in the euro area with housing bubbles emerging in some countries, like Spain, but not in others, like Germany. Thus, what is needed is both a macro- and micro-perspective to understand the full story.
54
Housing Market Challenges in Europe and the United States
From a macro-perspective liquidity is the real culprit. Without this excessive liquidity there would have been no bubbles – no credit, no bubble. Although one might point to some errors on the part of the Fed in removing the accommodation bias on a number of occasions in the last 10 years, ‘bad’ financial engineering has played by far a more important role in creating this prodigious liquidity. ‘Bad’ financial engineering purports to find loopholes in the law and the regulatory environment to make money. ‘Bad’ financial engineering has resulted in a ‘shadowbanking’ that developed and worked in parallel with regulated banking. The ‘shadow-banking’ operated outside regulation and control of the authorities. So, whatever was not allowed in regulated banking was developed in ‘shadow-banking’. The backlash of the greed of financial institutions is likely to increase calls for strict regulation of the industry. As the taxpayer is called upon to clean up the mess of the banks tougher regulation of the industry is very likely to ensue. But from a policy perspective it should be recognized that regulation is backward rather than forward-looking. Smart people will always take advantage of any given legislation by finding loopholes. Regulators will always react with a long lag to close the loopholes and in some occasions, like the current crisis, too late to prevent a calamity. A better approach than over-regulation is for the central bank to have a target on asset prices in a way that does not impede the functioning of free markets and does not prevent ‘good’ financial innovation. Since securitization implies the transfer of assets and the risk to the personal sector the ideal target variable for a central bank is the net wealth of the personal sector as a percentage of disposable income, which is a stationary variable and therefore a target range can be set. In the US, for example, this can be a range around 5-times the net wealth of the personal sector. In this way the central bank will monitor the implications of financial innovations as they impact on net wealth, even if it is ignorant of these innovations, as in the case of SIVs. With a wealth-target the central bank will act pre-emptively to curb an asset upswing cycle from becoming a bubble. Information on the constituent components of net wealth is available in the US with one-quarter lag, a month after the release of the NIPA accounts, thus making it useful for policy analysis and targeting. In the euro area there are huge efforts to compile such data, a prerequisite for targeting. Asset-led business cycles, like the current one, Japan in the 1990s and the US in the 1930s, produce a larger variability in output than inflation. In the upswing of the cycle output growth surpasses historical norms giving the impression that potential output growth has increased, thus
Philip Arestis and Elias Karakitsos 55
creating a general feeling of euphoria and prosperity, as it did in the second half of the 1990s in the US. But in the downswing the recession is deeper than normal, and even more important, it lasts for a long time with many false dawns, as in the case of Japan. As asset prices fall the past accumulation of debt becomes unsustainable and households and businesses engage in a debt reduction process by retrenching. This depresses demand putting a new downward pressure on asset prices thus creating a vicious circle. The policy implication is that in asset-led business cycles guiding monetary policy by developments in inflation alone will not prevent the bubble from becoming bigger than otherwise. Monetary policy should be formulated with at least two targets: Inflation and the output gap. In addition, there are merits for a mild, but not excessive, wealth targeting. The problem with excessive wealth targeting is that there are three targets and just one instrument – interest rates. Although a rate hike might reduce the output gap, diminish inflation and curb the net wealth of the personal sector, the impact on each target would be felt with a variable lag. This differential speed of adjustment of each target to monetary policy poses perils to the central bank task of stabilizing the economy along the potential output growth path. Thus, strict adherence to the fulfillment of each target by the central bank may cause instability rather than stability. But these are long-term policies, and as such they are not helpful in getting out of the current one. The burst of a bubble in the last 500 years has entailed asset and debt deflation that has triggered retrenchment on the part of households and firms with severe consequences for profits, the incomes of households and jobs. The deflation process is usually long and painful and the evidence of the last three episodes (1870s, 1930s and Japan in the 1990s) is that it usually lasts for 10 years. The policy-makers’ efforts so far have concentrated on unfreezing the credit markets and restoring confidence in banks by pumping liquidity and guaranteeing bank loans so that the interbank market can start to function again. They have also assigned public funds to recapitalize banks by buying mostly preferred shares and increased the guarantee limit on deposits to deflect runs on depository institutions. In the US the Fed has, in addition, extended credit facilities to non-depository institutions and has lowered the quality of assets that it accepts as collateral for lending. Although these measures may be adequate to ease the panic phase of the burst of a bubble, they are inadequate to deal with the crisis in the long run, as they deal with the supply side of credit, but not with the demand for it. The challenge for policy-makers is to break the vicious circle between falls in house prices and bank losses if they are to shorten the asset and
56
Housing Market Challenges in Europe and the United States
debt deflation process to less than 10 years. This requires preventing households from falling into negative equity; otherwise, delinquencies rise and bank losses mount; mortgage-lenders repossess the properties and dump them into the market that only causes lower house prices and even higher bank losses. Spending public money to cover the losses of the banks without supporting households to keep their homes and encourage others to obtain new mortgages is like throwing money into a black hole. Hence, the policies that should be pursued are on both sides of the credit market: Demand and supply. Unless demand for credit and demand for the general products of the banks are boosted in the months ahead, no amount of money can salvage the financial system. Dealing just with the supply side of credit by ignoring its dependence on demand will be a waste of resources. Hoarding of cash by banks, mutual funds, hedge funds, businesses and individuals will be a terrible blow to demand for credit that will trigger new losses for the financial institutions in the new-year, thus creating a vicious circle. We are now in what Keynes called a ‘liquidity trap’. Monetary policy does not work in this environment and neither does fiscal policy in the form of tax cuts; people will hoard the extra money – they will not spend it. What is needed is public works. A new Fannie Mae should be created, along the original model of the ‘New Deal,’ as the current one does not inspire confidence. The new Fannie should take from the banks the loans to all those who are threatened with foreclosure or business bankruptcy and offer them affordable loans to boost demand. Although the measures adopted so far are dealing with the panic, policy-makers are inconsistent in their long-term objectives in that they want both deleverage and high asset prices. They should either engineer an orderly deleverage, while at the same time accepting that in the new long-run equilibrium asset prices would be substantially lower; or they should flood the system with liquidity to prevent the erosion of asset prices, but knowing that deleverage would not materialize. In other words the policy-makers are not clear as to whether they target in the long run deflation or inflation. It is a hard fact of life, however, that from a long-term perspective the first target is what makes sense; otherwise, the excess liquidity that financed so many bubbles in the last 10 years will not be drained and will carry on financing new bubbles. Irrespective of whether the policy-makers target deflation or inflation, the forces of deflation are more powerful than those of inflation. So, even if the policy-makers wished to reflate asset prices, they might find it extremely hard to achieve their objectives.
Philip Arestis and Elias Karakitsos 57
3.7 Summary and conclusions We need to regulate financial engineering. Securitization implies a transfer of risk from banks to the personal sector and makes banks more willing to promote both lending and the sale of asset-backed securities to the personal sector. We should avoid the problem of fraud in the subprime arena; the problem has never been with the subprime model per se. It is this financial engineering that allowed US housing to become a bubble. Financial engineering is so complex that central banks would have a tough time if they wanted to measure, monitor and control the total liquidity in the economy. New policies are desperately needed, and targeting the net wealth of the personal sector is one such policy suggested in this contribution. Above all we should not lose sight of the fact that this crisis is the result of regulatory failure to guard against excessive risk-taking in the financial sector. Policy-makers must ensure that it does not happen again. Work has actually started to rebuild the architecture and the leading industrialized countries have already put forward recommendations for better prudential regulation, accounting rules and transparency. The role of credit agencies will also need to be rethought, with greater public scrutiny. In a globalized world, these efforts will have to be broad-based if they are to be effective. As for the real sector it ought to be emphasized yet again that under current circumstances public spending is the most effective means of getting the economy out of the current financial and economic trouble.
Notes 1. University Director of Research, Cambridge Centre for Economic and Public Policy, Department of Land Economy, University of Cambridge. 2. Associate Member, Cambridge Centre for Economic and Public Policy, University of Cambridge and Director of Guildhall Asset Management, PCC, Ltd. 3. Liquidity for the purposes of this chapter is to be understood not merely as reflecting monetary aggregates but also including ‘shadow’ banking. This is totally unregulated and provides loans that are financed by asset-backed securities. The latter’s multiplier could be infinite if the yield curve has a positive slope permanently – that is, the long-term rate is above the shortterm interbank rate, i.e. the LIBOR rate. The LIBOR (London Inter Bank Rate) is compiled by the British Bankers Association (BBA) and published daily between 11am and 12 noon London time. LIBOR rates are averages of interbank rates in major countries world-wide. They are submitted in panels, which comprise at least eight contributor banks; pound sterling, US dollar, euro and the yen have 16 banks (Gorton, 2008). The following BBA website provides further details: http://www.bba.org.uk/bba/jsp/polopoly. jsp?d=141
58
Housing Market Challenges in Europe and the United States
4. It was not just where financial liberalization was overtly introduced, but also where the authorities were required to operate under strict rules. An interesting example is the UK Financial Services Authority (FSA), set up in 1997 when the Bank of England was granted ‘independence’. Although the FSA was given sweeping jurisdiction over the British financial sector, it has regulated it ‘diffidently’. In the words of its first chairman ‘The philosophy of the FSA from when I set it up has been to say, “Consenting adults in private? That’s their problem, really.” ’ (Eisinger, 2008). 5. Established in 1933, the Glass–Steagall Act was repealed in 1999 thereby opening up competition among banks, securities and insurance companies. The Glass–Steagall Act prohibited a bank from offering investment, commercial investment and insurance services. See for full details: http://en. wikipedia.org/wiki/Gramm-Leach-Bliley_Act#Remaining_restrictions 6. ‘Subprime mortgage origination’ in 2005 and 2006 was $1.2 trillion, 80 per cent of which was securitized (see Gorton, 2008). The same study provides further data on the growth of the subprime mortgage market: ‘The outstanding amounts of Subprime and Alt-A combined amount to about one quarter of the $6 trillion mortgage market. Issuance in 2005 and 2006 of Subprime and Alt-A mortgages was almost 30 per cent of the mortgage market. Over the period 2000–2007, the outstanding amount of agency mortgages doubled, but subprime grew 800 per cent! Since 2000 the Subprime and Alt-A segments of the market grew at the expense of the Agency share, which fell from almost 80 per cent (by outstanding issuance) to about half by issuance and 67 per cent by outstanding amount’ (p. 8). The ‘Subprime and Alt-A’ term is defined to refer ‘to borrowers who are perceived to be riskier than the average borrower because of poor credit history’ (p. 7). 7. Gorton (2008) offers an interesting contrast between a subprime mortgage, as explained in the text, with ‘a standard, prime, 30 year, fixed rate mortgage’. Unlike the subprime mortgage, ‘(w)ith a prime mortgage, the borrower repays principal over time, and the mortgage matures after 30 years. The borrower may repay the mortgage, typically without penalty. The borrower may benefit from house price appreciation, but the lender does not (directly) benefit’ (p. 13). 8. A large number of SIVs’ assets were in the form of subprime residential mortgage-backed securities and commercial-backed securities. 9. In 2006 new synthetic indices of subprime risk were introduced; the so-called ‘ABX’ indices. For the first time ever information about subprime values and risks was gathered and made known. The ABX information and the lack of information about location of the risks led to the loss of confidence referred to in the text. 10. The Bank of England (2008, chart 6, p. 8) provides ‘projected ultimate credit losses on subprime asset backed securities’; these are estimated to reach $170 billion.
References Arestis, P. (2007) What is the New Consensus in Macroeconomics?, chapter 2 in P. Arestis (ed.), Is There a New Consensus in Macroeconomics?, Basingstoke: Palgrave Macmillan.
Philip Arestis and Elias Karakitsos 59 Arestis, P. and E. Karakitsos (2004) The Post-Bubble US Economy: Implications for Financial Markets and the Economy, Basingstoke: Palgrave Macmillan. Arestis, P. and M. Sawyer (2003) Reinstating Fiscal Policy, Journal of Post Keynesian Economics, 26(1), 3–25. Bank of England (2008) Financial Stability Report, No. 23, April. Bank of International Settlements (BIS) (2008) Annual Report, June. Eisinger, J. (2008) London Banks, Falling Down, downloadable from: http://www.portfolio.com/views/columns/wall-street/2008/08/13/Problems-inBritish-Banking-System Gorton, G.B. (2008) The Panic of 2007, NBER Working Paper Series, No. 14358, National Bureau of Economic Research: Cambridge, MA. King, M. (2005) Monetary Policy: Practice Ahead of Theory, Mais Lecture, Cass Business School, City University, London. Listokin, D., E. Wyly, L. Keating, K. Rengert and B. Listokin (2000) Making New Mortgage Markets: Case Studies of Institutions, Home Buyers and Communities, Fannie Mae Foundation Research Report.
4 Determinants of Homeownership Rates: Housing Finance and the Role of the State Elisabeth Springler and Karin Wagner1
4.1 Introduction The US subprime crisis of the last year led to a downturn in consumer demand and put not only the US housing finance system in severe difficulty but led also to a global financial crisis and a real economic downturn. Not only have the two major governmental agencies – Fannie Mae and Freddy Mac – counted on state bail-out, but numerous banks and industries too. Also households suffer in the current situation. RealtyTrac (2009) presented in the most recent study from February 2009 an increase in foreclosures by 6 per cent compared to January 2009 and an increase of 30 per cent compared to February 2008 (in June 2008 there was an increase of even 55 per cent year-on-year, December 2008, 41 per cent increase year-on-year). These developments were preceded by distortions in the mortgage sector. As numerous studies have pointed out clearly (see among others Shiller, 2008), the mortgage sector was driven by speculative trends on increasing housing prices accompanied by a rise in subprime lending to include households with less creditworthiness in the financial market. While in the light of the current financial crisis the main research focus in this field is laid upon the interrelations between house prices and financial developments (see among others Shiller, 2008; Girouard et al., 2006; Egert and Mihaljek, 2007; Pozdena, 1988; Miles, 1994) driven by defining the cost of housing,2 one major factor of empirical evidence is left aside: the development of homeownership rates. This chapter argues that administrative policies and developments in housing finance are the crucial determinants for the rise of homeownership. The increase in homeownership rates in turn can be seen as the main force for the current mortgage crisis. By doing so, the potential of similar 60
Elisabeth Springler and Karin Wagner 61
developments in other economies can be investigated and policy advice can be derived. The chapter proceeds as follows: After discussing the development in house prices in the forefront of the US subprime crisis, the potential of speculative trends on European housing markets is analyzed from a comparative statistical point of view. Section 4.3 focuses on the development and determinants of homeownership rates. In the following, in Section 4.4 the causal mechanisms between homeownership rates and housing finance as well as federal housing policies are explored. The assumptions made in this section are integrated into the empirical modeling of Section 4.5. The data-set for the cross country analysis is compiled from the European Tax Handbooks from the International Bureau of Fiscal Documentation and the European Mortgage Federation. Policy implications that can be derived from the empirical results are discussed in Section 4.6. Section 4.7 concludes.
4.2 What could be observed preceding the current financial crisis? When the current financial crisis took off in the United States, it was preceded by a burst of a house price bubble that had followed the continuous acceleration of house prices in the previous years. The peak was reached in December 2006, when the ratio of home values to incomes reached 184 per cent in the US (Weller and Sabatini, 2007, p. 4). In the course of these events the financial vulnerability of households rose, as the mortgage payments to income, the concentration of household assets to illiquid real estate assets and a concentration of highly leveraged homeowners increased. Furthermore, the high share of mortgages with variable interest rates had increased. Analyzing these indicators of vulnerability for the period 1989 to 2004, Weller and Sabatini (2007, pp. 15ff) show clearly that the financial position of US American households became more vulnerable towards house price shocks. The magnitude of the burst of the house price bubble showed significant regional and state-wise differences in house price developments as well as differences in the magnitude of the subprime crisis of the last years, as presented in the release of the Office of Federal Housing Enterprise of Oversight (2008). While the state-wide data cumulated a decrease in house prices of 1.7 per cent between the second quarter of 2007 and the second quarter of 2008, California faced a depreciation of more than 15 per cent in the same period, followed by Nevada with a decrease of 14.1 per cent and Florida with 12.4 per cent. On
62
Housing Market Challenges in Europe and the United States
the other hand, states such as Wyoming and Oklahoma experienced an increase in their house prices of more than 4 per cent in the same period. Although European economies did not experience a house price bubble or its burst in these last years, similarities between the developments of the housing market in the US and European economies can be seen. These similarities can be detected initially by the increase in housing prices in numerous European economies as well as by the increase in residential mortgage debt to GDP ratios. In the last years leading to the crisis most European economies experienced a sharp increase in housing prices but major differences in housing prices across member states of the European Union could be observed. While economies such as Spain, France, Italy and the UK faced an annual increase of more than 10 per cent on average for the period 2002 to 2006, Germany, Austria and Portugal showed a fairly stable house price structure with minor increases (European Mortgage Federation, 2007a). In the case of Germany even a decrease was observed. Contrary to these observations the latest data of the OECD (2008, p. 29) show that house prices are now only rising at low rates or have even decreased in the last months. The slump in the American housing market followed by the subprime crisis seemed to be the turning point for the house price cycle even in most European economies.3 Besides these developments, the increase in private household debt in some European economies is stated as a signal for changes in the housing sector and as an echo of the developments in the American housing market, although major structural differences exist among European economies.4 The striking feature in this respect is not so much the difference in the level of household indebtedness across European economies, as different levels of household indebtedness in general can be explained by gaps in homeownership rates and by cultural aspects connected to homeownership, but the sharp increase of household indebtedness in the period 2002–2006 compared to the average data for the period 1997–2001 (European Mortgage Federation, 2007a; International Union for Housing Finance, (n.d.); IMF, 2008). Despite the fact that a comparison between the sharp increases in house prices and the rise in private households’ indebtedness holds for some economies, it does not allow us to conclude that there is a direct causality between those two movements, as economies such as Spain and France faced a sharp price increase accompanied by just minor rises in private households’ indebtedness.
Elisabeth Springler and Karin Wagner 63
4.3 Homeownership rates: Development and determinants in academic literature As Figure 4.1 shows, the period between 1997 and 2004 is advocated as the period of massive house price increases. This period is characterized by a continuous increase in homeownership rates, which cumulate with an increase from 65.7 per cent to 68.1 per cent in a total rise of 2.4 percentage points in the respective period in the US. A similar trend can also be observed in some European economies in the same period. The determinants for homeownership rates discussed in academic research are as follows: 1. Interest rates: Their impact is two-fold. On the one hand, developments of interest rates have a direct impact on the cost of financing owner-occupied housing. A lower interest rate would increase therefore the demand for homeownership as homeownership becomes more attractive compared to other forms of housing (such as renting).5 To increase the effectiveness of this development studies suggest that also the cost of construction has to be taken into account. Lower interest rates could also facilitate construction via lower construction costs. Although the impact of interest changes for homeownership 1.50 UK USA
1.00
0.50
0.00 97
19
19
98
99
19
00
20
01
20
02
20
03
20
20
04
20
05
⫺0.50
06
06
04
20
20
19
– 97
20
– 97
19
⫺1.00
⫺1.50
Figure 4.1 Annual changes in homeownership rates in percentage points, 1997–2006 Source: UK National Statistics, US Census Bureau, Authors’ calculations.
64
Housing Market Challenges in Europe and the United States
rates via demand side or via supply relations is valuated differently in academic research,6 an overall inverse relationship between interest rates and homeownership is acknowledged. 2. House prices: Similar to the impact of interest rates, the empirical research of Painter and Redfearn (2002) for the USA, shows a positive impact of nominal house prices on homeownership rates. 3. Sociological and demographic factors: Apart from the research on the interrelations between interest rates, homeownership and the impact of income level of the household, academic literature on the development of homeownership rates focuses on the impact of age, ethnical backgrounds and household formation. In connection with these household characteristics, tax incentives are also integrated in country-specific research (Bourassa and Yin, 2006). The interrelations between interest rates, house prices and homeownership rates (although the causality is not determined in the empirical studies described) are especially connected to the research of homeownership rates in the literature on housing finance in the American and European housing markets. Although these studies aim to explain the development of house prices, the recommendations given for the housing finance sector similarly have a major impact for homeownership rates, as the provision of low interest rates is in the center of the analysis. On the basis of Goodhart and Hofmann (2008), Tsatsaronis and Zhu (2004), Girouard et al. (2006), Warnock and Warnock (2007), Zhu (2006), Committee on the Global Financial System (2006), Egert and Mihaljek (2007), the following categorization of the most important recommendations for housing finance can be proposed: • A widening in credit supply lowers interest rates and therefore
increases demand. The European Commission (Commission of European Communities, 2007) in the white paper on the Integration of EU Mortgage Credit Markets opted in this respect for stronger harmonization and liberalization of European mortgage markets with the aim of lowering costs of loans and promoting innovative mortgage products. • A stronger reliance on market-based channels and new modes of finance, such as securitization, increased the willingness of financial institutions for mortgage products, as the risk exposure can be minimized by off-balance-sheet selling of loans.
Elisabeth Springler and Karin Wagner 65 • Restructuring of the mode of mortgages, such as an increase in the
level of the loan to value ratios, which raises the ability of banks to lend against real estate collateral, has a positive effect for housing finance and drives up demand. Although the implementation of housing finance incentives for homeownership rates, as mentioned above, would deserve empirical analyses, cross country empirical studies focus on forms of mortgage interest rate setting. And the evaluation of their implementation is only rarely investigated. Zhu7 (2006, p. 62) categorizes housing finance systems for Asian economies using an ambitious sample of variables, including scores for effectiveness of regulatory regimes on the housing sector and a real estate transparency index, which reflects data availability and proxies of corruption. The empirical results obtained from this study are ambiguous. Nevertheless, the general importance of the structural background is acknowledged. Another attempt to grasp the institutional setting of housing finance from a macroeconomic point of view is put forward by Warnock and Warnock (2007). Again strong emphasis is set upon regulatory and legal aspects of the system including legal rights for borrowers and lenders with a set of variables. Their analysis of 62 developed and emerging economies leads to broad explanations of differences in housing finance systems. Unfortunately, a clear causality between differences in national housing finance systems and house price dynamics cannot be drawn. Although Mishkin (2007) does not present a model integrating housing finance variables, he shows clearly the causal relations of the impact of housing in the transmission mechanism.8 The conclusions drawn are similar to the effects of different national financial systems on growth. Market-based financial systems promote the innovation of new financing products more strongly than bank-based financial systems. They ease investment financing by lower financing costs and impose therefore a higher growth impact. Conversely, bank-based systems are less innovative, more stable but lead to lower overall economic growth.9 Modes of regulation to increase stability have to be enforced and should become an economic policy issue within European Union member states, once the impact of the housing finance system for European Economies is pointed out. Additional to features in housing finance, numerous studies (see among others Hoeller and Rae, 2007; Catte et al., 2004; ECB, 2003, 2009) put a strong emphasis on structural features of the housing sector,
66
Housing Market Challenges in Europe and the United States
incorporating the role of the state, such as housing taxes and subsidies. To the authors’ knowledge these structural features are not integrated into empirical studies in connection with developments in housing finance.10
4.4 Housing finance and the role of the state The gaps in existing literature between theoretical and empirical observations as well as the interrelations between homeownership rates, housing finance and structural characteristics open the door for deepening the theoretical model. Therefore, the interrelations between homeownership rates, housing finance and the role of the state by integrating national taxation structures as one part of national housing allowance schemes are explored. 4.4.1 Modes of housing finance From a structural point of view, four different indirect financing modes for the housing sector can be distinguished (Green and Wachter, 2007; Lea, 2001; Mooslechner, 1994). These forms may vary across economies and do not imply certain stages of economic development. Nevertheless, the most basic form for indirect finance, the simple deposit system to transmit funds, is mostly seen in less developed economies. European economies consist of a mixture of two or more of the following three forms: Contract savings systems, which are especially common in Austria and Germany, mortgage banks and secondary market systems. Developments in housing finance are on the one hand observable in terms of financial innovations in the area of mortgage banks and, on the other hand, by the increasing use of securitization products in terms of mortgage-backed securities. Table 4.1 gives an overview over the importance of securitization products in European economies and the USA. The differences in usage and even the introduction of those products are signs of structural differences in housing finance systems. Whenever economies heavily rely on mortgage banks and the usage of mortgage bonds, which do not imply a risk transfer when sold to investors, banks as financial intermediaries account for the main source of housing finance. Contrary to this, extensive use of mortgage-backed securities accounts for the importance of the secondary market as a source of funds. The housing finance system in the US can be seen as an example of strong use of the secondary market. Although the institutional background for these innovative products was already instituted in the
Elisabeth Springler and Karin Wagner 67 Table 4.1 Introduction and usage of securitization for housing finance in European economies and the USA Country
Introduction of securitization – mortgage-backed securities (MBS)
Usage of mortgagebacked securities
Austria Belgium Denmark Germany Greece Spain France Ireland Italy Luxembourg Netherlands Portugal Finland Sweden UK USA
– yes yes yes yes 1992 1999 second half 1990s yes yes yes yes 1989 yes 1987 yes
no limited limited limited limited extensive limited yes extensive yes extensive limited limited limited extensive extensive
Note: Especially for Finland and Germany Tsatsaronis and Zhu (2004) note that the introduction of securitization remained very limited in volume. MBS might be allowed from a legal point of view but not used for housing finance. With the change in the legal framework in Germany in 2005 (Funding Register Act) also a promotion of MBS, especially residential mortgage-backed securities are expected, as registration of special purpose vehicles is facilitated (European Mortgage Federation, 2007b). Nevertheless, especially due to the recent financial crisis∗ , the volume of MBS is still very limited compared to the extensive use in covered bonds in Germany. Latest comparable data (year 2006) for Germany displays a volume of a 35 336 million of Covered Bonds Issuance and a 6 200 million of issued Residential Mortgage-Backed Securities (European Mortgage Federation, 2008) ∗ Volk (2008) states that the total secondary market for MBS almost closed in the first quarter of 2008 compared to the first quarter of 2007, while the covered bonds market declined sharply in the same period but was still operating. Source: Springler (2008, Table 15); Tsatsaronis and Zhu (2004, Table 1).
1960s with the foundation of two entities, Fannie Mae and Ginnie Mae – the latter was supposed to guarantee mortgage-backed securities, whereas Fannie Mae issued them – the system only gained in importance and volume during the 1980s and 1990s (Colton, 2002, p. 8). Earlier, the housing finance system consisted primarily of long-term fixed rate products (Green and Wachter, 2007, p. 44). Nowadays the US housing finance system has experienced further diversification in mortgagebacked securities, which includes the increasing importance of issues by
68
Housing Market Challenges in Europe and the United States
so-called non-agencies, which – in contrast to the government-sponsored enterprises Freddy Mac and Fannie Mae – focus on the provision of mortgages to non-prime debtors (households) (Colton, 2002, p. 18; Florida, 1986, xiii; Frankel, 2006, p. 76). The total volume of mortgage-backed securities increased constantly in the 1990s and reached US$2.9 billion in 2005. The volume of mortgages provided by non-agencies increased even further latterly and accounted for more than 75 per cent of the total volume of mortgage credit in the year 2005 (Frankel, 2006, p. 77). Compared to these developments, the situation in Europe is nowhere near that evolved; although many economies introduced these innovative products of housing finance, as Table 4.1 shows. Among European Union member states, the UK secondary housing finance market is by far the most developed and accounts for 47 per cent of total residential mortgage-backed securities in Europe – this amounted to a18.4 billion of issuance in the first quarter of 2006 (European Securitisation Forum, 2006, p. 2). The causalities presented in Mishkin (2007) are integrated into our empirical model. The following causalities are assumed when applying this approach: When focusing on innovative products of housing finance, which on the one hand are driven by a strong presence of the financial markets in the system – as described by the ‘housing finance revolution’ of the USA or UK – and are on the other hand enforced by the integration of international mortgage credit markets, which widen the potentials of the banking system, the demand for housing is increased. This implies a positive house price dynamic. As financial structures of the housing sector consist in all economies of a mixture of the financing modes a categorization into different housing finance systems can only be done as a classification within a continuum. This development can be further fostered by the role of the state and its approach towards housing allowances, as presented in the next section. The investigation of housing finance structures requires an indirect approach due to inconsistencies in the international data provided. Two variables are used to design the housing finance system: To account for the potentials to decrease the cost of financing via lowering interest rates due to new innovative products and integrated markets after a ‘housing finance revolution,’ the variable representative interest rate on new mortgages is integrated into the model. 4.4.2 Role of the state on the housing market As regards the role of the state in the housing sector, three basic forms of intervention can be distinguished (Amann, 2000; Czerny, 2001).
Elisabeth Springler and Karin Wagner 69
UK
AT
NL
FR
DE 0 Figure 4.2 in 2001
10
20
30
40
50
60
70
80
Selected economies: Supply side allowances to GDP ratio in per cent
Source: Stagel (2004).
Supply side subsidies, which are given for the construction of new dwellings; demand side subsidies, which support a household directly and aim to increase the demand for owner occupied housing. (The third, tax incentives, will be discussed later.) In most cases the demand side subsidies are limited to lower income levels to enable them to become homeowners. As supply side subsidies involve a stronger role of the state on the housing sector, this mode of intervention was continuously aimed to be lowered in the past decades to reach higher levels of liberalization in the market. In contrast to that, demand side subsidies are seen as a social mechanism to provide lower income classes with suitable shelter. As Figure 4.2 shows, the structure of allowances varies substantially across countries. Although the structure of housing allowances has undergone a continuous change towards reducing supply side subsidies over recent years, major variations are still visible. The authors are aware of the fact that the structure of housing allowances and the underlying institutional features of public housing homeownerships are strongly interrelated with the respective national sociological and cultural11 characteristics of an economy, nevertheless these interdependencies will not (and in fact cannot) be covered in detail (within the model we try to cover this circumstance by calculating a fixed-effects model). Owing to structural differences in the implementation of these subsidies as well as the tendency to move to transmit the responsibility
70
Housing Market Challenges in Europe and the United States
for these measures from a national to a regional level, as was done in Austria in the 1990s, no reliable comparable data exist for the volume of supply and demand side subsidies for European economies. Nevertheless, the presence and impact of the structure of housing subsidies can be grasped indirectly when integrating the permissions for buildings into the model. Without integrating the institutional framework into the modeling the expected causal relation between buildings permissions and the rate of homeownership would be clearly positive. Once the impact of state intervention and the impact of non-profit organizations and public housing are acknowledged, the causal relation can be negative. As increased homeownership with additional state intervention by changes in the tax structure can also lead to a crowding out of housing supply of multistorey housing a negative relationship in overall permissions for the construction of housing units and homeownership rates can be expected. Figure 4.3 accounts for this relationship, by presenting the development of housing permissions and subsidized multistorey units and single houses in Austria between 1996 and 2003. While permissions were decreasing continuously in the late 1990s to 2001, and
Number of units 70 000 60 000 50 000 40 000 30 000 20 000 10 000
Subsidized multi storey units
Completed flats
Subsidized single family houses
Permissions of new residential units
2003
2002
2001
2000
1999
1998
1997
1996
0
Figure 4.3 Development of subsidized housing supply and housing permissions and completions in Austria, 1996–2003 Source: Amann (2005).
Elisabeth Springler and Karin Wagner 71
completions were also decreasing, homeownership rates were increasing in the respective period. Simultaneously subsidized housing supply was decreasing. To account for the role of the government in housing allowances and to show the impacts of the institutional background, this chapter focuses on the third mode of state intervention, on the structure of tax incentives of state intervention. Contrary to the situation presented above, the evolution of tax incentives was not accompanied by a vivid discussion but accounted for changes in the structure of public support. National tax systems affect the housing sector in numerous ways: Industrial view – taxation of capital income: Among others the studies of Goulder (1989) and Anas and Arnott (1997) focus on the impact of changes in capital income taxation on the costs of supplying new housing capital by the housing industry. While Goulder (1989) states that an immediate reallocation would be the result of an equilibrium housing model once capital income taxation is increased, Anas and Arnott (1997) embed their model into an analysis of housing allowances schemes and state that additionally introduced changes in the capital income taxation improve the efficiency of the housing allowance scheme. Additionally, developments in the size and quality of the housing sector were detected and increased the value-to-rent. Despite of methodic differences there is a strong interrelation between changes in capital income taxation and housing supply as well as house prices. Residential view – taxation of income: Another line of arguments follows a residential approach, where changes in the cost of financing are discussed from a private households’ point of view. Depending on the historical background of the respective studies, the starting point for the analysis is the change in house prices. Poterba (1984) for example focuses on indirect changes of tax subsidies. During the period of high inflation in the 1970s these increased, reduced the cost of owner-occupation and simultaneously raised the homeownership rate. As real house prices declined during the 1980s, Poterba (1991) investigated the reverse causal relationship, searching for house price determinants. Developments of real user costs of owner-occupied housing – primarily via changes in tax rates – have a significant influence in house price determination. For the developments of house prices of the 1980s, Poterba (1991) stresses the importance of real user costs of owner-occupied housing especially in combination with changes in demographic factors. These direct causal interrelations between house price developments and homeownership or owner-occupied housing are further developed on the one hand in microeconomic studies, which implement different income classes in
72
Housing Market Challenges in Europe and the United States
the modeling (see among others Green and Vandell, 1999). On the other hand, a set of studies develops the setting of costs for homeownership based on national tax systems in greater detail. Apart from methods and changes in interest deductibility from mortgage rate, also taxes on imputed rent as well as the taxation of capital gains have an influence on the cost of homeownership. Taxes on imputed rent, which aim to make costs of housing more comparable between tenants and homeowners, are not very common in European economies (see IFBD, 1996 to 2006). Neither may the effect of capital gains taxes due to various exemptions in national legislations be the driving force for changes in costs of homeownership (see Wolswijk, 2005). Van den Noord (2005) focuses on interest deductibility and property taxation when computing effective tax wedges. Both studies, Van den Noord (2005) as well as Wolswijk (2005), compare European economies, but focus on different causal relations. While van den Noord (2005) aims to explain different house price developments, Wolswijk (2005) analyzes the impact of tax wedges for the development of mortgage debt across European economies. Developments in mortgage credit growth become an important issue for the empirical analysis. This also means changes in the financial structure and the volatility of house prices are interrelated. As this chapter focuses on developments in residential housing, the cost of financing is integrated in the model used, building on the study of Van den Noord (2005). Nevertheless the methodological approach varies substantially as this chapter focuses not on the status of the current tax wedge in European economies but sets the developments between 1996 and 2006 in relation to homeownership.
4.5 Is homeownership determined by taxation and housing finance? 4.5.1 Constructing time series to represent financing conditions and tax deductibility We expand the model by Fukao and Hanazaki (1986). They assume a typical price of one unit of housing P (it is six times the disposable income of an average production worker). By incorporating marginal tax rates we got the after tax nominal interest rates ia (t). As there is limited duration of tax relief, the ia (t) are time-dependent. ∞ ∞ By the relationship 0 if Pe −(i−π)t dt = 0 ia (t)Pe−(i−π)t dt(π . . . inflation rate, P . . . unit of housing) the nominal cost of financing if result.
Elisabeth Springler and Karin Wagner 73
The relevant information on countries’ tax regulation and tax deductibility was taken from European Tax Handbooks (IBFD, 1996 to 2006). Furthermore, for data on marginal rates of income tax plus employees’ social security contributions and personal income tax we used OECD series (OECD, a–c). To derive the unit of housing we needed data on disposable income. We took data (OECD, a–c) of the takehome pay of a married couple with two children and one earner. This is gross wage earnings minus total payments to general government plus cash transfer from general government for two children. The time series calculated for the nominal cost of financing is shown in Table A4.1. 4.5.2 Empirical evidence for European economies and the US We used panel data of 1612 countries (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Italy, Ireland, Portugal, Spain, Sweden, Luxembourg, Netherlands, UK and US) for the years 1997 to 2006.13 The fixed effects model used is ln (ownert ) = β1i · ln (GDPcapti ) + β2i · ln (GDPcapt−1,i ) + β3i · ln (populationgrowtht−1,i ) + β4i · ln (permitst−1,i ) + β5i · ln (mortgdebtcapti ) + β6i · ln (costfinti ) + µi + vit · ln (interestnewti ) + µi + νit owner is the ownership rate, GDPcap the GDP per capita, populationgrowth the year-on-year population growth, permits is the number of permissions, mortgdebtcap stands for the residential mortgage debt per capita, the variable costfin represented the nominal cost of financing if14 and interestnew is the representative interest rate on new mortgage loans. Further variables explaining explicit institutional features (e.g. loan-tovalue-ratios (LTVs)) were not incorporated due to the fact that they are not available in a harmonized time-series manner. But one has to keep in mind that we calculated a fixed effects model to capture for these factors.
4.6 Results of the empirical model We tested a fixed effects panel model. Table 4.2 shows that the lagged population growth and the cost of financing are detected as most significant influences to the ownership rate. This makes sense as higher
74
Housing Market Challenges in Europe and the United States
Table 4.2
Fixed effects model for ownership rate
. xtreg owner gdpcap 1.gdpcap 1. populationgrowth 1.permit mortgdebtratio costfin, fe vce(robust) Fixed-effects (within) regresssion Group variable: varland R-sq:
within between overall
= = =
0.4288 0.0172 0.0176
corr(u_i, xb)
=
–0.4042
Number of obs Number of groups
= =
102 16
Obs per group: min avg max
= = =
2 6.4 9
F(6,15) Prob > F
= =
29.32 0.0000
(Std. Err. adjusted for 16 clusters in varland) Owner
Coef.
Robust Std. Err.
t
P>|t|
[95% Conf. Interval]
.1001613 –.521102
.0521042 .0560829
1.92 –0.93
0.074 0.368
–.010896 –.171648
.2112187 .0674276
gdpcap –. L1. population∼h L1. permit L1. mortgdebtr∼o costfin _cons
.0103811
.0039163
2.65
0.018
.0020337
.0187285
–.0304139 .0279287 –.0245943 3.811632
.0121474 .0192121 .0074942 .2912128
–2.50 1.45 –3.28 13.09
0.024 0.167 0.005 0.000
–.0563055 –.0130209 –.0405678 3.190926
–.0045223 .0688783 –.0086209 4.432337
sigma_u sigma _e rho
.20886496 .0153427 .99469296
(fraction of variance due to u_i)
Source: Authors’ calculations.
population growth leads to a higher demand for housing. Over the last 10 years many countries which showed remarkable growth rates in housing investment had experienced a rapid population growth, especially from immigration. Simultaneously ownership rates rose in this period. The significant coefficient in the results underlines these developments. The variable in the center of our focus which we took as indicator for the role of the state – the cost of financing – is significantly negative. As expected, the higher the costs for financing a home, the lower the ownership rate is. This important result ‘proves’ the fundamental role that tax deductibility and fiscal measures can play for the level of a country’s
Elisabeth Springler and Karin Wagner 75
ownership rate. Additionally, the number of permissions (as indicator for the supply of housing) and mortgage debt per capita are crucial for the level of the ownership rate. As discussed in Section 4.4.1, the design of the housing finance system affects households’ ability to finance their home loans. Availability and variety of mortgage instruments are important determinants for the structure of a country’s mortgage market. So a higher ownership rate may expect a higher mortgage debt ratio of households. Both significant coefficients – for permissions and for the mortgage debt ratio – showed the expected sign. The negative coefficient of the variable permits which represented the number of permissions is not surprising, when taking the arguments of Figure 4.3 above into account.15 Even if we include the time series of the representative interest rates on new mortgage loans, the coefficients remain nearly unchanged (see Table 4.3). Additionally we checked for endogeneity of interest rates (see Table A4.2 in the Annex). A dynamic GMM model showed that the period of 10 years is quite short to show significant dynamics inside the time period.16
4.7 Conclusions and policy implications Ownership rates vary much across euro area countries and the US. The chapter searches for the factors behind this empirical observation. The results obtained from the empirical models above add some policy as well as theoretical contributions to the existing research literature on the development and determination of homeownership rates. Concerning the contributions of this chapter to housing policy recommendations the following can be stated: The model confirms the importance of housing costs for the development of homeownership rates for European economies and the US. Despite these similar results, compared to other research studies presented above, the results obtained in this model further strengthen the importance of the role of the state and highlight the immediate effects that housing policies have. Although the aim to increase homeownership rates is not manifested in many European economies as clearly as in the US, state policies show their indirect willingness to increase homeownership. Regardless of the clear impact that tax policies have on the development of homeownership rates, the empirical results do not enable us to derive a normative statement for European housing policies. This means that the results do not suggest the
76
Housing Market Challenges in Europe and the United States
Table 4.3
Fixed effects model for ownership rate (incl. interest rates)
. xtreg owner gdpcap 1.gdpcap 1. populationgrowth 1.permit mortgdebtratio costfin interestnew, fe vce(robust) Fixed-effects (within) regresssion Group variable: varland R-sq:
within between overall
= = =
0.4630 0.0120 0.0124
corr(u_i, xb)
=
–0.4356
Number of obs Number of groups
= =
96 15
Obs per group: min avg max
= = =
2 6.4 9
F(7,14) Prob > F
= =
86.38 0.0000
(Std. Err. adjusted for 15 clusters in varland) Owner
gdpcap –. L1. population∼h L1. permit L1. mortgdebtr∼o costfin interestnew _cons sigma_u sigma _e rho
Coef.
Robust Std. Err.
t
P>|t|
[95% Conf. Interal]
.1517806 –.0960605
.0643399 .0672633
2.36 –1.43
0.033 0.175
.0137852 –.240326
.2897759 .0482051
.0133821
.0039155
3.42
0.004
.0049842
.02178
–.0368985 .0357541 –.041917 .0273823 3.670055
.0117886 .0237002 .0132649 .0202113 .2766834
–3.13 1.51 –3.16 1.35 13.26
0.007 0.154 0.007 0.197 0.000
.21984747 .01538199 .99516626
(fraction of variance due to u_ i)
–.0621825 –.0116145 –.0150778 .0865861 –.07036748 –.0134667 –.0159667 .0707312 3.076628 4.263482
Source: Authors’ calculations.
preference for the promotion of homeownership rates over other housing policy aims (such as providing an affordable rental sector, etc.) as the rise in homeownership rates is not discussed from a sociological or socioeconomic point of view in this study. Increasing homeownership rates should not be counted as ‘success of housing policies’. In this respect it would be necessary to analyze the cultural and sociological dimension of these developments. Although such an analysis is clearly beyond the scope of this study the empirical results presented stress their importance with the development of housing permissions in relation to homeownership rates. The negative coefficient in the regression model shows
Elisabeth Springler and Karin Wagner 77
that housing permissions decrease with an increase in homeownership rates. This suggests the potential of a form of crowding out on housing supply between the public and private sectors. Especially in economies with large rental sectors and a strong supply of housing from public or private limited-profit organizations, which fulfill in general similar social aims as public housing manufacturers, an increase in homeownership rates would imply a strong structural change in overall housing provision as well as a shift in the goal of social housing. Another aspect that can be pointed out in this empirical study is the effect of housing finance manifested in the development of interest rates. Other research studies suggested that an enlargement and deepening of the overall housing finance system would lead to a decrease in interest rates, which in turn would affect homeownership rates positively. The results obtained from our empirical models for European economies and the US suggest that a decrease in interest rates in combination with tax reliefs promotes homeownership further. The coefficients and results did not change significantly, when the representative interest rate on new mortgages – as an indicator for changes in the housing finance structure – was added. From a theoretical point of view the extension of already existing models of the literature allows for changes in tax structures during the period from 1997 to 2006 and leads to a gain in the explanatory value of federal housing policies. For the first time (to our knowledge) a time series of the user cost of financing is available and provides the possibility to test for its influence. Despite this contribution on macroeconomic level, weaknesses similar to other empirical studies can be found when it comes to the explanatory value for regional and national differences. The structure of the model does not allow for a distinctive national analysis. In combination with the policy conclusions that can be derived from a macroeconomic perspective, it opens the door and shows the necessity for further comparative socio-economic studies in this field.
Annex: Tax models used for the calculation of cost of financing • deduction with a ceiling but no time limit e.g. Austria 1999: Interest payments are deductible as special expenses up to ATS 20 000 (IBFD, 2000). ia = i − MTR ∗ min (20 000/P, i)
78
Housing Market Challenges in Europe and the United States
e.g. Ireland, 1997: deductibility of interest paid on a loan applied for the purchase, improvement or repair of only or main residence, up to . . . the lower of 80 per cent (100 per cent for a first-time buyer) of the interest paid or IEP 5 000 for a married couple, the first IEP 200 (married) is not deductible, except for first-time buyers (IBFD, 1998). ia = i − 0.8 ∗ MTR ∗ min (5000/P, i) + 0.8 ∗ MTR ∗ min (200/P, i) • deduction with or without a ceiling, taxation of imputed rent e.g. Belgium, till 2004: interest on mortgages may be deducted from taxable income up to the total amount of income from immovable property. Imputed rental income from the taxpayer’s main dwelling-house included in the taxable income. The basic rate of the levy is 1.25 per cent for the Brussels and the Walloon regions (IBFD, 2002). ia = i − MTR ∗ min (i − 0.012, 0) e.g. Belgium, since 2005: Additionally, interest on a mortgage contracted on or after 1 January 2005 may be deducted up to a2000 for the first 10 years and a1500 thereafter. ia =
i − MTR ∗ min (2000/P, i − 0.0125), t ≤ 10 i − MTR ∗ min (1500/P, i − 0.0125), t 10
• deduction of a fixed fraction of the acquisition value e.g. Germany, 1996: a taxpayer acquiring or constructing a new owneroccupied dwelling receives a cash grant up to 5 per cent of the construction or acquisition costs in the year of completion or acquisition and in the following seven years, with a ceiling of DEM 5000 per annum. Acquisition or construction cost is the cost of the dwelling and the cost of the land (IBFD, 1997). ia =
i − MTR ∗ min (0.05, 5000/P), t ≤ 8 i, t 8
Germany, since 2000 no tax relief: The tax relief for owner-occupied dwellings abolished. Instead of tax relief, the taxpayer may currently be entitled to a tax-free cash grant for acquiring or constructing a new dwelling to be used by himself (IBFD, 2001). ia = i
Elisabeth Springler and Karin Wagner 79
• tax credit with indefinite duration e.g. Italy, 2003: A tax credit equal to 19 per cent of certain personal expenses is granted. The expenses which qualify for credit are among others: interest paid on mortgage loans contracted to finance the purchase of an owner-occupied dwelling-house, up to a maximum credit of a686.89 (IBFD, 2004). ia = i − min (686.89/P, 0.19i)
Additional regression results Table A4.1
Cost of financing – time series in %
Austria Germany France Nether- Belgium Denmark Finland Sweden lands 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
5.195 5.153 4.191 4.184 5.071 n.a. 4.409 3.687 3.625 2.996 3.418
5.925 5.525 4.593 4.190 5.263 4.798 4.783 4.071 4.037 3.353 3.763
5.919 5.582 4.640 4.609 5.394 n.a. 4.860 4.130 4.098 3.410 3.797
4.686 3.686 3.265 3.250 3.825 n.a. 3.054 2.614 2.622 2.115 2.174
6.493 5.753 4.752 4.749 5.593 5.131 4.987 4.181 4.153 2.931 2.397
3.805 3.350 2.726 2.675 3.092 2.829 2.867 2.443 2.423 1.916 2.146
UK
Greece
Italy
Spain
Ireland
Luxembourg
6.748 6.069 4.742 4.513 4.794 n.a. 3.342 3.068 3.300 2.987 2.931
11.270 8.090 6.165 5.500 5.359 4.789 4.660 3.628 3.617 2.868 3.256
9.314 6.050 4.205 4.728 4.922 n.a. 4.427 3.645 3.664 2.972 3.473
7.426 5.441 4.108 3.771 4.144 n.a. 3.966 3.308 3.280 2.722 2.839
5.405 4.719 3.680 3.617 4.354 n.a. 4.008 3.306 3.261 2.663 2.996
5.137 4.821 3.714 3.158 3.908 n.a. 4.066 2.511 3.158 2.477 3.051
Source: Authors’ calculations.
5.338 4.345 3.190 3.270 4.108 n.a. 4.233 2.957 2.990 2.294 2.766
3.306 4.258 3.213 3.157 3.478 3.361 3.622 2.992 2.978 1.624 1.793
Portugal United States 7.704 5.297 4.388 3.605 4.460 n.a. 3.907 3.106 3.093 2.681 3.164
3.150 3.108 2.576 2.765 2.981 2.483 2.282 2.192 2.337 2.358 2.634
80
Housing Market Challenges in Europe and the United States
Table A4.2
Fixed effects model with instrumental variables
Fixed-effects (within) IV regresssion Group variable: varland R-sq:
within = 0.3749 between = 0.1765 overall = 0.1507
Coef.
gdpcap –. .0909136 L1. –.043141 interestnew .0180565 population∼h L1. .0085948 mortgdebtr∼o L1. .0262298 costfin –.0290365 _cons 3.434435
102 15
Obs per group: min = avg = max =
1 6.8 9
Wald chi2(6) Prob > chi2
corr(u_i, xb) = –0.5152 Owner
Number of obs = Number of groups =
Std. Err.
z
P>|z|
= 6.80e+06 = 0.0000
[95% Conf. Interval]
.0769655 .0724821 0.136501
1.18 0.238 –.059936 –0.60 0.552 –.1852032 1.32 0.186 –.0086972
.2417632 .0989213 .0448101
.0047537
1.81 0.071 –.0007223
.0179118
.0163827 .0137941 .2639153
1.60 0.109 –.0058797 –2.10 0.035 –.0560724 13.01 0.000 2.91717
–.0583392 –.0020006 3.951699
sigma_u .21459137 sigma _e .01597688 rho .99448737 (fraction of variance due to u_i) F
test that all u_i = 0:
F(14,81) = 481.92
Prob > F = 0.0000
Instrumented: gdpcap L.gdpcap interestnew Instruments: L. populationgrowth L.mortgdebtratio costfin interestnew gdpcap L.gdpcap Source: Authors’ calculations.
Notes 1. The authors wish to thank Dieter Gstach for his valuable comments. 2. Additionally, in a few cases the explanatory value of institutional factors is acknowledged, but not integrated into the econometric modeling. Additionally, innovations in the area of housing finance and institutional factors facilitating or hindering the access of households to housing finance are mostly neglected. 3. Apart from the fact that also most European economies showed a strong upward trend in house prices over recent years, which seems to follow the trend observed in the US in the past, European economies were also affected by the financial crisis of the US because of the strong interrelations of banks
Elisabeth Springler and Karin Wagner 81
4. 5.
6.
7. 8.
9. 10.
11.
12. 13.
14. 15.
16.
and financial institutions on international financial markets (see Shiller, 2008, p. 8). See the discussion in Czerny and Wagner, 2003. This field of research includes numerous studies, which try to connect household characteristics, such as income, with the impact of interest rates. This is done especially in the light of the subprime market in the US analysis. See, among others, Barth et al., 2008; Bicakova and Sierminska, 2008; Weicher, 2007. See, in this respect, the discussion of economic analysis on the impact of interest rates on homeownership in Painter and Redfearn (2002, pp. 245ff and pp. 259ff). Zhu (2006) focuses in his empirical analysis on the impact of housing finance for house prices. Homeownership rates are not investigated. Furthermore, Mishkin (2007, p. 381) claims that ‘the innovations we have been seeing in mortgage markets have the potential to weaken the response of overall aggregate demand to changes in income driven by monetary policy, thereby altering the overall transmission of monetary policy shocks to the economy.’ These causalities presented can be interpreted as evidence of the important role of consumer spending and household debt for the macroeconomic development of the US economy in recent years. Innovative mortgage products enabled a continuous increase in household debt and facilitated economic growth via the transmission channel. See among others Levine, 1997, 2002; Allen and Gale, 2000. The models of Poterba, 1984 and Van den Noord, 2005, analyzing the impacts of taxation on house prices, are used as a starting point for the modeling in this chapter and are described in more detail below. Esteban and Altuzarra (2007) point clearly in their study on the Spanish housing market at the specific cultural and sociological dimension of the sector, which affects housing demand. In the case of Spain, on the one hand the demand for secondary housing is comparatively high, while on the other the rental sector is not highly developed, so that households lack any real choice between homeownership and renting a home. Sometimes 15 countries are included, depending on the model used. We applied such a restricted time horizon as we wanted to use balanced panels within the calculations. Starting earlier and employing interpolation or other estimation techniques of unbalanced panels seemed not that favorable to us as we wanted to test for ‘real’ influencing factors to the size of ownership rates. Furthermore, within the last 10 years the harmonization of interest rates and debt statistics took place, data of earlier years are even less harmonized. For the equations used to calculate ia and then if , see the Annex. The negative coefficient with GDP per capita is also found by other papers. For example, also Fisher and Jaffe (2002) found in their cross country data that homeownership rates are negatively correlated with GDP per capita. Another possibility for dynamic panels are the so-called pooled mean group estimators (PMG) for dynamic panels (Pesaran et al., 1999) where intercepts, short-run coefficients and error variances are allowed to differ across groups but the long-run coefficient is fixed. Similar to the dynamic GMM these models also did not seem suitable for our situation as the time horizon is quite short.
82
Housing Market Challenges in Europe and the United States
References Allen, F. and Gale, D. (2000) Comparing Financial Systems, Cambridge, MA: MIT Press. Amann, W. (2000) Schwerpunkt Subjektförderung. Auswirkungen und Optionen einer substantiellen Mittelverlagerung, Endbericht, FGW Publishing. Amann, W. (2005) Die Zukunft der Wohnbauförderung Chancen und Perspektiven in den Bundesländern, FGW Publishing. Anas, A. and R. Arnott (1997) Taxes and Allowances in a Dynamic Equilibrium Model of Urban Housing with a Size-quality Hierarchy, Regional Science and Urban Economics, Vol. 27, 547–80. Barth, J., T. Li, T. Phumiwasana, and G. Yago (2008) Perspectives on the Subprime Market, Miljen Institute, www.srrn.com Bicakova, A. and E. Sierminska (2008) Mortgage Market Maturity and Homeownership Inequality among Young Households: A Five Country Perspective, DIW Discussion Papers 778. Bourassa, S. and M. Yin (2006) Housing Tenure Choice in Australia and the United States: Impacts of Alternative Subsidy Policies, Real Estate Economics, Vol. 34, 303–28. Catte, P. et al. (2004) Housing Markets, Wealth and the Business Cycle, OECD Economics Department Working Papers, No. 394, OECD Publishing. Colton, K.W. (2002) Housing Finance in the United States: The Transformation of the US Housing Finance System, Joint Centre for Housing Studies Harvard University W02-5, July 2002. Commission of European Communities (2007) White Paper on the Integration of EU Mortgage Credit Markets, COM (2007) 807 final, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2007:0807:FIN:EN:PDF Committee on the Global Financial System (2006) Housing Finance in the Global Financial Market, CGFS Working Group Report, No. 26, BIS Publishing. Czerny, M. (2001) Wohnungswirtschaft vor neuen Herausforderungen, Vienna: WIFO Publishing. Czerny, M. and K. Wagner (2003) Structural Factors in the Austrian Housing and Real Estate Market, Focus on Austria, 3/2003, OENB, http://www.oenb. at/en/img/foa_20033_tcm16-8301.pdf ECB (2003) Structural Factors in the EU Housing Markets, ECB Publishing. ECB (2009) Housing Finance in the Euro Area, ECB Occasional Paper, No. 101, March. Egert, B. and D. Mihaljek (2007) Determinants of House Prices in Central and Eastern Europe, BIS Working Paper Series, Working Paper No. 236, September 2007, BIS Publishing. Esteban, M. and A. Altuzarra (2007) A Model of the Spanish Housing Market, Paper presented at the Macro Workshop of the Boeckler Stiftung, Berling, October 2007. European Mortgage Federation (2007a) Hypostat 2006, Annual Report. European Mortgage Federation (2007b) Factsheet Germany 2007. European Mortgage Federation (2008) Hypostat 2007, Annual Report. European Securitisation Forum (2006) ESF Securitisation Data Report, Spring 2006. Fisher, L.M. and A.J. Jaffe (2002) Determinants of International Home Ownership Rates, paper proposed to be presented at the 7th Annual Conference of the
Elisabeth Springler and Karin Wagner 83 Asian Real Estate Society jointly held with the American Real Estate and Urban Economics Association at Seoul, Korea, 4–6 July, 2002. Florida, R. (1986) Overview, in: Florida, R. (ed.), Housing and the New Financial Markets, State University of New Jersey, Centre for Urban Policy. Frankel, A. (2006) Erstklassig oder auch nicht: Finanzierung von Wohneigentum in den USA im neuen Jahrhundert, BIS Quarterly Review, March 2006, BIS Publishing. Fukao M. and M. Hanazaki (1986) Internationalisation of Financial Markets: Some Implications for Macroeconomic policy and for the Allocation of Capital, Economics Department Working Paper, No. 37, OECD, Paris. Girouard, N. et al. (2006) Recent House Price Developments: The Role of Fundamentals, OECD Economics Department Working Papers, No. 475, OECD Publishing. Goodhart, C. and B. Hofmann (2008) House Prices, Money, Credit and the Macroeconomy, ECB Working Paper Series, No. 888, April 2008, ECB Publishing. Goulder, L. (1989) Tax Policy, Housing Prices and Housing Investment, Regional Science and Urban Economics, Vol. 19, 281–304. Green, R.K. and K. Vandell (1999) Giving Households Credit: How Changes in the US Tax Code Could Promote Homeownership, Regional Science and Urban Economics, Vol. 29, 419–44. Green, R.K. and S.M. Wachter (2007) The Housing Finance Revolution, paper presented at the Symposium of Housing, Housing Finance and Monetary Policy at the Federal Reserve Bank of Kansas City, 30 August–1 September, 2007. Hoeller, P. and D. Rae (2007) Housing Markets and Adjustment in Monetary Union, OECD Economics Department Working Papers, No. 550, OECD Publishing. International Bureau of Fiscal Documentation (IBFD): European Tax Handbook, years 1996 to 2000 and 2002 to 2006. International Union for Housing Finance (n.d.) Fact Sheet USA, www.housingfinance.org. IMF (2008) World Economic Outlook, Housing & Business Cycle, April 2008. Lea, M. (2001) Overview of Housing Finance Systems, International Housing Finance Sourcebook 2000, International Union of Housing Finance Publishing. Levine, R. (1997) Financial Development and Economic Growth: Views and Agenda, Journal of Economic Literature, 35 (2), 688–726. Levine, R. (2002) Bank-Based or Market-Based Financial Systems: Which Is Better?, Journal of Financial Intermediation, 11(4), 398–428. Miles, D. (1994) Housing, Financial Markets and the Wider Economy, Chichester, New York, Brisbane, Toronto, Singapore: John Wiley & Sons. Mishkin, F. (2007) Housing and the Monetary Transmission Mechanism, Working Paper Finance and Economics Discussion Series, Division of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C. Mooslechner, P. (1994) Das System der Wohnbaufinanzierung in makroökonomischer Perspektive, in: Korinek, K. and E. Nowotny (eds), Handbuch der gemeinnützigen Wohnungswirtschaft, Wien: Orac, pp. 185–204. OECD (a) The tax-benefit Position of Production Workers, OECD Publishing, editions 1996 and 1997.
84
Housing Market Challenges in Europe and the United States
OECD (b) The tax-benefit Position of Employees, OECD Publishing, editions 1998 and 1999. OECD (c) Taxing Wages, OECD Publishing, editions 2000 to 2006. OECD (2008) Economic Outlook, Vol. 2008/1 No. 83, June, OECD Publishing. Office of Federal Housing Enterprise Oversight (2008) House Price Index, Report 2nd Quarter 2008, http://www.ofheo.gov/media/pdf/2q08hpi.pdf Painter, G. and Redfearn, C. (2002) The Role of Interest Rates in Influencing LongRun Homeownership Rates, Journal of Real Estate Finance and Economics, 25(2/3) 243–65. Pesaran, M.H., Y. Shin and R.P. Smith (1999) Pooled Mean Group Estimation of Dynamic Heterogeneous Panels, Journal of the American Statistical Association, 94, 621–34. Poterba, J.M. (1984) Tax Subsidies to Owner Occupied Housing: An Asset Market Approach, Quarterly Journal of Economics, 99 (4), 729–52. Poterba, J.M. (1991) House Price Dynamics: The Role of Tax Policy and Demography, Brookings Papers on Economic Activity, Vol. 2, 143–203 (including discussion). Pozdena, R.J. (1988) The Modern Economics of Housing, Westport, Connecticut: Quorum Books. RealtyTrac (2009) U.S. Foreclosure Market Report, February 2009, www.realtytrac.com Shiller, R. (2008) Understanding Recent Trends in House Prices and Homeownership, Paper presented at the Symposium of Housing, Housing Finance and Monetary Policy, at the Federal Reserve Bank of Kansas City, 30 August–1 September, 2007. Springler, E. (2008) Wohnbaufinanzierung aus volkswirtschaftlicher Sicht, in: Lugger, K. and M. Holoubek (eds), Die österreichische Wohnungsgemeinnützigkeit ein europäisches Erfolgsmodell, Wien: Manz, pp. 281–91. Stagel, W. (2004) Wohnbauförderung und Wohnversorgung im internationalen Vergleich, ISW Endbericht, im Auftrag der oö. Landesregierung, Ressort Wohnbau. Tsatsaronis, K. and H. Zhu (2004) What Drives Housing Price Dynamics: Cross Country Evidence, BIS Quarterly Review, March 2004, BIS Publishing. Van den Noord, P. (2005) Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence, Économie international, Vol. 101, 29–45. Volk, B. (2008) Rmbs vs. Covered Bonds, in: European Covered Bond Council (ed.) European Covered Bonds Factbook. Warnock, V.C. and F.E. Warnock (2007) Markets and Housing Finance, www.srrn.com Weicher, C. (2007) The Long and Short of Housing: The Home Ownership Boom and the Subprime Foreclosure Bust, Networks Financial Institute Policy Brief, 2007PB-09, www.srrn.com Weller, C. and K. Sabatini (2007) On Shaky Ground: the US Mortgage Boom and its Economic Consequences, unpublished working paper, presented at the 11thconference of the research network Macroeconomic Policies of the Boeckler Stiftung, October. Wolswijk, G. (2005) On Some Fiscal Effects on Mortgage Debt Growth in the EU, European Central Bank Working Paper, No. 526, September. Zhu, H. (2006) The Structure of Housing Finance Markets and House Prices in Asia, BIS Quarterly Review, December, BIS Publishing.
5 The Rental Housing Market Dieter Gstach1
5.1 Introduction This chapter surveys some recent scholarly literature about the macroeconomic role of housing with regard to its contribution to an understanding of rental housing. It will also present major stylized facts about rental housing in the form of simple statistics, partly collected from other sources, partly compiled from the most recent wave of EU-SILC microdata collection for 2006. Together this material constitutes an informed starting point for a discussion about the potential role of rental housing in macroeconomic analysis. The focal point of the following account is the question of whether or not it is justified to proceed by assuming, like in Ortalo-Magné and Rady (2004), that ‘(t)here is no rental market.’ To increase contrast, the material to be presented will be organized according to two working hypothesis: 1) Renters and homeowners differ significantly in economic terms. 2) Rising house prices exert a negative long-run impact upon aggregate consumption via rising rents. The reputation of rental housing is not very good. TV gossip shortly before Michael Jackson’s death revealed that he even had to move into a rental unit.2 The stress was on rental, implying that the tenure type as such constituted the disaster, irrespective of the value of housing services associated with this rental. This reputation contrasts with the typical asset market approach to housing where it is assumed that individuals in equilibrium would be indifferent between renting or owning. In industrialized countries housing markets typically consist of four thirds:3 Two-thirds of owner-occupiers, one-third of landlords renting out and, as ‘inevitable byproduct,’ one-third of renters. This classification is related to but not identical to the fact that housing is, simultaneously, 85
86
Housing Market Challenges in Europe and the United States
an investment and a consumption good. However, in the ongoing discussion about the build-up and subsequent burst of the housing bubble the fourth third is rarely mentioned as if it were not directly affected and, at any rate, played no role in transmitting house price shocks to the wider economy. This is surprising for two reasons: One is that in the consumption demand perspective one may suspect changing house prices to translate into reverse changes in non-housing consumption because of lack of substitutability and strong income effects of housing consumption; second, because the burst of the housing bubble particularly affects renter households. More precisely: Those who were at the verge of ownership before mortgage markets deepened and lending standards were relaxed and who turned into owners afterwards. Clearly, these are exactly the households who are now, after the subprime crisis started to unfold, in greatest danger of arrears and foreclosures. The latter point is stated very clearly in Leamer (2007). It should be added, that among renter households, those at the verge of ownership are the ones relatively well off. Arguably, also most of the other renters would choose ownership, if only they were found creditworthy.4 This preference in practice arises, apart from idiosyncratic motives, simply because tenants through their term of lease must pay also for the landlords’ profit margin upon the house price itself, while this margin is based on a desired amortization period much shorter than a renter’s lifespan. This means, as often stated, that it is not the lack of income that prevents these renters from owning but credit constraints. The remainder of this chapter is organized as follows: In the next section some recent contributions to the literature about the macroeconomics of housing are briefly surveyed with respect to their rental market content. Section 3 discusses relevant data to describe rental markets and associated conceptual and availability problems. Section 4 presents some stylized facts based on EU-SILC data about differences between tenants and owners and the role of the rental in tenant household expenditures and Section 5 summarizes.
5.2 Selected macroeconomic literature A recent literature review on ‘Macroeconomics and housing’ by Leung (2004) suggests that the rental market is of no macroeconomic relevance: No hint regarding its mere existence can be found in the text. But also the other articles of volume 13/4 of the Journal of Housing Economics dedicated exclusively to the macroeconomics of housing, feature such an implicit
Dieter Gstach 87
irrelevance theorem. The situation improves only marginally in the special issue 24/1 of the Oxford Review of Economic Policy with the same focus. From the key article ‘Housing Markets and the Economy: The Assessment’ by Muellbauer and Murphy (2008), a distinct macroeconomic role of the rental sector cannot be inferred. The majority of the relevant literature is of an econometric nature, and studies particularly potential real estate wealth or collateral effects and, more generally, the monetary transmission mechanism in the face of changing housing market conditions. See e.g. Case (1992), Girouard and Blöndal (2001), Boone and Girouard (2002), Catte et al. (2004), Bernanke (2007), Mishkin (2007), André and Girouard (2008), Karakitsos (2008) and IMF (2008). The dominant impression arising from this rather scattered literature is that increasing house prices have positive effects on aggregate consumption and that the interest sensitivity of output has increased substantially along with mortgage market liberalizations. Only exceptionally, like in Kiss and Vadas (2005), is the possibility of a distinct macroeconomic role of the rental sector mentioned. Theoretically, as has been repeatedly noted, the impact of increasing house prices upon consumption is unclear and could as well be negative. See e.g. Catte et al. (2004), Al-Eyd et al. (2005) or Attanasio et al. (2009). The last point out that the observed co-movements of house prices and consumption might be due to common causality rather than anything else. This at least leaves the theoretical possibility of a relevant rent channel, through which rising house prices could ultimately unfold a dampening effect upon aggregate consumption. In Sheiner (1995) the working of such a channel has been studied by estimating the savings response of renters to rising house prices. As Sheiner argues, theoretically this response could be positive or negative: Positive, because due to rising house prices renters would be forced to increase savings for a larger future down-payment. Negative, because renters could also react with delaying the purchase of a house or skipping this idea completely. Her estimation results, based on micro-data from the 1984 US panel study of income dynamics, provide strong evidence for significantly increased savings of renters as a result of house price appreciation. As no evidence of a significant real estate wealth effect was available at that time, Sheiner conjectured that the net impact of rising house prices upon savings would be positive as well. In other words, the macroeconomic effects resulting from renters’ response dominates those from owner-occupiers. Sheiner might correct her assessment in the light of today’s evidence about positive collateral effects of rising house prices, but this would leave the relevance of her primary results untouched.
88
Housing Market Challenges in Europe and the United States
Another form of such a rental channel is studied empirically in Gstach (2007) based on panel data for several OECD countries over the 1970– 2004 period. Starting point of this study is a typical consumption function specification found in the above-cited literature on real estate wealth effects. This is augmented by real rental rates, rented market shares and an interaction term of these two variables (to allow for country specific results) to test for the significance of rental rates in explaining consumption. The results provide robust evidence for a negative contribution of rising rents upon consumption, coexisting with a significantly positive wealth effect quantitatively in line with the above-cited literature. Furthermore, assuming long-run proportionality between changes of rents and house prices (for given interest rates and other user cost relevant variables), his results indicate negative net effects from rising house prices upon aggregate (!) consumption, albeit insignificant ones for typical rented market shares. Similarly, Muellbauer and Murata (2008) find a negative long-run impact of land prices upon consumption for Italy and Japan. But they view the illiquid mortgage markets in these countries as ultimately responsible for this negative net effect, rather than a rental channel. A serious challenge for the analysis of a potential macroeconomic role of rental housing is the slow response of rental rates to changing house prices (see e.g. Muellbauer and Murphy, 2008). This implies for house prices as triggering variable, that the role of rental housing in the transmission of such changes to the wider economy could be revealed only in a long-run perspective. On the other hand, various policy measures influence rents rather more directly. For example, rent regulations, as described for major European countries in Haffner et al. (2007), which affect also sitting tenants, exert their influence comparably fast. This would make estimation of macroeconomic effects associated with changing rental market conditions possible without the need to assume questionable long-run equilibrium relationships. But such attempts would be subject to another major challenge: The notoriously poor quality of available rental data. This is stated for example in Muellbauer and Murphy (2008), who question the strategy of testing the existence of a house price bubble by analyzing the priceto-rent ratio. But while the deficiencies of housing related statistics for the owner-occupied segment are well documented (see e.g. the contributions of Murphy, 2008 and Eiglsperger, 2008 in this volume), the lack of reliable and internationally comparable data on rental housing markets is rarely discussed. So I conjecture that the statement by Leamer (2007) about the macroeconomics of housing, namely that there is ‘. . . too much theory and not enough data’, applies in particular to rental housing.
Dieter Gstach 89
However, such data are necessary to test whether indeed rental housing is irrelevant for macroeconomic analysis as this survey of some recent relevant literature suggests. Given the general neglect of rental rates in actual consumption function specifications in the applied macroeconomic housing literature, the latter displays a surprisingly widespread awareness of the negative impacts of rising house prices upon renters. Muellbauer and Murphy (2008), for example state the redistributing effects of rising house prices ‘towards the haves from the have-nots’, where the latter are meant to include renters. The same point is also made in Leamer (2007), who explicitly talks about the undesirable redistribution from renters to current homeowners in the course of rising house prices. Edelstein and Lum (2004) also discuss the ambiguity of wealth effects. As they note ‘. . . house price inflation widens the real income gap between the “haves” and the “have-nots” but the distributional effects depend on relative numbers of private and public housing owners and renters.’ Similarly in Goodhart and Hofmann (2008) who claim that a permanent increase in house prices ‘works in favor of a positive wealth or collateral effect (. . .) on consumption’, because of the ‘asymmetry between gainers and losers’, the latter being smaller and consisting of tenants and first-time buyers. However, this awareness only goes half-way towards a macroeconomic conception of rental housing, because it does not include the possibility of a significant feedback from renters to aggregate consumption. A potential feedback channel of rental housing upon the wider economy is related to the size of this market: When home ownership is associated with less mobility, as found for example in Barcélo (2006), higher rented market shares might contribute to lower overall unemployment. Consequently, as indicated e.g. in Henley (1998), ECB (2003) or OECD (2005), one must ask whether the actual size of rental markets is too small compared with what should be regarded as healthy for labor mobility. As Oswald (1997) formulates it, ‘the large rise in European home-ownership may be the missing piece of the unemployment puzzle.’ Unfortunately a serious quantification of this potential feedback is a demanding econometric exercise, which may explain why it does not exist hitherto. The relationship between house prices and rents is a regular topic in housing research and is usually tackled from the perspective of capital asset pricing. In this perspective house prices equal the discounted present value of market rents under arbitrage in perfect capital markets. This suggests a rather simple relationship between the two variables and calls for a consideration of rents as additional explanatory variable in house price determination. This was stated earlier for example by
90
Housing Market Challenges in Europe and the United States
Muellbauer and Murphy (1997). As they note, ‘(t)his is an issue future modellers will need to address’ (p. 1722). But unanimous findings about this relationship exist only for the short run and these clearly reject the idea of proportionality between house prices and rents. It remains dubious, however, whether the often found stickiness of rents is the reason for, or merely a reflection of, this lacking proportionality. According to Meen (2006), it may be due to the fact that discount rates in the user cost formula are likely to be time-varying and that nominal interest rates affect house prices too. But also for the long run corresponding problems exist. To remedy the latter Murphy (2008) suggests that a ‘measure of credit constraints should also appear in equilibrium housing rent to house price equations.’ Looking at the relationship between house prices and rents from reverse angle it is tempting to speculate that rental prices will simply follow house prices, at least in the long run. But the corresponding cointegration analysis of Gallin (2004) does not yield clear-cut results in this respect. This calls for the introduction of a separate rental rate variable in econometric models along with house prices as in Gstach (2007), where both variables turned out to be significant in the long run also. A recent study on the economically relevant differences between owners and renters and wealth and income distributions within these two groups is Bicáková and Sierminska (2007). They investigate microdata from the Luxembourg Wealth Study and confirm earlier findings about the relative economic disadvantage of renters compared to owner-occupiers. Regarding the economic behavior of these two groups, Mishkin (2007) notes that they would change their spending in similar ways in response to movements in house prices. This clearly contradicts the findings by Sheiner (1995), cited above. The increasing importance of the land component in house prices is discussed explicitly in Davis and Heathcote (2007) for the US and, more cursorily, also e.g. in Mishkin (2007). Of course, this issue also applies to rentals. But here this change appears in the form of rising economic rents, usually the main reason for rent controls as stated e.g. ECB (2003). For homeowners, instead, this land component in house prices is a different matter owing to their simultaneous ownership of the rent-creating factor and consumption of its services. So, particularly in the rented market segment, this issue calls for a distinct treatment of land and structure in the analysis of supply side effects associated with house price/rental rate changes. Rent indexation, i.e. the automatic increase of rentals in line with the CPI, is analyzed in Gstach (2006). It is shown, unsurprisingly, that
Dieter Gstach 91
such indexation has an inflation accelerating consequence, which is reinforced by the use of imputed rents in the CPI. Rent controls in many countries were changed in the past to allow for more widespread use of this instrument. At the same time the comparable instrument of wage indexation has been largely abolished, also because of ample evidence about the undesirable consequences of such indexation. While the attention of economists at the moment may be rather more attracted by deflationary developments, the issue of rent indexation certainly deserves further analysis. Unlike in the above-cited macroeconomically oriented literature, rental markets are covered well in the microeconomic literature, populated by rational, optimizing agents. One strand of this literature has analyzed the (in)efficiency of rent control, see for example the overviews of Arnott (1998) or Klappholz (2004). Another currently more active branch of this literature investigates the role of credit constraints and the question of tenure choice. See e.g. the recent related contributions of Davis and Heathcote (2005), Jeske (2005), Ortalo-Magné and Rady (2005), Finocchiaro and von Heideken (2007), Silos (2007) or Calza et al. (2007). But the key question of the feedback of rental rates and rented market shares upon the wider economy is not tackled. It should also be noted, that in these models after removal of credit constraints a true preference for owner-occupation rather than rented accommodation requires the introduction of some market distortion. Of course this can be found easily in reality in form of the widespread subsidization of home ownership, for example, through deductibility of mortgage interest payments and/or renunciation of taxing imputed rents. Recent accounts of this are found in Wolswijk (2008) and Springler and Wagner (2008). Also this issue has been analyzed in a long series of papers initiated by Poterba (1984). But this does not capture the essential argument from the introductory section about the higher relative price of rental housing.
5.3 Data issues The main variables to characterize rental housing markets are rents, rented market shares, tenant turnover and vacancy rates of rental units. The relevance of each of these and some related variables is discussed below. Recent figures for these variables will be presented along the way to put things into perspective. The treatment of turnover rates for the two market segments is postponed to the next section, while rent-to-income ratios will be discussed here and in the following section.
92
Housing Market Challenges in Europe and the United States
5.3.1 Rental rates Despite their obvious significance, rental rate data comparable across time and countries are in fact hard to find. Country-specific housing market regulations gave rise to rather different national practices of defining the relevant cost items to be included in the rental measure. Furthermore, in many countries these regulations have changed considerably in recent decades. Whether it is energy, property management, parking space or other accommodation related items, these have been treated differently across time, as many breaks in national series tell, and across countries, as the ILO tables about national definitions of rental costs highlight. The problem of controlling adequately for the quality of housing, at any rate, cannot be used as explanation for this poor state of affairs, because various proposals for dealing with this issue can be found in the extensive literature covering the same problem for owner-occupied housing (see e.g. the discussion in Murphy, 2008). But there are also reasons to expect data availability on rental rates to be actually better than for house prices: The frequency with which the rental for an individual unit changes is much higher than for owneroccupied units. This has two reasons: First, the typical tenure duration of renters is considerably shorter than that of owner-occupiers. According to the EU-SILC data, for example, the median is seven years for renters compared with 18 years for owner-occupiers. This implies a turnover rate in rental units of almost three times that of owner-occupied units. Secondly, typical tenancy durations exceed the duration of the underlying rental agreements, which may be renegotiated after extension of each single contract. So, present rental housing market conditions are priced into monthly rates even faster than the typical tenure duration suggests. If at all, rental rate data are usually available only in index form containing no level information. This lack of level information, a typical feature also of house price statistics, can either be ignored by talking about growth rates only, or, with panel data covering several countries, by assuming some sort of structural homogeneity across national economies and addressing cross-country differences in price levels via unobserved components techniques. The latter approach is particularly useful because it stands to reason that cross-country differences in rental levels may affect non-housing consumption. Alternatively, rent-to-income ratios could provide relevant real level information for comparative international studies. For some countries this information is provided below. The OECD, the ECB and the BIS compile series for house prices that are regularly used in relevant studies, despite some lack of international comparability. But, to my knowledge, the only source for a listing of
Dieter Gstach 93
rental price indices for many countries over more than just a few years is the ILO. The ILO compiles nominal rental rate series as table 7F in the consumer prices category with yearly coverage typically starting 1970. The sources of these data are the national statistical offices, which publish domestic rental rates in one form or another. Although the ILO tries to harmonize the national series by occasionally choosing indices of more appropriate subcategories of ‘rent’, the extensive documentation reveals remaining shortcomings of the so constructed indices. Figures 5.1 and 5.2 show these ILO data as yearly growth rates of the deflated series (by country-specific harmonized index of consumer prices – HICPs), i.e. real rental growth: 12 EU countries are displayed for
US UK JP
4
2
0 ⫺2 1990 Figure 5.1a
1995
2000
2005
Yearly growth rates of real rents in per cent (US, UK and Japan)
Source: ILO data.
DE IT ES
4
2
0 ⫺2 1990
1995
2000
2005
Figure 5.1b Yearly growth rates of real rents in per cent (Germany, Italy and Spain) Source: ILO data.
94
Housing Market Challenges in Europe and the United States
the 1987–2007 period. The period 1970–1987 (not shown) was characterized by much higher volatility of real rental rates, but this is about the only commonality across countries to be found for this period. As the figures show, volatility of real rental rates has generally decreased further over the 1980s and the 1990s. The Figure 5.1a shows data for the US, the UK and Japan. What distinguishes them from the other countries is the consistent rise of real rental rates in the years around 2000 and an almost uninterrupted rise in real rentals for the whole period considered. The data for Germany, Italy and Spain are displayed in Figure 5.1b. In these countries there has been a
FR NL AT
4
2
0 ⫺2 1990
1995
2000
2005
Figure 5.2a Yearly growth rates of real rents in per cent (France, Netherlands and Austria) Source: ILO data.
CA DK CH
4
2
0 ⫺2 1990
1995
2000
2005
Figure 5.2b Yearly growth rates of real rents in per cent (Canada, Denmark and Switzerland) Source: ILO data.
Dieter Gstach 95
massive increase in real rents in the 1990s until about 2000 with yearly growth rates exceeding 4 per cent followed by little change thereafter. The striking commonality of the group of countries including France, the Netherlands, Austria, shown in Figure 5.2a, and (not shown) the Nordic countries is a more or less consistent and significant increase of real rents up to 2000 followed by marked slump, contrary to developments in the first group described above. This common feature is surprising because these countries are rather heterogeneous. The housing market characteristics of the Netherlands, for example, are usually described as more similar to those of the UK than to those of France. The French housing market in turn is usually considered incomparable to that of the Nordic countries. Figure 5.2b comprises Canada, Denmark and Switzerland, all of which exhibit comparably low variability and only a modest overall increase or even decrease in real rental rates between the mid-1990s and 2007. The case of Canada is particularly striking as it not only displays by far the lowest variability of real rentals, but also the only instance of continuous reduction of real rentals since the mid-1990s. 5.3.1.1 Price-to-rental ratio An important variable constructed via rental rates is the price-to-rental ratio. In the textbook housing market equilibrium, to be found for example in Miles (1994) or DiPasquale and Wheaton (1996), this ratio equals the reciprocal of the user cost of housing. So, deviations from this equality might be used to check for the existence of a house price bubble as suggested e.g. in Case and Shiller (2003). Even simpler instead, one might consider deviations of this variable from its own long-run trend as potential indicator for a real estate bubble, as in Ayuso and Restoy (2006). An overview of corresponding figures in this deviation form is found in André and Girouard (2008) for OECD countries. As a criterion in tenure decisions or to assess the rentability of real estate investment (for which it has to be compared with user costs) the original form is required. Girouard et al. (2006) reports corresponding figures for various countries. 5.3.1.2 Rent-to-income ratio A variable often used to address the question of affordability of housing is the rent-to-income ratio. Typical values of this ratio are around 25 per cent when disposable household income is used for comparison. More precise figures for this ratio for major EU countries derived from the EUSILC data-set will be given below. Based on evidence for stationarity of user costs (including house prices but not rents!) in the UK, Meen (1996) argues why the house-price-to-income ratio is not necessarily constant in
96
Housing Market Challenges in Europe and the United States
the long run. Given co-integration of rent and income, this would imply the non-existence of a stable long-run relationship between rent and income too. This is potentially relevant information for mortgage lenders as it renders simple extrapolative procedures for risk assessment flawed. 5.3.2 Rented market shares The macroeconomic literature on housing typically does not distinguish between renters, landlords and owners, although a few empirical studies cited above indeed have used rented market shares as explanatory variable. This neglect may lead to questionable results, if, for example, the consumption behavior of these groups differs significantly. This in turn could be due to various reasons such as different incomes, different composition of renter households compared with owner-households, different exposure to unemployment risk or different mortality rates, all of which have been investigated and found significant indeed. Therefore, the potential feedback of rental markets to the wider economy hinges crucially on the size of the rental market. Accounting for this in empirical analysis involving several countries, consequently requires the usage of an indicator for rental market size in the regression specification. The size of this rented market is usually expressed via the share of rented accommodations and varies widely between countries in a range of roughly 10 to 50 per cent. As documented below, rented market shares around the globe followed a secular decline in the past decades. One reason behind this development, apart from rising per capita incomes, is that economic policy in most countries fosters home ownership in various ways. This decline in rented market shares was rather steep for some countries (Italy, Spain,5 UK and Belgium) where corresponding figures dropped by more than 30 per cent between 1970 and 2007, with a more modest drop (less than 15 per cent) for many others (including the US, Canada and France). In Japan, Germany and Australia rented market shares did not change significantly during this period, while it increased considerably only in New Zealand and Finland. But, as the example of Finland nicely illustrates, figures about rented market shares should be interpreted cautiously in each case: While Catte et al. (2004) (citing the Royal Institute of Chartered Surveyors as source) report a rented market share for Finland in 2002 of 42 per cent, the Danish National agency for enterprise and housing (NAG) gives a figure of only 36 per cent for the pooled group of regular renters and all other non-owners. Furthermore, the latter source sees rented market shares in Finland in 2000 at 42 per cent, while ECB (2003) for the same year report 32. Likewise, figures reported for Germany in 2000 range from 53 per cent (Statistisches Bundesamt) to 60 per cent (ECB, 2003), while
Dieter Gstach 97
those for Sweden cover an even wider range from 39 (Catte et al., 2004) to 54 per cent (NAG). This gives an impression of the ambiguity of available data. It is not clear to what extent this problem could be remedied simply by stating proper definitions of reported data. It would help avoid some misunderstandings, for example, if it were mentioned whether rented market shares apply to households or individuals. For a typical rented market ratio of 30 per cent on a household basis and owner-occupier households’ sizes 15 per cent larger than for renters (see below), the corresponding rented market ratio would only be 27 per cent on an individual basis. Likewise, it would also be highly desirable if authors made clear whether their reported rented market shares apply to all contracts or just contracts with market rents. That this distinction is quantitatively important is illustrated by the rented market shares reported in O’Sullivan and de Decker (2007) for the competitive segment only, which, for some countries, are only a fraction of the corresponding figures in the literature cited above. In ECB (2003) it was noted that the shrinking of the rental sector may damage the functioning of the whole housing market. So it could be seen as good news, that the long downward trend in rented market shares in quite a few countries has come to a halt or even showed signs of reversal. Clearly, the evidence in this regard must be cautiously interpreted in the light of the above-mentioned data problems. Nevertheless, taken at face value, the figures for the recent 10 years suggest that this trend break in rented market shares applies for example to Denmark, Belgium (Ball, 2005), Spain (European Commission, 2005, p. 50), Japan, Norway, Australia (Catte et al., 2004) and the US (JCHS, 2007). According to Census Bureau data, for example, the US rented market share reached an all time low of 30.8 per cent in the fourth quarter of 2004 and since then increased again (32.2 per cent in first quarter 2008). 5.3.3 Vacancy rates The indicator most directly reflecting relative strength of demand and supply in housing markets are vacancy rates, which is true for owneroccupied housing and rental housing alike. Meen (2006), citing a study by Evans and Hartwich (2005) and the US Census Bureau as source, provides some vacancy rate estimates for the rental market: UK 3.4 per cent, Germany 8.2 per cent, France 6.8 per cent, Italy 20 per cent, US 10 per cent. It should be added that these vacancy rates often display significant volatility, as a look at the Census Bureau figures shows. This high volatility in the case of the rental market also reflects the sluggish price response (particularly downward) upon changing demand
98
Housing Market Challenges in Europe and the United States
conditions in this market. As Leamer (2007) points out, a strong rental market does not suffice to prevent a slowdown in new construction, when house prices tumble. Rather, changes of the vacancy rate seem to absorb much of the short-run fluctuations in demand. So, using the rate of new construction of rental units instead of vacancy rates may give a highly misleading impression of current market conditions. Unfortunately, internationally comparable vacancy rates for extended periods of time are compiled nowhere, leaving yet another important statistical job for the future.
5.4 Stylized facts from EU-SILC data This section presents some stylized facts about renters and rental housing based on the most recent EU-SILC micro-data for 2006. This data set about social indicators and living conditions (SILC) also contains a useful set of tenure related information. While EU-SILC data are collected in all EU member states, the former eastern European countries as well as some special cases like Luxembourg were skipped to improve data comparability. This yielded data covering 19 countries with a total sample of roughly 80 000 households consisting of 230 000 individuals. The average rented market share on a household basis in this sample is around 25 per cent, significantly lower than the average figure of around 30 per cent, which can be calculated from the figures reported by national statistical offices. This indicates a slight bias of the sampling design towards owner-occupiers. 5.4.1 Renters vs. owners Income differences between renters and owners are the key distinguishing feature of these two groups from an economic point of view. As EU-SILC data confirm, renter households on average have significantly lower disposable incomes than owner-occupiers. As Figure 5.3 shows the ratio based on the medians is roughly 3:2 between owner incomes (wide, light-shaded bars) and renter incomes (narrow, dark-shaded bars) for most countries.6 Particularly high differences are observed for the Nordic block (Denmark, the Netherlands, Sweden) where the ratio is more around 5:3 and comparably low differences for Italy and Spain. In any case, being renter rather than owner goes along with significantly lower incomes, although the causality is not clear and in fact may run in both directions. Note that a comparison of income levels for the UK and Denmark with those of the other countries is not meaningful because they are based on exchange rates from 2004 rather than on purchasing power parities for 2006.
99
Income (1000 Euro)
60 50 40 30 20 10 0 #(Owners): #(Renters):
DE 5095 4171
ES 7054 793
FR 5161 2681
IT 11459 2255
UK 4844 1457
Figure 5.3a Distribution of disposable household income (Germany, Spain, France, Italy and UK) Legend: Boxes represent the two interquartile ranges of incomes. Whiskers extend to the most extreme data points which are no more than 50 per cent of the interquartile range from the box. This typically covers around 90 per cent of the sample observations. Source: EU-SILC data.
Income (1000 Euro)
60 50 40 30 20 10 0 #(Owners): #(Renters):
AT 2238 1364
DK 2432 741
IE 2645 464
NL 4657 1924
SE 3021 1381
Figure 5.3b Distribution of disposable household income (Austria, Denmark, Ireland, Netherlands, Sweden) Legend: Boxes represent the two interquartile ranges of incomes. Whiskers extend to the most extreme data points which are no more than 50 per cent of the interquartile range from the box. This typically covers around 90 per cent of the sample observations. Source: EU-SILC data.
100 Housing Market Challenges in Europe and the United States
It is worthwhile noting that this distinction is not only a European phenomenon. Statistics New Zealand, for example, also reports incomes by tenure type based on ‘residual income’. This is defined as disposable equivalized income after deduction of mandatory housing related costs, in particular rents and mortgage repayments. Using the corresponding values of this variable to construct the ratio between renter and owner incomes leads to very similar results compared to the above for Europe (US$31 400 for owner households versus US$22 600 for renter households in 2004). Since renters, according to EU-SILC definitions, include all kinds of non-owners, one may suspect that the subgroup of renters actually paying market rents would be much better off. However, the exclusion of subsidized renters would not change the median income of renters significantly. The exceptions to this indifference claim are the UK and Ireland, where median incomes would increase to roughly a26 000 when excluding subsidized renters compared to the a20 000 reported in Figure 5.3. This indicates an extraordinarily high level of segregation in the UK rental market, which has no parallel in mainland Europe.7 Next to income the exposure to unemployment risk is another important distinguishing criterion between renters and owners. The reference group in constructing unemployment ratios consists of the selfemployed, employed and unemployed persons (all together referred to as ‘active’ below) and excludes, according to EU-SILC definitions, retired and other inactive persons. It turns out that the unemployment ratio among active renters over all countries is around 12.9 per cent as opposed to 6 per cent for active owners. So, not only is the average income of renters far lower than that of owner-occupiers, they are also subject to an a priori risk of becoming unemployed that is more than twice as high. This finding, of course, must not be interpreted causally, as if the ownership status as such would lower unemployment. The above-cited studies on labor mobility should have made this clear. Changing perspective one finds, furthermore, that the renter ratio among the unemployed is 36.7 per cent, significantly above the average of roughly 25 per cent of renters in the overall sample. These findings strengthen the impression of economic disadvantage of renters. One might suspect that these disadvantages are in fact much smaller when taking household size into account. But as EU-SILC data tell, this leads to only minor corrections of the above picture. Owner households have average equivalized size of 1.75, slightly above the size of 1.52 of renter households, but the median size of 1.5 is the same for both tenure types.
Dieter Gstach 101
The rented market shares in the different household size classes (measured in heads) are displayed in Figure 5.4. The share of renters for the class ‘single parent households’ (SP) is displayed separately.8 The situation in southern European countries (Italy, Spain, Portugal, Greece), represented via narrow, dark-shaded bars is depicted separately from that in northern EU countries represented via the wide, light-shaded bars. The repeatedly reported different housing habits for these two groups of countries suggest this distinction. As can be seen, the idea that rented market shares would decrease monotonically with household size is wrong, irrespective of North or South. In northern EU countries about 23 per cent of households with six persons or more live in rented accommodation compared with only 17 per cent for households with only five members. In southern Europe the rented market share already starts increasing again with households of four members. EU-SILC data about rented market shares by age are shown in Figure 5.5, again distinguishing northern from southern EU countries. EU-SILC reports (at most) two responsible persons per household. If two are reported, the older person is used for the age classification of the household. As can be seen, the rented market share for the northern European countries falls with the age cohort and reaches a minimum for the group of households with head aged 50–60 years. But for the groups with older heads the rented market share starts increasing again reaching 32 per cent for those aged 80 years and above.9 It can also be seen, that 70 Percentage of renters
EU North 60 EU South 50 40 30 20 10 0 #(North, all ): #(South, all):
1 2 3 13820 18818 8200 5690 8035 6183
4 8842 5747
5 3306 1625
Household size Figure 5.4 Rented market shares by household size Source: EU-SILC data.
6⫹ 1065 598
SP 1510 276
102 Housing Market Challenges in Europe and the United States
70 Percentage of renters
EU North 60 EU South 50 40 30 20 10 0 20 #(EU–North): #(EU–South):
4274 1042
40 60 9360 12516 11531 9096 3933 5455 5584 5178
80 5556 4410
3028 2542
Age Figure 5.5
Rented market shares by age group
Source: EU-SILC data.
this U-shaped figure of rented market shares across age profiles does not apply to southern Europe. There, instead, it seems that apart from the highest age group (more than 80 years) rented market shares would be ever increasing. 5.4.2 Rent to income ratio The rent concept of EU-SILC refers to the total monthly amount paid for the use of an unfurnished dwelling used as main residence of a household. Rentals may include payments for the use of a garage in connection with the dwelling but should exclude payments for electricity, heating, repairs and maintenance. As Figure 5.6 shows, renter households (wide, light-shaded bars) typically spend around 22 per cent of disposable income on rent. Restricting attention to tenancies of less than five years duration (narrow, dark-shaded bars) reveals, that for more recent tenancies this ratio is significantly higher than the average in almost all countries, implying rising rent-to-income ratios. This rise is in the 10 per cent range with Denmark being an outlier as more recently formed rental households pay almost 30 per cent more than the average of all rental households. The decrease of the rent-to-income ratio of more recent rentals in Ireland may be explained also with the extremely high starting level compared to all other European countries. The Netherlands and Sweden did not report tenancy durations.
103
Rent/income in %
50 40 30 20 10 0 #(All): #(Dur<=5):
DE 3581 968
ES 549 283
FR 1524 795
IT 1542 557
UK 699 356
Figure 5.6a Distribution of rent-to-income ratio in per cent (Germany, Spain, France, Italy, UK) Legend: Boxes represent the two interquartile ranges of incomes. Whiskers extend to the most extreme data points which are no more than 50 per cent of the interquartile range from the box. This typically covers around 90 per cent of the sample observations. Source: EU-SILC data.
Rent/income in %
50 40 30 20 10 0 #(All): #(Dur<=5):
AT 1110 304
DK 741 81
IE 104 761
NL 1924
SE 1337
Figure 5.6b Distribution of rent-to-income ratio in per cent (Austria, Denmark, Ireland, Netherlands, Sweden) Legend: Boxes represent the two interquartile ranges of incomes. Whiskers extend to the most extreme data points which are no more than 50 per cent of the interquartile range from the box. This typically covers around 90 per cent of the sample observations. Source: EU-SILC data.
104 Housing Market Challenges in Europe and the United States
Combining a typical figure for more recent rent-to-income ratios of 25 per cent together with average rented market shares of 30 per cent in industrialized countries would amount to a minimum value of roughly 5.5 per cent of aggregate consumption taking the form of rental payments. This minimum would be reached only when the average consumption propensities of renters and owners were equal. But the actual weights of rents in national CPIs are distributed around this 5.5 per cent value in a fashion that cannot be explained by country-specific rented market shares and/or usage of imputed rents to cover owneroccupied housing. This again reflects the above-mentioned problem of very different conventions of defining rental rates across countries.
5.5 Summary This chapter highlights the potential importance of the rental market for the macroeconomics of housing and its neglect in large parts of the relevant literature. It is argued, in particular, that the possibility of negative effects from rising rentals rates upon aggregate consumption should be better investigated. A survey of some recent macroeconomic literature shows that the rental segment as a potentially important feedback route for changing housing market conditions is largely neglected. But this neglect is not based on econometric evidence but solely on the implicit assumption of the irrelevance of rental rates. The poor quality of available data on rental housing is identified as the main obstacle towards a better understanding of related problems. As recent EU-SILC micro-data confirm there exist significant socioeconomic differences between owners and renters: Owner-occupiers have roughly 50 per cent higher disposable household incomes than renters and are much less exposed to unemployment risk. So, the consumption patterns of these two groups are likely to differ which would imply non-neutrality of income redistributions between renters and landlords. The typical expenditure share of rents relative to renter households’ disposable income is found to be above 20 per cent. More recently-formed households already pay 25 per cent, indicating an adverse dynamic of decreasing housing affordability. Based on ILO data for major OECD countries it is shown that real rentals indeed have risen considerably over the past two decades. This implies a strong income shift from renters towards landlords and, in the light of the above argument, a possible dampening effect upon aggregate consumption. Meen (1996) argued that the relationship between housing and the macro-economy was poorly understood in the mid-1980s. Meanwhile,
Dieter Gstach 105
as the literature survey indicates, two particular shortcomings he mentioned (lacking appreciation of increased interest rate sensitivity and equity withdrawal effects upon consumption) have been removed. An updated account of the state of affairs should refer to the lacking appreciation of the particular role of rental housing feedback upon the wider economy. A prerequisite for corresponding research is improved data on rental housing. Therefore, the wish list of Murphy (2008) regarding desirable housing statistics for future research, which hitherto comprises house prices, (housing) wealth variables and (mortgage) credit conditions, should be extended to include at least two more items: rental rates and rented market shares. Without such data it is impossible to substantiate the often assumed irrelevance of rental housing. Or – as I would conjecture – its distinct role in the macroeconomics of housing.
Notes 1. Vienna University of Economics,
[email protected] 2. Author’s translation from the original German comment from ‘brisant’ on ARD, 24 January 2009, 5:40.p.m. 3. 33 per cent is a rough proxy for the mean rented market share in industrialized countries, with the US, UK, Canada and Australia being very close to this mean. For more on the share of the rented market see below. 4. In the absence of credit constraints only a minority of people would prefer rented accommodation. For example, those who have only temporary accommodation needs in a particular area and for whom the transaction costs involved in purchase and resale of a home are a deterrent. 5. Arévalo and Ruiz-Castillo (2004) report a decline in Spain from 30 per cent in 1970 to 10 per cent in 2000. 6. This difference does not yet take imputed rent into account, which, according to recommendations of the ‘Canberra group on household income measurement’, should be added to owner-occupiers’ disposable income. But Eurostat plans this inclusion for the future. 7. The UK poverty site http://www.poverty.org.uk/76 has more on this segregation and the associated outlier character of the UK and Ireland. 8. Only single-parent households with up to three children can be identified from the underlying equivalized household size figures. 9. To remedy the problem with top coding of age at 80, I simply defined everyone classified as being of age 80 to fall in the class aged 80–90 years.
References Al-Eyd, A., R. Barrell, E.P. Davis and O. Pomerantz (2005) Consumption, Income and Wealth: a Time Series Based Panel Data Study of the Euro Area, paper presented at the AIECE (Association des Instituts Européens de Conjoncture) working group on medium term and structural changes, Brussels, 2 November 2005.
106 Housing Market Challenges in Europe and the United States André, C. and Girouard, N. (2008) Housing Markets, Business Cycles and Economic Policies, presentation at the workshop Housing Market Challenges in Europe and the United States – Any Solutions Available?, Oesterreichische Nationalbank, Vienna, 29 September, 2008. Arévalo, R. and Ruiz-Castillo, J. (2004) The Rental Equivalence Approach to Nonrental Housing in the CPI. Evidence from Spain, WP 04-17, Departamento de Economia, Universidad Carlos III de Madrid. Arnott, R.J. (1998) Rent Control, in: Newman, Peter (ed.), The New Palgrave Dictionary of Economics and the Law, Vol. 3, New York: Palgrave Macmillan, pp. 305–10. Attanasio, O.P., L. Blow, R. Hamilton, and A. Leicester, (2009) Booms and Busts: Consumption, House Prices and Expectations, Economica, 76(1), 20–50, earlier as WP05/24, 2005, Bank of England. Ayuso, J. and Restoy, F. (2006) House Prices and Rents in Spain: Does the Discount Factor Matter?, Documentos de Trabajo No. 0609, Banco de España. Ball, M. (2005) RICS European Housing Review 2005, Technical report, Royal Institute of Chartered Surveyors, London, UK. Barcélo, C. (2006) Housing Tenure and Labor Mobility: A Comparison across European Countries, Documentos de Trabajo No. 0603, Banco de España. Bernanke, B.S. (2007) Housing, Housing Finance, and Monetary Policy, speech at the Federal Reserve Bank of Kansas City’s Economic Symposium, Board of Governors of the Federal Reserve System. Bicáková, A. and E. Sierminska (2007) Homeownership Inequality and the Access to Credit Markets, Working Paper No. 5, Luxembourgh Wealth Study. Boone, L. and N. Girouard (2002) The Stock Market, the Housing Market and Consumer Behaviour, OECD Economic Studies 35, 2002/2, 175–200. Calza, A., T. Monacelli, and L. Stracca (2007) Mortgage Markets, Collateral Constraints, and Monetary Policy: Do Institutional Factors Matter?, Discussion Paper No. 6231, Centre for Economic Policy Research, London. Case, K.E. (1992) The Real Estate Cycle and the Economy: Consequences of the Massachusetts Boom of 1984–1987, Urban Studies 29(2), 171–83. Case, K.E. and R.J. Shiller (2003) Is there a Bubble in the Housing Market?, Brookings Papers on Economic Activity 2, 299–362. Catte, P., N. Girouard, R. Price, and C. André, (2004) Housing Markets, Wealth and the Business Cycle, Economic Department Working Papers No. 394, OECD, Paris. Davis, M.A. and J. Heathcote (2005) Housing and the Business Cycle, International Economic Review 46(3), 751–84. Davis, M.A. and J. Heathcote (2007) The Price and Quantity of Residential Land in the United States, Journal of Monetary Economics 54, 2595–620. DiPasquale, D. and W.C. Wheaton (1996) Urban Economics and Real Estate Markets, Prentice Hall, Englewood Cliffs, NJ. ECB (2003) Structural Factors in the EU Housing Markets, Technical Report, European Central Bank, March, Frankfurt. Evans, A.W. and O.M. Hartwich (2005) Unaffordable Housing: Fables and Myths, Policy Exchange, London. Edelstein, R.H. and S.K. Lum (2004) House Prices, Wealth Effects, and the Singapore Macroeconomy, Journal of Housing Economics 13(4), 342–67, special issue Housing and the Macroeconomy: The Nexus.
Dieter Gstach 107 Eiglsperger, M. (2008) Residential Property Price Statistics for the Euro Area and EU Countries, Presentation at the workshop Housing Market Challenges in Europe and the United States – Any Solutions Available?, Oesterreichische Nationalbank, Vienna, 29 September. European Commission (2005) Housing Statistics in the European Union 2004, National Board of Housing, Building and Planning, Sweden and ‘Ministry for Regional Development of the Czech Republic’, Boverket. Finocchiaro, D. and V.Q. von Heideken (2007) Do Central Banks React to House Prices?, Sveriges Riksbank Working Paper 217, Sveriges Riksbank, Stockholm. Gallin, J. (2004) The Long-run Relationship between House Prices and Rents, Working Paper of the Finance and Economics Discussion Series 2004-50, Federal Reserve Board, Washington, D.C. Girouard, N. and S. Blöndal (2001) House Prices and Economic Activity, Economics Department Working Papers No. 279, OECD. Girouard, N., M. Kennedy, P. van den Noord, and C. André (2006) Recent House Price Developments: The Role of Fundamentals, Economics Department Working Papers No. 475, OECD. Goodhart, C. and B. Hofmann (2008) House Prices, Money, Credit, and the Macroeconomy, Oxford Review of Economic Policy 24(1), 180–205. Gstach, D. (2006) Der Einfluss steigender Wohnungsmieten auf den Konsum. Eine makroökonomische Untersuchung, Forschungsbericht, Arbeiterkammer Wien, Vienna/Austria. Gstach, D. (2007) The Housing Rental Rate Elasticity of Aggregate Consumption: A Panel Study for OECD Countries, European Journal of Housing Policy 7(4), 367–82. Haffner, M., M. Elsinga, and J. Hoekstra (2007) Balance between Landlord and Tenant? A Comparison of the Rent Regulation in the Private Rental Sector in Five Countries, presentation, International Conference of the European Network for Housing Research, Rotterdam. Henley, A. (1998) Residential Mobility, Housing Equity and the Labor Market, Economic Journal 108, 414–27. IMF (2008) Housing and the Business Cycle, World Economic Outlook, April 2008. JCHS (2007) The State of the Nation’s Housing 2007, Report, Harvard University, Joint Center for Housing Studies. Jeske, K. (2005) Macroeconomic Models with Heterogeneous Agents and Housing, Economic Review (Q4), 39–56, Federal Reserve Bank of Atlanta. Karakitsos, E. (2008) The Lessons from the Current Crisis for Macro-theory and Policy, presentation at the workshop ‘Housing Market Challenges in Europe and the United States – Any Solutions Available?, Oesterreichische Nationalbank, Vienna, 29 September, 2008. Klappholz, K. (2004) Rent Control, in: John Eatwell, Murray Milgate and Peter Newman (eds) The New Palgrave, Vol. 4, Basingstoke: Palgrave, pp. 143–6. Kiss, G. and Vadas, G. (2005) The Role of the Housing Market in Monetary Transmission, MNB background studies, 2005/3, Magyar Nemzeti Bank. Leamer, E.E. (2007) Housing IS the Business Cycle, presentation at symposium, Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, 30 August–1 September. Leung, C.K.-Y. (2004) Macroeconomics and Housing: A Review of the Literature’, Journal of Housing Economics 13(4), 249–67.
108 Housing Market Challenges in Europe and the United States Meen, G. (1996) Ten Propositions in UK Housing Macroeconomics: An Overview of the 1980s and early 1990s, Urban Studies 33(3), 425–44. Meen, G. (2006) Ten New Propositions in UK Housing Macroeconomics: An Overview of the First Years of the Century, ICHUE Discussion Paper No. 2, University of Reading. Miles, D. (1994) Housing, Financial Markets and the Wider Economy, New York: John Wiley and Sons, Inc. Mishkin, F.S. (2007) Housing and the Monetary Transmission Mechanism, NBER Working Paper 13518, National Bureau of Economic Research, Cambridge/MA. Muellbauer, J. and K. Murata (2008) Consumption, Land Prices and the Monetary Transmission Mechanism in Japan, paper presented at the workshop ‘Japan’s Bubble, Deflation and Long-term Stagnation’, Columbia University, March 21. Muellbauer, J. and Murphy, A. (1997) Booms and Busts in the UK Housing Market, Economic Journal 197(445), 1701–27. Muellbauer, J. and Murphy, A. (2008) Housing Markets and the Economy: the Assessment, Oxford Review of Economic Policy 24(1), 1–33. Murphy, A. (2008) House Price and Housing Market Data: A User’s Perspective, Presentation at the workshop Housing Market Challenges in Europe and the United States – Any Solutions Available?, Oesterreichische Nationalbank, Vienna, 29 September, 2008. OECD (2005) How Persistent Are Regional Disparities in Employment? The Role of Geographic Mobility, OECD Employment Outlook 2005, Chapter 2. Ortalo-Magné, F. and S. Rady (2004) Housing Transactions and Macroeconomic Fluctuations: a Case Study of England and Wales, Journal of Housing Economics 13(4), 287–303. Ortalo-Magné, F. and S. Rady (2005) Housing Market Dynamics: on the Contribution of Income Shocks and Credit Constraints, Discussion Paper 2005-01, University of Munich. O’Sullivan, E. and P. de Decker (2007) Regulating the Private Rented Housing Market in Europe, European Journal of Homelessness 1, 95–117. Oswald, A.J. (1997) The Missing Piece of the Unemployment Puzzle, Inaugural Lecture, Department of Economics, University of Warwick. Poterba, J.M. (1984) Tax Subsidies to Owner-occupied Housing: An Asset Market Approach, Quarterly Journal of Economics 99(4), 729–52. Sheiner, L. (1995) Housing Prices and the Savings of Renters, Journal of Urban Economics 38(1), 94–125. Silos, P. (2007) Housing Tenure and Wealth Distribution in Life Cycle Economies, B.E. Journal of Macroeconomics 7(1). Springler, E. and K. Wagner (2008) Financial Systems, Structural National-wide Differences in Housing Finance, presentation at the workshop Housing Market Challenges in Europe and the United States – Any Solutions Available?’, Oesterreichische Nationalbank, Vienna, 29 September, 2008. Wolswijk, G. (2008) Fiscal Aspects of Housing in Europe, presentation at the workshop Housing Market Challenges in Europe and the United States – Any Solutions Available?, Oesterreichische Nationalbank, Vienna, 29 September, 2008.
6 Housing Markets, Business Cycles and Economic Policies Christophe André and Nathalie Girouard1
6.1 Introduction From the mid-1990s to 2007, the vast majority of OECD countries experienced an exceptional expansion of their housing markets, both in terms of magnitude and duration. Deviating from historical patterns, the latest housing upswing has been disconnected from the business cycle, as the economic downturn of the early years of the century was not accompanied by a slowdown of housing markets. At the same time, household debt has reached record levels in many countries, largely as a result of low interest rates and a wide range of financial innovations on mortgage markets. Housing has contributed to the expansion of economic activity by enhancing the effect of interest rate cuts on economic growth: residential investment has been strong and wealth effects have supported private consumption, in particular in countries where mortgage markets were well developed. Econometric investigations point to a transmission of housing wealth to private consumption through the refinancing of mortgages and home equity loans – more generally housing equity withdrawal. Financial innovations have contributed to excessive lending and housing booms, which are at the root of the current financial turmoil. Tax systems favoring home ownership through tax deductibility or subsidies have sometimes exacerbated the problem. These developments have revived the debate on the role of monetary policy in asset price cycles and on the opportunity of introducing asset price evolutions in central banks reaction functions. While loose monetary policy has fuelled the housing boom, many excesses in credit expansion would have been prevented by adequate regulation and supervision. In particular, a dramatic growth in the share of assets held outside the traditional banking system, especially in investment banks, 109
110 Housing Market Challenges in Europe and the United States
structured investment vehicles, conduits and hedge funds, has weakened the control over the financial system. Creating conditions for a sustainable development of housing markets will imply reinforcing the supervision and regulation of the financial system. In the very short term, the resolution of the crisis requires stopping house prices from overshooting on the downside, restoring the normal functioning of financial markets and providing fiscal stimulus to the economy.
6.2 OECD housing markets in a historical perspective The main characteristics of real house price cycles from 1970 to the mid-1990s can be summarized as follows: the average cycle lasted about 10 years. During the expansion phase of about six years, real house prices increased on average by close to 40 per cent. In the subsequent contraction phase, which lasted around five years, the average fall in prices has been in the order of 25 per cent. Hence, at least since 1970, real house prices have fluctuated around an upward trend. This tendency is generally attributed to rising demand for housing space linked to increasing per capita income, growing populations, supply factors such as land scarcity and restrictiveness of zoning laws, quality improvement – to the extent it is imperfectly controlled for in house price indices – and comparatively low productivity growth in construction (Girouard et al., 2006a). The current house price cycle differs from historical patterns in three important respects. First, the size of the real price gains during the upturn is striking. In most OECD countries, the cumulative increases recorded in the recent episode have far exceeded those of previous expansions. Between 1995 and 2007, real house prices have been multiplied by about 31/2 in Ireland, 23/4 in the United Kingdom and have more than doubled in Australia, Denmark, Finland, France, the Netherlands, Norway, New Zealand, Spain and Sweden. They have generally peaked – between the end of 2006 and 2007 – at levels well above previous cyclical peaks. Second, the duration of the upturn has surpassed that of earlier phases of large real price increases in almost all countries. The latest upswing lasted around 10 years, compared with an average expansion of about six years in previous cycles. Third, during the last housing cycle upswing, there has been a disconnection between the evolution of real house prices and the business cycle. Figure 6.1 compares the deviation of OECD real house prices from their trend with the output gap, which measures the deviation of real GDP from its potential. Until 2000, there was a strong connection between the two indicators – peaks roughly coincided, while
Christophe André and Nathalie Girouard Percentage deviation from trend 20
111
Percent of potential
Real house prices (left scale)
Output gap (right scale)
8
15
6
10
4
5
2
0
0
⫺5
⫺2
⫺10
⫺4
⫺15
⫺6
⫺20 1975
⫺8 80
85
90
95
2000
05
Figure 6.1 OECD real house prices and the business cycle Note: Real house prices have been detrended using a linear trend. The OECD real house price index is an aggregate of the 17 countries for which time series start in the early 1970s, computed using purchasing power parity-adjusted GDP weights. Source: OECD calculations.
troughs in house prices lagged activity troughs by three to four years – but in the early years of this century, the divergence between the two indicators has been remarkable. Recently, the two measures have started to move in parallel again, the reversal in the real house price trend being accompanied by lower levels of activity. A summary measure commonly used to assess housing market conditions is the ratio of nominal house prices to per capita disposable income, a gauge of whether housing is within reach of the average buyer. As shown in Table 6.1, around the peak of the cycle, in 2007, priceto-income ratios, as well as price-to-rent ratios were substantially above their long-term averages in most countries – often exceeding their longterm averages by at least 25 per cent and sometimes by 60 per cent or more. Real house price developments can, to some extent, be explained by changes in fundamental house price determinants, such as real and nominal interest rates, mortgage credit conditions – in particular the lengthening of loan durations, which has loosened the borrowing constraint of households – and in some countries demographic evolutions and supply-side rigidities (Girouard et al., 2006a). It is, however, likely that many housing markets have been overheating in recent years, although the exact magnitude of the overvaluation is difficult to assess as structural changes in housing and mortgage markets are likely to be shifting equilibrium levels. In addition, some changes in fundamentals
112 Housing Market Challenges in Europe and the United States Table 6.1
House prices in real terms and relative to rents and income Per cent annual rate of change
Level relative to long-term average1 Price-torent ratio
2000–2006 United States Japan Germany France Italy United Kingdom Canada Australia Denmark Finland Ireland Netherlands Norway New Zealand Spain Sweden Switzerland Euro area3,4 OECD4
5.4 –4.3 –2.9 9.5 6.1 8.8 6.7 7.1 7.9 4.7 8.3 2.9 5.5 9.2 11.2 6.7 1.7 4.5 4.1
2007
20082
–0.4 –1.1 –1.2 4.9 3.1 8.4 8.5 8.8 2.9 5.6 –1.8 2.6 11.4 8.3 2.6 8.6 1.3 2.0 1.5
–6.2 –2.4 –2.7 –1.1 –1.1 –2.3 –3.4 0.4 –6.0 –2.4 –9.6 0.7 –5.2 –6.6 –3.7 0.0 0.2 –2.0 –3.8
Price-toincome ratio 2007
127 70 72 158 125 170 186 174 166 151 198 153 165 162 195 162 85 128 126
109 67 66 140 115 150 131 143 156 110 159 155 134 155 158 124 75 114 109
Note: House prices are deflated by the Consumer Price Index. 1. Long-term average = 100. 2. Average of available quarters where full year data are not available. 3. Germany, France, Italy, Spain, Finland, Ireland and the Netherlands. 4. Average of the countries shown in the table, using 2000 GDP at PPP weights. Source: National sources.
may be long-lasting while others may prove unsustainable. For example, a reduction of mortgage rates resulting from lower inflation expectations or increased efficiency and competition in financial markets may lower borrowing costs in a durable way and thereby increase equilibrium house prices. But a fall in mortgage rates driven by the underestimation of risks is likely to be reversed sooner or later as risk premiums are reassessed, leading to a correction in house prices. Over the past decade, housing demand has been boosted by an increasing ability of households to borrow. Debt relative to disposable income has reached record levels in a number of countries – e.g. 142 per cent in
Christophe André and Nathalie Girouard
113
the United States, 144 per cent in Spain, 186 per cent in the United Kingdom, 260 per cent in the Netherlands in 2006. As a result, the household sector may have become more vulnerable to adverse macroeconomic or asset price developments, even though total household wealth has increased even more dramatically than debt. This large stock of assets provides households with a financial cushion against a negative shock. However, households in a number of countries have leveraged balance sheets and the sensitivity of borrowers to house price and interest rate developments has likely increased. The recent downturn in the United States has caused some concern about homeowners with negative equity, which are more likely to default than others.2
6.3 The role of housing markets in monetary policy transmission The sharp reduction in interest rates at the beginning of the decade has generated a large increase in housing demand and stimulated residential investment, which reached high levels as a percentage of GDP in several countries. While long-term averages for housing investment relative to GDP are typically in the four to six per cent range, this ratio exceeded six per cent in 10 OECD countries at the peak of the cycle in 2006, reaching the highest levels in Greece (7.8 per cent), Spain (9.3 per cent) and Ireland (14 per cent). Low interest rates, in addition to their impact on the consumption of durable goods, affect household consumption via three channels: income, wealth and liquidity effects. A drop in interest rates lowers the financial burden on indebted households and hence increases their disposable income (income effect). However, revenues of creditor households are reduced and the global effect on consumption ultimately depends on the relative proportion of indebted and creditor households, the level and structure of their debts and assets, and their respective marginal propensity to consume. As indebted households tend to have a greater marginal propensity to consume than creditor households, it is likely that the income effect of a fall in interest rates is positive. A reduction in interest rates generally triggers a rise in asset prices and consequently an increase in the value of housing wealth, which can lead to an increase in household consumption (wealth effect). However, because households both own housing assets and consume the housing services deriving from them, capital gains to the owner are partly or fully offset by the higher discounted value of future imputed rents when house prices rise.3 Unlike a rise in equity prices, which can reflect an
114 Housing Market Challenges in Europe and the United States
increase in the economy’s expected productive potential, and thus of future income, higher house prices may simply reflect increased scarcity owing to higher demand, with no net change in either the quantity or the quality of the services they can provide. In that case, there would be no change in national wealth. Nevertheless, even if aggregate wealth is unchanged, house price increases usually affect the relative positions of specific groups of people – for example, of current homeowners vis-à-vis would-be home buyers. These wealth transfers can have macroeconomic effects if these categories’ propensities to spend differ. According to the life-cycle theory, consumption depends on households’ lifetime income and wealth. In order to keep their consumption level fairly constant over their lifetime, households tend to borrow when they are young and earn a relatively low income, to repay their debt as they get older and their income increases and finally to accumulate assets which will allow them to maintain their consumption level during retirement. However, in an uncertain world with imperfect financial markets, some households have only limited access to borrowing: even when available, uncollateralized consumer credit tends to be prohibitively expensive. Since housing assets constitute the most important form of collateral available to households, an increase in their value allows more borrowing to finance current consumption (liquidity effect). Estimates based on the life-cycle model suggest that the long-run marginal propensity to consume out of housing wealth is in the range of between 0.05 and 0.08 for Australia, Canada, the Netherlands, the United Kingdom and the United States, while it is between 0.01 and 0.02 in Italy, Japan and Spain and statistically insignificant in France and Germany. In the former five countries, the housing wealth effect appears to be larger than the financial wealth effect. The estimated long-run marginal propensities to consume out of financial wealth vary between 0.03 and 0.07 for Australia, Canada, Japan, the Netherlands, the United Kingdom and the United States and between 0.01 and 0.02 for France, Germany, Italy and Spain (Table 6.2). The empirical link between housing wealth and household consumption could result from a ‘pure’ wealth effect, whereby an increase in the value of household wealth would trigger a reassessment by households of the level of their permanent income and therefore of their desired consumption. But it could also result from a liquidity effect, the increase in housing wealth providing collaterals for additional borrowing by liquidity-constrained households. The size of the long-run effect of housing wealth on consumption appears to be positively correlated with indicators of mortgage market
Christophe André and Nathalie Girouard
115
Table 6.2 Short-term and long-term impact of financial and housing wealth on consumption Estimated short-term and long-term marginal propensities to consume out of financial and housing wealth Short-term
Australia Canada France Germany Italy Japan Netherlands Spain United Kingdom United States
Long-term
Housing
Financial
Housing
Financial
0.02 0.03 – – – 0.01 0.02 0.01 0.08 –
– 0.03 – 0.01 0.01 – – – 0.03 0.02
0.07 0.06 – – 0.01 0.01 0.08 0.02 0.07 0.05
0.03 0.04 0.02 0.02 0.01 0.07 0.06 0.02 0.04 0.03
Source: OECD estimates (Catte et al., 2004).
size, such as household mortgage debt ratios, suggesting that the mortgage market is pivotal in translating house price shocks into spending responses (Figure 6.2, panel A). The influence of housing market developments on consumption as well as the rapidity of the response of private consumption to changes in housing wealth depends on the extent to which homeowners are able to borrow against housing assets. Housing equity withdrawal (HEW) is usually calculated by subtracting the household sector’s residential investment from the net increment in its mortgage debt. HEW indicates the extent to which the household sector as a whole is extracting liquidity from the housing market. It is more likely to be positive when households are able to renegotiate existing mortgage loans or to contract second mortgages on the same property to take advantage of an increase in housing wealth. The size of HEW is closely correlated with the impact of housing wealth on consumption (Figure 6.2, panel B). The addition of a variable representing HEW in a consumption function allows an examination of the liquidity effect. The results support the hypothesis that rising HEW may have increased the consumption level in Australia, Canada, the Netherlands, the United Kingdom and the United States (Table 6.3). In these countries the marginal propensity to consume out of HEW – approximating a liquidity effect since this variable
116 Housing Market Challenges in Europe and the United States A: Mortgage debt ratios
B: Housing equity withdrawal
MPC out of housing wealth
MPC out of housing wealth 0.10
0.10 Netherlands
2
R ⫽ 0.63
0.08
2 R ⫽ 0.85 Netherlands
0.08
Canada
0.06
United Kingdom
Australia
Australia United Kingdom
Canada
0.06
United States 0.04
0.04
United States Spain
0.02
Spain
0.02
Italy
Italy, Japan
Japan Germany
0.00
France
0.00
Germany
France ⫺0.02
⫺0.02 0
10
20
30
40
50
60
70
⫺10
80
Mortgage debt 2002 (% of GDP)
⫺8
⫺6
⫺4
⫺2
0
2
4
HEW average level 1990-2002 (% of disposable income)
Figure 6.2 Marginal propensities to consume out of housing wealth and mortgage market indicators Note: MPC is for marginal propensity to consume; HEW is for housing equity withdrawal. Source: OECD estimates (Catte et al., 2004).
Table 6.3
Long-term impact of housing equity withdrawal on consumption
Marginal propensity to consume Financial wealth Housing wealth Housing equity withdrawal
Australia
Canada
Netherlands
United Kingdom
United States
0.04 – 0.63
0.06 0.05 0.63
0.11 – 0.20
0.08 – 0.89
0.03 – 0.20
Source: OECD estimates (Catte et al., 2004).
represents a cash flow – appears significant, and its magnitude ranges between 0.20 for the Netherlands and the United States and 0.89 for the United Kingdom. By contrast, no effect of HEW is found for France, Germany, Italy, Japan and Spain. Where the HEW variable is significant, it seems to capture most of the impact of housing wealth on consumption, suggesting that such impact is channeled to a large extent through greater access to liquidity. In fact, when HEW is included among the explanatory variables, the effect of housing wealth is generally no longer statistically significant. This is consistent with the fact that in the countries where HEW plays an important role – Australia, Canada, the Netherlands, the United Kingdom and the United States – it is also strongly correlated with house prices.
Christophe André and Nathalie Girouard
117
6.4 Structural factors behind differences in housing wealth effects Differences in the impact of changes in housing wealth on consumption across countries are linked to structural differences in national housing and mortgage markets. In particular, consumption responses to changes in housing wealth can be expected to be higher, ceteris paribus, in countries where financial markets provide easy access to mortgage financing and to financial products that facilitate equity withdrawal; where there is a high rate of owner-occupation, which implies a wider distribution of housing wealth; and where there are low housing transaction costs and housing wealth is exempted from capital gains taxes, both of which would encourage owners to perceive housing assets as more liquid. English-speaking countries, Denmark and the Netherlands appear to have the most complete mortgage markets in terms of the range of products offered, such as second mortgages and equity release products, as well as a choice between alternative interest rate adjustment and repayment structures. They are also able to cover a broader range of potential borrowers, including for example younger or older households, and borrowers unable to certify their income. Some of these product or borrower coverage options also exist in the traditionally less sophisticated markets such as Italy and Germany, but they are less common, having been introduced more recently. High loan-to-value (LTV) ratios allow borrowers to take out more debt. In a number of OECD countries there are regulatory ceilings on LTV ratios, and in most of them a loan’s LTV ratio influences its weighting for the purpose of capital adequacy requirements, so that high-LTV loans are more costly to fund. Changes in short-term interest rates have a stronger impact on consumption in countries where variable-rate mortgages prevail, or where fixed-rate mortgages can easily be refinanced. In some countries, refinancing is costly, making it unattractive when interest rate declines are small. In France, for example, fixed-rate mortgages have typically been available for a term of 15 years, but refinancing penalties amount to up to six months interest or 3 per cent of the balance that is being prepaid. In Germany, rates on mortgages are typically fixed for 10 years, and it is very difficult to refinance. By contrast, in Denmark and the United States, where most loans are also fixed-rate, penalty-free prepayment options are common, as mortgage lending is largely funded through callable mortgage-backed securities. In Ireland, Spain and the United Kingdom, mortgage rates are usually variable and interest rate changes feed through rapidly to changes in monthly service payments.
118 Housing Market Challenges in Europe and the United States
Potentially amplifying the importance of the mortgage-market structure is the extent of owner-occupation. Housing tenancy structures differ considerably across OECD countries. Broadly speaking, the share of owner-occupied housing is very high in southern European countries, relatively low in Germany, the Netherlands and in some Nordic countries and around two-thirds in most other countries (Catte et al., 2004). To some extent, these differences reflect tax incentives and differences in access to mortgage financing. However, some of the countries with the highest owner-occupation rates – such as Italy and Spain – are among those that have, or had until recently, the least developed mortgage markets. This suggests that other mechanisms for providing access to home ownership are available in these countries, like for example inter-generational wealth transfers. Thus, while high rates of owneroccupation may be a necessary condition for a housing wealth channel to open up, it is not a sufficient one, and the cross-country correspondence between owner-occupation and the sensitivity of consumption to real house prices is weak. Housing transaction costs also differ considerably across countries. Taxes – stamp duties, transfer and cadastral taxes, VAT – are one component. In addition, the fees to be paid to intermediaries can be set directly by regulations or be influenced by regulations on entry into the market for real estate services. Estimates of housing transaction costs for different countries on a comparable basis are difficult to find. Those available cover only a limited number of countries. The European Mortgage Federation (EMF, 2006) categorizes countries into high – Belgium, Spain, Italy, Greece – medium – France, Hungary, Germany and Poland – and low transaction costs countries – Sweden, the United Kingdom and Denmark.4 Higher transaction costs are likely to weaken the link between housing wealth and consumption transmission mechanism by making housing assets less liquid.
6.5 The role of taxation The taxation of capital gains on housing assets can be seen as having similar effects to transaction costs if the tax is levied when the gains are realized, as is usually the case. However, while most OECD countries apply capital gains taxes to residential property, a majority exempt owner-occupied dwellings that are the owner’s main residence, sometimes under the condition that the property is held for a minimum number of years – e.g. Austria, Belgium, Finland and Germany – or that the proceeds are reinvested – Spain and Portugal. In the few countries
Christophe André and Nathalie Girouard
119
where gains are taxed but no exemption exists for principal owners, such as Norway, Sweden and Austria, this tax may be perceived as a significant additional transaction cost. Other fiscal provisions, such as mortgage interest deductibility and the level of property taxes potentially play an important role too. In equilibrium, assuming that arbitrage exists between house ownership and rental, house prices are linked to the level of rents and the user cost of housing: P = R/(ia + τ + f − π) where P represents the level of house prices, R the level of rents, ia the after-tax interest rate on mortgages, τ the property tax rate, f the depreciation rate and π the expected growth rate of rents. The price level is a decreasing function of the after-tax mortgage interest rate and the property tax rate. Mortgage interest rate deductibility, by lowering ia , has a positive impact on the price level. A low property tax rate similarly tends to increase prices. The property tax rate is generally rather unrelated to the evolution of housing market prices. An exception is Denmark, where the property tax rate is revised annually, contributing to macroeconomic stabilization (Muellbauer, 2006). A tax system favoring home ownership through subsidies and tax deductibility not only raises the equilibrium price of housing but also increases price volatility (van den Noord, 2005). The demand and supply shocks affecting the housing market are amplified compared with the situation that would prevail in the presence of a neutral tax system. Mortgage interest deductibility does not in itself increase house price volatility. However, if expectations are extrapolative, which is likely in housing markets, an increase in house prices tends to generate expectations of further increases and a tax system which subsidizes housing tends to increase price volatility. Van den Noord (2005) shows that euro area countries with the highest subsidies for home ownership – Netherlands, Ireland, Finland, Spain – have the most volatile house prices. Moreover, mortgage interest deductibility encourages households to keep LTV ratios high by refinancing their mortgages, increasing the proportion of mortgages at risk of falling into negative equity – i.e. owning a property that is worth less than the amount of the loan to repay – in the case of a fall in house prices, an important feature of the current housing crisis in the United States (Ellis, 2008). Residential mortgages in the United States are generally non-recourse loans. In the event of default, the lender can foreclose the property used to secure the loan,
120 Housing Market Challenges in Europe and the United States
but can make no further claim on the borrower. Households with negative equity have thus an incentive to walk away from their homes. The resulting increase in foreclosures depresses house prices further, driving more homeowners into negative equity and eventually into foreclosure. As foreclosures increase, losses on mortgage-backed securities rise and the stress on financial markets intensifies. Therefore, the resolution of the crisis requires stopping house prices from overshooting on the downside and containing the number of foreclosures.
6.6 Asset prices and monetary policy The main objectives of monetary authorities in OECD countries are price stability, minimum deviations of output from its potential and financial stability.5 Insofar as asset prices, in particular those of housing, can provide useful information on future production and inflationary pressures, they can influence the monetary policy stance. In some countries, owner-occupied housing costs are incorporated into the consumer price index, but it is not always the case. In particular, the European Union harmonized index of consumer prices (HICP) excludes owneroccupied housing costs, whereas the most widely monitored consumer price indices in the United States take them into account. In times of booming housing markets, ignoring owner-occupied housing costs may lead to an underestimation of inflationary pressures. Furthermore, there are several ways of accounting for owner-occupied housing costs – e.g. imputed rents, user costs – and the method used can have a noticeable impact on consumer price indices (Ahearne et al., 2005; Cournède, 2005). Beyond these technical issues, the debate among economists and policy-makers on the opportunity for monetary policy to respond to asset price movements, further than taking into account what is implied by their short-term effects on output, employment and inflation, has been revived recently. The question is whether monetary policy should try to counter evolutions that would make asset prices drift from their fundamentals. For monetary policy to be efficient in containing asset price bubbles, three conditions need to be met (Kohn, 2006): First, monetary authorities need to be able to detect the development of a bubble at an early stage; second, a modest tightening of monetary policy needs to be able to control speculation; third, expected gains from avoiding a bubble in terms of medium-term macroeconomic performance must be substantial. This third condition can be expressed in a slightly different way: preventive action by central banks is only warranted if one considers
Christophe André and Nathalie Girouard
121
that it will prove more difficult to deal with the effects of a burst bubble. This could be the case, in particular if there are deflation risks or if nominal interest rates close to zero leave little margin for the central bank to reduce lending costs.6 Or, if developments in asset markets are perceived as putting the stability of the financial system at risk. Acting against the building of a bubble implies being able to identify the latter with a reasonable degree of certainty in real time. This assumption appears problematic, especially as it implies that central banks would have an information advantage over private markets.7 In addition, only about two-thirds of the major price increases which have taken place in OECD countries since 1970 have been followed by sharp declines.8 A large price increase is not necessarily associated with a bubble and can be the result of a favorable evolution of fundamentals. Fundamentals can, however, be difficult to observe or to interpret in real time. These difficulties are compounded by the lag in the reaction of the economy to monetary policy. Historically, price-to-rent and price-to-income ratios have tended to revert to their long-term average, albeit with relatively long adjustment delays (Girouard et al., 2006a; Ahearne et al., 2005). A significant deviation of these ratios from their historical norms implies a large probability of correction in the medium term. High investment rates relative to their long run average are also indications of potential overheating, as construction booms tend to end with sharp adjustments (Hoeller and Rae, 2007). Another feature of housing booms is that they are usually associated with strong credit growth (Ahearne et al., 2005; Detken and Smets, 2004; ECB, 2005). The increase in demand permitted by credit expansion puts pressure on prices, but the resulting increase in housing wealth also provides collaterals which can be used for further borrowing, a mechanism described as the financial accelerator (Bernanke et al., 1988). If all the indicators described above may help in assessing the state of the housing market, detecting bubbles in real time remains highly challenging. Assuming that a bubble could be identified with a reasonable degree of certainty, it is necessary to assess whether monetary policy is the most suited instrument to counter its development. Housing bubbles imply sharp increases – often double digit – in prices. In this situation, it is unlikely that moderate hikes in policy rates are enough to bring house prices back to their equilibrium level. An increase in interest rates which would be sufficient to impact significantly on house prices would probably be damaging for the rest of the economy.9 Furthermore, if speculation can lead to a misallocation of resources – e.g. overinvestment in the construction sector – an increase in interest rates aimed at
122 Housing Market Challenges in Europe and the United States
cooling asset markets might crowd-out investments that are economically sound and socially useful (Kohn, 2006). Finally, taking the evolution of asset prices into account when setting the monetary policy stance might reduce the clarity of monetary policy objectives in the eyes of the public and therefore make macroeconomic stabilization by the central bank more difficult (Nickell, 2005; Mishkin, 2007). Some central banks consider, however, that monetary policy has a role to play in controlling financial imbalances resulting from irrational developments in asset prices.10 Thereby, they might be able to improve macroeconomic performances in the medium term. In this approach, central banks aim at cooling asset markets and avoid the building up of bubbles by raising interest rates moderately when markets seem to be becoming too buoyant and credit growth looks excessive.11 Detken and Smets (2004) note that monetary policy has often been relatively accommodative at the end of an asset price cycle upswing. Taylor (2007) argues that if the federal funds rate had been set according to a Taylor rule in the past few years, much of the latest housing boom in the United States would have been avoided.12 However, high saving and strong reserve accumulation by central banks in surplus-generating emerging markets and large saving relative to investment in the OECD countries’ corporate sector have also contributed to the unusually low long-term interest rates of recent years (Ahrend et al., 2008). The building up of a bubble on financial or housing markets is a source of serious concern for monetary authorities because its bursting is likely to have a sizeable impact on growth and employment, and can even cause a deflationary spiral as was seen in Japan in the 1990s. Indeed, most large downturns in house prices have been associated with sharp slowdowns in GDP and household consumption growth, suggesting that housing cycles can amplify the business cycle considerably. According to an International Monetary Fund study on developed countries in the postwar period, on average contractions related to housing market busts last twice as long as those related to stock market crashes and associated output losses are twice as large – about eight per cent after three years against four per cent for stock market-related crises (IMF, 2003). Similarly, Cecchetti (2006), in a multi-country study, concludes that housing booms lower future growth prospects and entail considerable risks of particularly adverse outcomes. Detken and Smets (2004) and ECB (2005) also emphasize the high costs of recessions coupled with falls in house prices. Four main reasons explain the larger macroeconomic repercussions of recessions associated with housing crises, compared with those linked to stock market developments: First, housing wealth is much more widely
Christophe André and Nathalie Girouard
123
distributed than financial wealth, which is largely concentrated among the wealthiest households. These households’ consumption is presumably less sensitive to current income developments than that of less affluent households; second, as house prices are less volatile than equity prices, their variations are more likely to be perceived as permanent, and hence to affect consumption; third, housing wealth is more often used as collateral for borrowing and its level has therefore a larger impact on households’ financial constraints; fourth, housing crises often coincide with banking crises and a credit crunch, which delays the economic recovery. Hence, housing crises create important risks for the economic growth outlook. The higher housing wealth and mortgage debt is relative to GDP, the more serious the risk is. As noted earlier, these aggregates have increased considerably in recent years in most OECD countries. The current low level of inflation is also likely to lengthen significantly adjustments on housing markets and entails additional downside risks. Nominal house prices are generally sticky on the downside. Hence, real prices tend to adjust more easily when levels of inflation are higher. Based on the historical record, declines in real house prices, following large runups, have been more protracted when increases in the overall price level were small (Girouard et al., 2006a). Furthermore, when mortgage repayments are fixed, inflation tends to erode the value of household debts in real terms, thus improving the balance sheet of indebted households and facilitating the revival of domestic demand. Conversely, deflation – a fall in the level of prices – increases the value of debts in real terms, putting additional downward pressure on prices, with the ultimate risk of throwing the economy into a negative spiral, as happened during the Great Depression of the 1930s. Risks for growth, employment and financial stability resulting from housing market developments could justify preventive monetary policy interventions. Preventive action is warranted if expected gains from avoiding a bubble in terms of medium-term macroeconomic performance are substantial. Burst bubbles have often been associated with dismal macroeconomic performance and devising policy responses to such situations is generally tricky. In Japan, during the 1990s, economic policies were unable to lift the economy out of stagnation, presumably because of an inadequate initial monetary policy response and an inability to address the structural weaknesses of the financial system. Sweden was much more successful in getting out of a severe financial and economic crisis in the early 1990s (Ergungor, 2007). The current housing crisis in the United States is driving the world economy into the worst recession in decades and poses a major challenge
124 Housing Market Challenges in Europe and the United States
to policy-makers around the world, even more thorny as the global dimension of the crisis does not allow falls in domestic demand to be mitigated by increases in exports, as happened in previous country-specific or regional crises. In such circumstances, resolute fiscal policy stimulus is warranted, but it is not without consequences for the sustainability of public finances. These events would plead for preventive action to prevent housing booms. Nevertheless, while prudent monetary policy can help avoid fuelling asset price booms, prudential regulations also have a decisive role to play in that respect.
6.7 Systemic risk and financial system supervision The large macroeconomic impact of housing crises over history is notably linked to the fact that they are often concomitant with banking crises (Detken and Smets, 2004). An appropriate assessment of the capacity of the financial system to absorb a fall in house prices by its supervisors is therefore essential, particularly at times of deregulation of financial markets (Bernanke, 2002; White, 2006). This capacity essentially depends on the capital basis of financial institutions, the adequacy of provisioning – especially in relation to business cycle positions – and underwriting policies based on sound risk assessment. Banking regulations – in particular the capital requirements set out in the Basel II agreement – should guarantee that international banks hold adequate capital. However, in the past few years, there was a dramatic growth in the share of assets held outside the traditional banking system, in particular in investment banks, structured investment vehicles, conduits and hedge funds.13 Many traditional banks have sponsored offbalance-sheet vehicles to which they have contingency commitments to provide liquidity. These developments have increased the leverage of the financial system. Moreover, in a system where balance sheets are marked to markets, financial institutions eager to maintain their leverage are encouraged to take on more debt and make more investments as asset prices increase. In turn, this puts additional pressure on asset prices, improves bank balance sheets and encourages further lending. Capital standards, also pro-cyclical under the Basel II agreement, are unable to restrain credit expansion. As financial institutions search for ever more borrowers, they tend to ease lending standards, creating the conditions for the following crisis. Lax underwriting criteria were particularly evident in the United States subprime mortgage market – a market catering to borrowers that do not qualify for conventional loans, because of poor credit history and/or
Christophe André and Nathalie Girouard
125
insufficient guarantees to qualify for prime loans. From 2004 to 2006, this market expanded dramatically, reaching about 20 per cent of mortgage originations in 2006. This expansion proved unsustainable and the subprime market collapsed in 2007, with repercussions far beyond the housing market. At the current juncture, the amount of losses on mortgage-backed securities remains highly uncertain, but could be around $500 billion (latest updates based on Blundell-Wignall, 2008). These events have vividly shown that even though housing loans have historically been fairly safe – contrary to commercial property loans, which have contributed to several financial crises at the beginning of the 1990s – it is necessary to remain vigilant about underwriting conditions, especially in the presence of principal–agent problems. Banks also need to assess, in particular through stress-testing, the sensitivity of their portfolios to unfavorable market and macroeconomic developments. Similar exercises should be conducted at the system level by regulators, focusing on systemic risks in addition to risks to individual institutions (Geithner, 2008). More transparency in financial information and accounting, as well as improved financial literacy of the public, would also support sustainable development in housing finance. Provisioning policies are also key to financial stability. Unfortunately, provisioning tends to be pro-cyclical, with risks usually underestimated when the economy is expanding (Hoeller and Rae, 2007). In fact, most provisioning methods are backward-looking, i.e. future default rates are simply assumed to be similar to those observed in recent years. An exception is Spain, where a ‘dynamic provisioning system’ has been put in place, where estimated future losses depend on past losses evaluated over a whole business cycle (White, 2006). Rating agencies also tend to issue ratings on the basis of past defaults rather than perform forwardlooking assessments. Over recent years, this problem has probably been compounded by conflicts of interests – rating agencies receiving fees from the issuers of the securities they rated could have been incited to underestimate risks associated with these securities. As a result, residential mortgage-backed securities and collateralized debt obligations linked to subprime residential mortgage-backed securities were given top ratings. When investors realized that the risks associated with these securities were much greater than implied by their ratings, demand for these products collapsed, triggering a deep liquidity crisis, which prompted exceptional interventions by central banks on both sides of the Atlantic. In addition, government assistance had to be provided to several financial institutions in the United States (e.g. Bear Stearns, Fannie Mae and Freddy Mac, AIG) and Europe (e.g. Fortis, UBS, Northern Rock). Financial
126 Housing Market Challenges in Europe and the United States
market conditions worsened further after the collapse of Lehman Brothers in September 2008, driving the world economy into a severe recession. This episode has shown that if securitization can offer great opportunities for the development of housing finance, it also creates new challenges for the regulators and supervisors of the financial system. Between the mid-1980s and the early 2000s, the evolution of the structure of housing markets seems to have contributed to reducing the volatility of housing investment and GDP in the United States (Gordon, 2005). Housing finance in the United States has experienced sweeping changes since the Savings and Loans crisis of the 1980s. Before that crisis, mortgages were mainly granted by local and regional depository institutions, leading to a credit cycle driven by local financial conditions. Today, a national, market-based system of securitized mortgage finance accounts for about 70 per cent of outstanding mortgages, allowing a distribution of risks among a wide range of investors (Schnure, 2005; Green and Wachter, 2007). As a result, US housing markets have become less dependent on local financial conditions and had become more stable, until recently. In the past few years, however, large amounts of subprime mortgage securities were concentrated in highly leveraged financial institutions, leading to the ongoing housing and financial crises. Outside the United States, the evidence on the impact of mortgage market structures on the resilience of the economy to a shock on house prices is mixed. In the study quoted above, the IMF concludes that bank-based financial systems are more vulnerable to house price busts than market-based ones (IMF, 2003). But Cecchetti (2006) finds little evidence of a role of the financial structure in the real economic impact of asset price booms. Ultimately, any assessment of the respective merits of bank-based and market-based systems will strongly depend on the quality of the loan portfolio. Neither system is likely to perform well when there is excessive risk-taking and poor underwriting standards.
6.8 Conclusions Easy monetary conditions and financial innovations have contributed to excessive lending and housing booms, which are at the root of the current financial and economic crises. In some countries, tax provisions – especially mortgage interest deductibility – have also encouraged excessive borrowing. But many excesses in credit expansion would have
Christophe André and Nathalie Girouard
127
been prevented by adequate regulation and supervision of the financial system. It is therefore crucial that regulators keep pace with financial innovations, control the level of leverage of financial institutions, avoid pro-cyclical provisioning and capital standards, improve risk management and transparency and make sure that underwriting standards are appropriate. However, policy-makers want to avoid a regulatory overreaction, which would discourage legitimate activity and push risktaking vehicles into the shadows – e.g. towards hedge funds or offshore – where they would be even further away from the reach of regulators. Geithner (2008) and OECD (2008) have put forward proposals for reforming the US regulatory system in order to prevent future financial crises. Prudential regulations also need to be improved in many other countries and, given the global interconnections in financial markets, international cooperation between financial authorities is essential. In the short term, the resolution of the crisis requires stopping house prices from overshooting on the downside – in particular by ensuring that mortgages remain accessible to creditworthy borrowers and containing the number of foreclosures – and restoring the normal functioning of financial markets – through the provision of State guarantees, the separation of good from bad assets and the recapitalization of banks (Blundell-Wignall et al., 2008). Finally, given the devastating impact of the financial crisis on economic activity worldwide and the limited effectiveness of monetary policy in the current environment, it is essential to support overall demand through targeted fiscal stimulus.
Notes 1. This chapter draws extensively on research carried out in the OECD Economics Department, in particular Catte et al. (2004) and Girouard et al. (2006a, 2006b). The authors would like to thank Jørgen Elmeskov, Vincent Koen and other OECD colleagues as well as participants of the Austrian Central Bank’s Workshop ‘Housing Market Challenges in Europe and the US – Any Solutions Available?’ for their useful comments. The views expressed in this chapter are those of the authors and do not necessarily reflect those of the OECD. 2. Mortgage holders with negative equity were about 7 per cent at the end of 2006. A further 15 per cent drop in house prices from this point would bring the number of mortgage holders with negative equity to 21 per cent (Greenlaw et al., 2008). 3. The extent of the offset depends on the owners’ effective time horizon, that is, on whether they intend to sell their housing assets during their lifetime or to pass it on to their offspring. If current wealth holders fully internalize the welfare of future generations, so that their economic planning horizon
128 Housing Market Challenges in Europe and the United States
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
is effectively infinite, the expected cost of future imputed rents fully offsets the value of housing assets. High, medium and low transaction costs countries are defined as countries where transaction costs are respectively above nine per cent, between four and nine per cent, and below four per cent of property value. These objectives are generally considered by central banks, even though the mandates and scope of responsibility of these institutions vary across countries. Zero nominal interest rates do not completely eliminate the possibility of monetary policy action. The Bank of Japan adopted a policy of quantitative easing between March 2001 and March 2006, apparently with some positive results, although the magnitude of the impact on the economy is difficult to evaluate (Yamasawa, 2006; Spiegel, 2006). Several central banks, including the US Federal Reserve, are now implementing, or contemplating, unconventional monetary policy easing. Gurkaynak (2005) shows that, despite recent advances, econometric detection of asset price bubbles cannot be achieved with a satisfactory degree of certainty. To qualify as a major increase, the appreciation had to feature a cumulative real price increase equaling or exceeding 25 per cent over a five-year period. The detail of these episodes can be found in Girouard et al. (2006a). For example, Nickell (2005) estimates that to eliminate the surge in house price inflation in 2003–04 in the United Kingdom, a 300 basis points rise in interest rates during 13 quarters would have been necessary. This would have taken off more than 1/2 per cent of GDP in 2003. The influence of financial imbalances on future inflation is one of the justifications for the monetary pillar of the ECB (ECB, 2005). Bean (2003) and White (2006) also stress the need to take into account the consequences of monetary policies at a longer horizon than the two years or so generally considered in monetary analysis. Such ‘leaning against the wind’ policies have, in particular, been carried out recently by central banks in the United Kingdom, Australia and Sweden (Ahearne et al., 2005; White, 2006; Ingves, 2007). Ahrend et al. (2008) also provide evidence that policy rates persistently and significantly below what a Taylor rule would prescribe are associated with increases in asset prices, especially housing. Geithner (2008) estimates that in early 2007 the combined assets of such structures amounted to more than $10 trillion which is roughly equivalent to the total assets of the US banking system.
References Ahearne, A.G., J. Ammer, B.M. Doyle, L.S. Kole and R.F. Martin (2005) Monetary Policy and House Prices: a Cross-Country Study, International Finance Discussion Paper, No. 841, Board of Governors of the Federal Reserve System (US). Ahrend, R., B. Cournède and R. Price (2008) Monetary Policy, Market Excesses and Financial Turmoil, OECD Economics Department Working Paper, No. 597.
Christophe André and Nathalie Girouard
129
Bean, C. (2003) Asset prices, Financial Imbalances and Monetary Policy: Are Inflation Targets Enough?, Bank for International Settlements Working Paper, No. 140. Bernanke, B., M. Gertler and S. Gilchrist (1998) The Financial Accelerator in a Quantitative Business Cycle Framework, NBER Working Paper, No. 6455, National Bureau of Economic Research. Bernanke, B. (2002) Asset-price ‘bubbles’ and monetary policy, speech before the New York Chapter of the National Association for Business Economics, 15 October. Blundell-Wignall, A. (2008) The Subprime Crisis: Size, Deleveraging and Some Policy Options, OECD Financial Market Trends, Vol. 2008/1, No. 94. Blundell-Wignall, A., P. Atkinson and S.H. Lee (2008) The Current Financial Crisis: Causes and Policy Issues, OECD Financial Market Trends, Vol. 2008/2, No. 95. Catte, P., N. Girouard, R. Price and C. André (2004) Housing Markets, Wealth and the Business Cycle, OECD Economics Department Working Paper, No. 394. Cecchetti, S. (2006) Measuring the Macroeconomic Risks Posed by Asset Price Booms, NBER Working Paper, No. 12542. Cournède, B. (2005) House prices and inflation in the Euro Area, OECD Economics Department Working Paper, No. 450. Detken, C. and F. Smets (2004) Asset Price Booms and Monetary Policy, European Central Bank Working Paper, No. 364. ECB (2005) Asset Price Bubbles and Monetary Policy, European Central Bank Monthly Bulletin, April. Ellis, L. (2008) The Housing Meltdown: Why Did It Happen in the United States?, Bank for International Settlements Working Paper, No. 259. EMF (2006) Study On the Cost of Housing in Europe, European Mortgage Federation, Brussels. Ergungor, O.E. (2007) On the Resolution of Financial Crises: The Swedish Experience, Federal Reserve Bank of Cleveland Policy Discussion Paper, No 21. Geithner, T. (2008) Remarks at the Economic Club of New York, New York City, 9 June. Girouard, N., M. Kennedy, P. Van den Noord and C. André (2006a) Recent House Price Developments: The Role of Fundamentals, OECD Economics Department Working Paper, No. 475. Girouard, N., M. Kennedy and C. André (2006b) Has the Rise in Debt Made Households more Vulnerable?, OECD Economics Department Working Paper, No. 535. Gordon, R.J. (2005) What Caused the Decline in U.S. Business Cycle Volatility?, CEPR Discussion Paper, No. 5413. Green, R. and S. Wachter (2007) The Housing Finance Revolution, paper presented at Housing, Housing Finance, and Monetary Policy Symposium, sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, WY, 30 August– 1 September. Greenlaw, D., J. Hatzius, A. Kashyap and H.S. Shin (2008) Leveraged Losses: Lessons from the Mortgage Market Meltdown. US Monetary Policy Forum, New York, February. Gurkaynak, R. (2005) Econometric Tests of Asset Price Bubbles: Taking Stock, Finance and Economics Discussion Series, No. 2005–04, Board of Governors of the Federal Reserve System (US).
130 Housing Market Challenges in Europe and the United States Hoeller P. and D. Rae (2007) Housing Markets and Adjustments in Monetary Union, OECD Economics Department Working Paper, No. 550. IMF (2003) When Bubbles Burst, World Economic Outlook, April, Chapter 2. Ingves, S. (2007) Housing and Monetary Policy: A View from an Inflation Targeting Central Bank, paper presented at Housing, Housing Finance, and Monetary Policy Symposium, sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, WY, 30 August–1 September. Kohn, D. (2006) Monetary Policy and Asset Prices, speech at Monetary Policy: A Journey from Theory to Practice, a European Central Bank Colloquium held in honor of Otmar Issing, Frankfurt, Germany, 16 March. Mishkin, F. (2007) The Role of House Prices in Formulating Monetary Policy, speech at the Forecasters Club of New York, 17 January. Muellbauer, J. (2006) Housing and Personal Wealth in a Global Context, paper prepared for the United Nations–WIDER Project Meeting,Personal Assets from a Global Perspective, Helsinki, Finland, 4–6 May. Nickell, S. (2005) Practical Issues in UK Monetary Policy, 2000–2005, British Academy Keynes Lecture, 20 September. OECD (2008) Economic Survey, United States. Schnure, C. (2005) Boom–Bust Cycles in Housing: The Changing Role of Financial Structure, International Monetary Fund Working Paper, No. 05/200. Spiegel, M. (2006) Did Quantitative Easing by the Bank of Japan Work?, Federal Reserve Bank of San Francisco Economic Letter, No. 2006-28, October 20. Taylor, J.B. (2007) Housing and Monetary Policy, paper presented at Housing, Housing Finance, and Monetary Policy Symposium, sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, WY, 30 August–1 September. Van den Noord (2005) Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence, Économie internationale, No. 101, 29–44. White, W.R. (2006) Procyclicality in the Financial System: Do We Need a New Macrofinancial Stabilisation Framework?, Bank for International Settlements Working Paper, No. 193. Yamasawa, N. (2006), An Analysis: Quantitative Easing Policy Was Effective in Buoying the Japanese Economy, Japan Center for Economic Research Staff Report, August 24.
7 European Rental Markets: Regulation or Liberalization? The Spanish Case Montserrat Pareja-Eastaway1 and María Teresa Sánchez-Martínez2
7.1 Introduction Housing markets are currently undergoing a curious transformation both in Europe and in the USA as they reflect not only the current economic scenario but also offer evidence regarding the results of long-term housing policies. It is difficult to identify a common European pattern in the housing market as most countries have specific traditions and use different strategies in order to deal with their housing problems. This is particularly true when analyzing tenure; the European housing situation offers quite a diverse scenario in terms of the dominance of one form of tenure or another. Historically, the size of the rental market has evolved differently throughout Europe. In some cases, the percentage of dwellings in this housing segment has remained stable (i.e. Germany), while in others it has fallen dramatically (i.e. Spain). In many cases, rent controls or regulations have been implemented and removed over time as a mechanism to counteract the market outcome. Despite its tenure composition, the latest data on housing accessibility in Europe shows that neither the countries which have traditionally tended to encourage the rental sector (Central and Western Europe) nor the countries that have decided to provide mechanisms to facilitate access to homeownership (Mediterranean countries) have found a proper solution for current housing needs. Consequently, while many countries are making an effort to encourage homeownership through the privatization of part of their public housing stock, others, such as Spain, are promoting their limited rental markets (Trilla, 2001). A key issue in this debate, and in terms of the potential strategies that have to be adopted by all countries, is the need to reconsider public housing 131
132 Housing Market Challenges in Europe and the United States
provision focusing on greater economic efficiency without neglecting social objectives. In this context, the question posed in the title of this chapter reflects a dichotomy in housing markets, specifically in the rental sector. There are different reasons behind the need to regulate or control the market and it is necessary to ask to what extent the evolution of this market in terms of rent and size depends upon the characteristics of this regulation? In other words, are there any other variables that need to be taken into account in order to analyze and forecast the relative importance of rental markets in Europe? Undoubtedly, the answers depend on the role played by the rental dwellings, i.e., whether or not they perform a ‘social’ function in addition to other questions such as encouraging labor mobility or first-time housing access. If this is the case, which mechanisms ensure sufficient provision of this type of tenure? Therefore, the aim of this chapter is to analyze European rental markets in the light of the current academic debate regarding the provision of social housing. In order to do so, Section 7.2 will briefly examine the different housing contexts and trends in Europe over the last few decades. The following section, Section 7.3, will describe the main features of the theoretical analysis of the role of the tenure system in providing proper housing. Specifically, regulations in the rental sector will be examined in depth in order to assess their usefulness in terms of facilitating access to proper housing. A detailed analysis of the Spanish case is carried out in Section 7.4 as this reflects a historically varied combination of policy approaches to the rental sector. Finally, some conclusions are put forward in order to provide some guidelines for future action in terms of European housing policy development.
7.2 Housing systems in Europe and tenure The forces of both supply and demand determine the coexistence and predominance of different tenure typologies in the housing stock; family structure (i.e. single-family households, sheltered accommodation for the elderly, etc.), disposable income or labor requirements shape certain preferences in terms of the demand for tenure. Simultaneously, expectations regarding housing prices, property owner guarantees or interest rates play a key role in developing dwellings with selected tenancies on the supply side. Furthermore, historical policy approaches and public policy instruments aimed at facilitating affordable housing (i.e., interest rate subsidies, rent regulation and control) might also influence the
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 133
tenure composition in a certain period and also the percentage and volume of social rents. The different tendencies within the field of European housing systems play a key role in the contemporary analysis of tenure and the relationship between landlords and tenants. As Haffner et al. (2008) point out: ‘The balance between the interests of tenants and landlords not only depends on the way the system is designed, but also on the way the system interacts with the housing market in allocating advantages and disadvantages’ (Haffner et al., 2008, p. 231). The specific characteristics of supply and demand together with historical trends in housing polices in each country establish the main priorities of national housing policies. In particular, the combined role of public intervention and the forces of supply and demand in facilitating housing access to low-income families have shaped a particular stock of social housing with considerable differences between countries, and reflect notably different conceptualizations of the meaning of social housing. As Whitehead points out, the nature of social housing: ‘depend[s] on the impact of past policies and the strength of current organizational and regulatory structures’ (Whitehead, 2003, p. 151). Many authors have described the evolution of European housing systems by means of comparative analysis and the selection of representative case studies (Oxley and Smith, 1996; McCrone and Stephens, 1995; Harloe, 1985; Kemeny, 1995; Kleinman, 1996; Balchin, 1996; Boelhouwer and van der Heijden, 1992; Clapham et al., 1996; Haffner et al., 2008). Although, in most European countries massive construction of social housing, which was mostly public rental, took place after the Second World War in order to provide accommodation for low-income families. Since that period, the tenure pattern has been developed and transformed by each country according to its own socio-political system, the particular balance of the forces of supply and demand and its housing policy priorities. Despite this diversity, some common trends can be summarized as follows: a) The percentage of homeownership had increased by early 2000 in the majority of European countries compared to the situation at the beginning of the 1980s (see Figure 7.1). b) In general, a fall in terms of the size and quality of rental dwellings is detected all over Europe. c) Rental provision has lost its significance and changed its predominant form: at the beginning of the century, the main form of rental
134 Housing Market Challenges in Europe and the United States
d)
e)
f)
g)
provision was private renting, followed by the dominance of public (social) rental housing after the Second World War. Rents have mainly increased (except in the cases of Portugal and Denmark) in the long run mainly because of the increase in the cost of land for construction, the cost of housing services and the development of quality improvements. Real rent is less cyclical than house prices. Fluctuations are usually linked to regulatory reforms and are less sensitive to speculation or to macroeconomic cycles. A decrease in social programs aimed at facilitating access to housing has occurred in the majority of countries: the private rental sector has thus emerged as an alternative to social housing. Identification of affordability problems in certain social groups – mainly low-income families, the young and the elderly.
There have been multiple and diverse explanations for these general trends. Some authors (McCrone and Stephens, 1995) have highlighted the increase in the standard of living of citizens, who would therefore
Early 1980s
Early 2000s
Germany Netherlands Denmark Austria UK Sweden France Portugal Luxembourg Belgium Italy Hungary Greece Ireland Spain 0
20
40 %
Figure 7.1
Rented housing in several European countries
Source: Housing Statistics in the European Union 2004.
60
80
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 135
need less social housing and would consequently appreciate incentives to become homeowners. The demand for homeownership has certainly received a major boost from the increased wealth of private households but also to other factors related to the increased attractiveness of the sector itself. This has mainly been due to decreasing interest rates for mortgages, favorable tax and subsidy policies aimed at homeownership and increasing expectations of capital gains from rising housing prices until the real estate crisis in 2007. The strictness of rent controls may have reduced the supply rental accommodation as property owners receive fewer incentives to place the dwellings on the rental market given the considerable risk they face when there is no possibility to adapt to the changing housing context (ECB, 2003). This phenomenon is also associated with the diminishing force of the welfare state since the 1980s along with the lack of strong public policies in contemporary life (Kemeny, 2005; Navarro, 2004). The notable decrease in social policies aimed at guaranteeing access to housing might have meant fewer incentives designed to stimulate the (public) rental sector in the housing system. As can be seen in Table 7.1 and according to the latest available data, the percentage of social rental housing ranges from a low figure of 1 per cent in Spain to 35 per cent in the Netherlands. Germany exhibits the highest percentage of private rental housing. The private rental sector fulfils many (socially) useful functions in the housing market. Firstly, it provides housing for elderly, single or divorced households adjusted to their housing needs (i.e. less space, local Table 7.1
Housing tenure
Austria Netherlands Denmark** Sweden UK France** Ireland Germany Hungary Spain**
Owner occupation
Private rental
Social rental
55 54 52 59* 70 56 80 46* 92 82
20 11 17 21 11 20 11 49 4 10
25 35 21 20 18 17 8 6 4 1
Note: * Owner occupation include: Sweden, cooperatives, and Germany, shared ownership/equity ‘Genossenschaften’. ** Does not include ‘other’. Source: Whitehead and Scanlon (2007).
136 Housing Market Challenges in Europe and the United States
environment, etc.); secondly, it allows easy access to housing for young people and those belonging to flexible labor markets. Finally, it retains a residual role of housing provision for those unable to access homeownership or social housing. Furthermore, the existence of a considerable rental sector might even help to counteract overheated owner-occupied housing markets. Given the role performed by the rental sector, it is evident that it cannot be fully substituted by any other form of tenure ‘If private rental housing did not already exist, there would be good reasons to invent it’ (Priemus and Maclennan, 1998, p. 203). However, although the role played by the rental sector is clear, nowadays it is marginal, especially in countries such as Spain (ECB, 2003). As the sector has shrunk in the majority of EU countries, an in-depth review of legislation and regulations has taken place over the last few decades in order to identify the causes of this decline. One of the topics currently debated in academic and housing policy circles is the role played by the rental sector in the provision of affordable housing/social housing for the population (Van der Heijden, 2002). As will be seen later, the huge tenure imbalance in favor of homeownership displayed by certain markets such as Spain inevitably leads to a debate about the specific housing features and peculiarities that characterize the role of ‘social housing’ in these markets. Comparative analysis suggests that the majority of European countries achieve their social targets without considering homeownership. Nevertheless, countries such as The Netherlands or the UK, with a strong tradition of providing social housing throughout the public rental sector, show a turning point in the conceptualization of this provision. These countries show a downward trend in the direct provision of housing and, simultaneously, they stimulate competitiveness in terms of supply, both in quality and quantity, affecting all of the actors involved (public, private and NGOs). Indeed, the upward trend in the costs of housing provision, especially when oriented to low-income collectives, is not isolated from this fact. In summary, the diverse evolution of European housing systems has led to quite different scenarios for current housing policies. Tenure composition allows for different mechanisms to provide housing for certain collectives. Changes in rent regulations covering a wide spectrum, from strict regulation to open liberalization, may respond to the need for better provision but to what extent are these changes able to counteract the market trend? The following section will explore the pros and cons of both alternatives stressing the importance of other additional variables that might alter policy results.
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 137
7.3 Theoretical debate; regulation versus liberalization in the rental sector The current regulation scenario in the rental sector mainly reflects the historical evolution that has taken place in the housing systems of particular countries (Ellingsen and Englund, 2003). Several authors have developed in-depth and comparative analyses of the tenure system in Europe and the regulations affecting the rental sector in Europe (Haffner et al., 2008; Scalon and Whitehead, 2004; ECB, 2003; Priemus and Mandic, 2000; Van der Heijden and Boelhouwer, 1996). As a general conclusion, they have all stressed how regulations in the rental sector are rooted in the historical development of the country, the role played by its housing policies and its legislative framework. Besides, there is no clear advantage or disadvantage in using a more permissive or restricted rent regulation, as Turner and Malpezzi (2003) summarize, ‘Regulation per se is neither good nor bad. What matters are the costs and benefits of specific regulations under specific market conditions’ (p. 15). Public rental housing, social rental housing and private rental housing are, among others, the typical options in Western Europe. Evidently, they are not exactly the same nor do they provide the same type of service and their distinctive features are actually different from one country to another. Many authors have identified common dichotomies and dissimilar characteristics among them (Maclennan and More, 1997; Priemus and Dieleman, 2002; Priemus and Mandic; 2000): public vs. private provision; the targeting or not of low-income families; regulatory mandates or absence of regulatory control and profit and non-profit rental provision are, among others, distinctive features of these different typologies. As Mandic (2000) points out, these terms are to a great extent ‘context dependent’ and defined by the ‘institutional, legal, policy and financial frameworks in which a provider operates’ (p. 219). Statistics on social renting are hardly comparable as the national definitions of ‘social’ rental housing vary from one country to another. Given this situation, this chapter focuses mainly on the legal regulations that govern the private rental sector as part of housing policy instruments to facilitate housing access to certain demand segments. The main instrument used to intervene in the rental sector is rent control, especially in situations where there is a considerable excess of demand for housing, in particular, rental housing. Developed economies have used them as a measure of economic policy in order to regulate the market output, either completely or partially. The objectives, which underlie the adoption of these measures, go beyond the achievement of
138 Housing Market Challenges in Europe and the United States
specific housing policy targets (Haffner et al. 2008; ECB, 2003; Glaeser, 2003; Lind, 2001, 2003; Sánchez-Martínez, 2002; Arnott, 1995). Arguments in favor of using rent control mechanisms justify intervention related to the achievement of the following goals: • Better income distribution: Rent control implies income transfer from
•
•
•
•
property owners to tenants, similarly to the redistributive effects of taxes (to property owners) and subsidies (to tenants). Implicitly, through this objective rent control this assumes the capacity of property owners to obtain excessive profits given the existence of a monopolistic power. Overall, rent control reveals itself to be a progressive measure since it provides additional income for tenants who are, theoretically, poorer than property owners. Higher efficiency when the market is not ‘competitive’: The housing market and, specifically, the rental sector do not display the traditional features of a competitive market. In essence, the asymmetry in the information provided by landlords to tenants and the monopolistic power of proprietors mainly due to the sunk costs of tenants represent an attractive market for intervention. Greater price stability: Rent regulation favors the establishment of rents at a ‘stable’ and ‘fair’ level. During inflationary periods, rent control can act as a policy instrument to keep inflation down. Better security of tenure: Rent control has opposite effects on the security of tenure experienced by tenants. On the one hand, rent control will reinforce lease arrangements; on the other hand, it will provide economic security against the risk of ‘economic eviction’, i.e. being forced to move out given the rent rises in the local market. Reduced segregation: Instruments such as rent control may help to reduce segregation in urban areas, as it would integrate poorer and richer households within the same urban agglomeration.
Nevertheless, rent controls might also create some non-desirable side effects: • Reduced tenant mobility: Rent control can set rents below the market
level dissuading existing tenants from moving to other dwellings. This situation occurs when the dynamics of the housing market increase market prices for rental faster than the updates in regulated rents. Tenants thus prefer to remain in their current dwelling rather than search for another option. This phenomenon also contributes to a reduction
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 139
in vacant rental dwellings and, therefore, is partly responsible for an increase in searching costs for new entrants. • Reduction in new construction: Rent control provides arguments against the development of new construction aimed at the rental sector. Variations in demand cannot be translated directly in terms of supply reactions as rent controls prevent any possible update in prices. As Lind (2003) emphasizes, uncertainty about future legislation increases the risk for developers given the potential for changes in rent controls. • Reduced quality of rental dwellings: If rent controls exist, landlords receive few incentives to renovate their rental properties as they cannot pass on these expenses through an updated rent. Therefore, rental dwellings under rent control will probably display less quality than dwellings in the owner-occupied sector and there may even be pockets of low-quality housing in high-quality neighborhoods. As Glaeser (2003) stresses, the decrease in segregation may only affect those tenants already living in a rental flat under rent control but it does not affect the potential demand of the rental sector. Therefore, other instruments might be more helpful to reduce segregation rather than the use of rent control. • Creation of differences between existing and new tenants: Rent regulations exclusively reduce the expenditures of those households currently living in a rental dwelling. They establish distinctions between those already occupying a rental dwelling and those about to enter the housing market. This also affects urban segregation since rent control may cause an increase in segregation if it contributes to a reduction in the quality of regulated dwellings. Since rent control systems are heterogeneous and make use of many different instruments in order to regulate the rental sector, it may be almost impossible to generalize about their effects. The perspective adopted when analyzing rent control, assuming a perfectly competitive housing market or recognizing the limitations on perfect competition, will undoubtedly affect the possibilities of improving welfare in a welldesigned rent control program (Arnott, 2003). As Turner and Malpezzi (2003) have summarized, although empirical findings suggest no (or a very poor) redistributive transfer between landlords and tenants as a consequence of rent control, the final outcome of many rent regulations vary in line with market conditions and industrial organization. Therefore, diversity in the output efficiency of rent controls is the rule and not the exception.
140 Housing Market Challenges in Europe and the United States
However, there are certain fundamental aspects to be considered when analyzing rent control systems: Firstly, regulations governing how the initial rent will change in a multi-year contract; secondly, the control on the initial rent negotiated in a new contract; and finally, regulations governing contract termination (eviction). Arnott (1995) distinguishes three main types of rent regulation: • First generation rent control: This type of rent control is characterized
by nominal rent freezes. This mechanism, which was typically used in post-war periods because of the scarcity of affordable housing, evolved in different ways according to the country. While in some cases the restrictions were removed once the housing shortage was solved, in other cases they remained in place for longer. • Second generation rent control: After the removal of barriers to rent updates, a second type of control emerged when automatic rent increases took place according to a particular price index, used mainly from the 1970s onwards as an inflation control mechanism. • Third generation rent control: Once the previous controls are no longer frequent, they have mostly been substituted by rent regulation within the tenure, providing security for current tenants but not protecting new ones; this has been called third generation rent control corresponding to rent control within a tenancy and not between tenancies (defined by consecutive different contracts between the same parties). Most EU countries have progressed to the use of a system that allows annual rent indexation in a simple way (i.e., CPI, housing costs) and a simultaneous process of decontrolling rents in new contracts. Provisions for the renewal contracts (limitations on landlord negotiation power) have shown a tendency to free negotiation of the duration and contract termination rules as there is widespread agreement among policy-makers that rents should reflect market conditions. Over the last decades, using different argumentations but with similar objectives, many European countries and the USA have implemented policies aimed at stimulating access to owner-occupation for low-income households. Although certain authors (Green and White, 1997; Rossi and Weber, 1996) have stated that home ownership creates positive externalities, other research has highlighted the negative aspects which involve an almost non-existent rental sector such as imposing limits on household mobility or contributing to the creation of a captive demand (i.e., young people) given the lack of rental housing (DiPasquale and Glaeser, 1999).
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 141
Despite the good intentions of the objectives underlying under rent control, there is no general conclusion concerning the effects of rent regulations on the possible benefits experienced by households. Rent regulations in general appear to achieve little in terms of income redistribution and do so at the cost of considerable efficiency losses (Turner and Malpezzi, 2003). The Spanish case has to be considered an example of the use of a varied range of regulations (and deregulations) without any real impact on the rental sector. As we shall conclude later, other measures rather than strict or lax legislation might offer better results in terms of supporting a larger rental sector.
7.4 The Spanish case The housing system in Spain is significantly different from that of other European countries since the welfare state and, in particular, housing policies have followed a significantly different path and have created a different framework for public policies. Political and economic differences also mean that Spain has specific housing features: four decades of dictatorship followed by a relatively sudden membership of the European Union in 1986 created both political and economic struggles and the need to overcome the tensions emerging in Spanish society. Housing affordability has been, and remains, one of the main problems for public authorities. In this sense, priorities have changed since the beginning of the 1990s: From a general desire to facilitate housing access for a huge majority of the population to more segmented and specifically oriented help for certain collectives (Leal Maldonado, 2000; Rodríguez-López, 2006). Until very recently, Spain has enjoyed favorable economic conditions – a low unemployment rate, high rates of GDP growth – partially supported by an extraordinarily dynamic construction sector: property developers and investors generated great expectations based on the rise in house prices, one of the most substantial rises in Europe. This process also reinforced the upward trend of second home construction, especially in some coastal areas. Simultaneously, low interest rates until 2006 and considerable financial facilities have encouraged demand in the housing market, which has also been boosted by the emancipation of youngsters from the baby-boom generation and the massive influx of immigrants. According to the literature focused on Southern European countries and, in particular, Spain (Allen, 2006; Allen et al., 2004; Leal Maldonado, 2005; Pareja-Eastaway and San Martín, 1999, 2002; Sánchez-Martínez,
142 Housing Market Challenges in Europe and the United States
2002), there are some particularities in the Spanish housing market that need to be highlighted: • The tenure system is completely unbalanced in favor of homeown-
ership and social housing represents quite a scarce percentage of the housing stock. • There are a significant number of vacant properties and second homes, especially on the Mediterranean coast. • Family and relatives play an important role in facilitating housing access for other members of the household, reinforcing the assumption that the relationship between housing and household structure is different in southern European countries in comparison to the rest of Europe.3 • Spain displays a more developed housing provision system than other southern European countries, with a strong real estate sector within a complex urban planning system and important links to financial markets. This general framework is relevant in order to define the scenario where regulations in the rental market take place. As mentioned previously, regulation and liberalization of the rental sector are context-related, as they should complement other related housing policies thus contributing to a more satisfactory policy outcome. Although Spain does possess a quite dynamic real estate sector, particularly until 2007, it is simultaneously a country with one of the highest housing deficits for certain social groups such as young people, low-income families or immigrants (Leal Maldonado, 2005). Therefore, public policies possess a strategic capacity to provide housing for certain demand segments. Over the last few decades, a constant housing policy target has been the stimulation of homeownership. Low interest rates and the liberalization of Spanish financial markets have contributed to this objective since the beginning of the 1990s. The scarcity of rental dwellings along with the lack of public social housing has worsened the situation of certain social groups in Spain. Moreover, difficulties in terms of access to homeownership are currently increasing: according to the estimates of the Spanish Central Bank, the typical household effort made to access an average dwelling is around 46.8 per cent of gross salary4 and this figure is even higher in urban agglomerations such as Barcelona or Madrid. In addition to this situation, and given the considerable increase in house prices until 2007 and the increase in unemployment, as a direct consequence of the current economic crisis house buyers have recently encountered
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 143
enormous difficulties in meeting their mortgage payments. Indeed, the historical public encouragement of the owner-occupied sector together with the favorable economic and financial conditions of the past few years have led to high levels of debt in Spanish households. This might create additional tensions in the market as Spanish households are currently encountering considerable difficulties in dealing with mortgage repayments: mortgage arrears went from 0.03 per cent in 2006 to 2.4 per cent in the fourth term of 2008. Nowadays, the key issue at stake in Spanish housing policies is the need to identify mechanisms to stimulate the rental housing supply, both from vacant dwellings or from the housing stock bought as an investment. This is not new. For many years, there has been an insistent claim to enlarge the rented sector from many Spanish agents (including civil society and varied range of authorities). However, a certain gap between the political will of stimulating the rented sector and the measures finally adopted is detected. As García Montalvo (2007) stresses: ‘For many years, specific measures to promote the rental sector as an alternative to homeownership have not accompanied public intentions to favor this sector (. . .) only 10 per cent of the housing stock is under this form of tenure. It reveals the clear malfunction of the Spanish real estate and its anomalous behavior, hindered by a complex framework of measures supporting homeownership which conform to the so-called “ownership culture”’ (p. 150). The following sections offer an in-depth analysis of the elements that need to be taken into account in order to give proper policy alternatives to improve this state of affairs. Firstly, we will give a broad overview of the close relationship between tenure and housing policies in Spain. Secondly, evolution of rental regulations and their impact, if any, on the housing market will be examined. Finally, given the urgent need to boost the Spanish rental sector; current policy measures will be explained. 7.4.1 Tenure and housing policies Although there is currently a tenure imbalance in Spain, with a predominance of the owner-occupied sector, this has not always been the case (Pareja-Eastaway and San Martín, 2002). The decreasing tendency of the rental sector started in 1950, as a consequence not only of successive public stimuli of owner-occupation but also of rent regulations aimed at freezing and encouraging long-term tenancies (see Table 7.2). An additional characteristic of the housing market in Spain is the high percentage of housing that has been completely paid for – more than 50 per cent. Consequently, around 3.2 million households had mortgage
144 Housing Market Challenges in Europe and the United States Table 7.2
Tenure in Spain
in % Social (public) rent Private rent Rent Owner-occupation Others Total
1950
1960
1970
1981
1991
2001
3 51 54 46 0 100
2 43 45 51 4 100
2 30 32 64 4 100
2 21 23 73 4 100
1 15 16 78 6 100
1 10 11 82 6 100
Source: Population and Housing Census, 2001.
Number of contracts
Monthly rent/sq.metre
Number 30 000
EUR 16 14 12 10 8 6 4 2 0
25 000 20 000 15 000 10 000 5 000
Figure 7.2
06 20
04 20
02
00
20
98
20
19
96 19
94 19
92 19
90
88
19
86
19
19
19
84
0
Contracts and rents in Barcelona, 1984–2007
Source: Secretaria d’Habitatge, Generalitat de Catalunya.
payments pending in 2001 (22.8 per cent). This fact has consequences that go beyond the housing market itself: Given the enormous increase in housing prices since the beginning of the 2000s in Spain, Spanish homeowners have enjoyed an important ‘wealth effect’. In a few years they have noticed a considerable increase in the value of their property which has a positive effect on their security – although this is fictitious, as they still need to consume housing services – although it may have contributed to the huge increase in household expenditure and debt over the last few years. The Spanish rental sector has not been unaffected by the real estate boom and not only house prices but also rents have increased in recent years. Taking Barcelona as a case in point (Figure 7.2), it is noticeable that
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 145
there has been an enormous increase in monthly rent per square meter since 1998, the start of the real estate boom. It is also evident that the new legal framework established since 1994 may have stimulated new contracts after a few years, parallel to the increase of real rents due to the enormous increase in house prices. The reduced dimension of the rental market in Spain is in part due to the relative scarcity of other formulae used in other European countries such as cooperative housing or public rental housing. Evidently, the role of housing policy has been influenced by this situation. In broad terms, policy intervention in housing markets in Spain follows a different path to other European countries. The restructuring process that took place after the Civil War was notably slow given the economic restrictions that affected Spain at that time. The relatively late industrialization process and the migration of population from the countryside to the urban conurbations of large cities such as Barcelona or Madrid took place mainly during the 1950s and 1960s, when the public shortage of housing to cater for migrants was most evident. Public authorities reacted by developing housing programs that encompassed different degrees of public involvement. Under the socalled ‘Vivienda de Protección Oficial’ or Regulated Housing (RH), policies stimulated owner-occupation rather than the rental sector, theoretically targeting low and medium-income households although this objective was not totally accomplished:5 developers received grants and many types of subsidies in order to build dwellings for low and medium-income households. Nevertheless, mainly high-income households also bought them. Furthermore, fiscal policies in Spain have also favored house buying, and the purchase of first and even second homes has received different forms of tax relief. There are at least two main focal points in Spanish housing policies; on the one hand, the provision of financial support, basically aimed at the demand side of the market and, on the other, the reactivation of the building sector as a means of expanding the whole economy. Depending on the objective to be achieved, the government has used different instruments ranging from fiscal assistance to direct aid. Before the 1980s, public expenditure on housing was mainly oriented towards ‘bricks and mortar’, mainly providing subsidies to the supply side of the market because of its strong participation in the whole economy and its capacity to generate new employment. From the 1980s onwards, i.e. during the period known as ‘the transition’, there was a shift from supply side policies to more demand-oriented ones. Since then, housing policies have reinforced ‘social’ questions
146 Housing Market Challenges in Europe and the United States
i.e. they prefer to grant subsidies to the demand side by subsidizing interest rates and accompanying these measures with construction sector-oriented support, which focuses particularly on developers. Therefore, the development of public (or private) rental housing as an alternative to providing social housing was scarce since public initiatives did not really support other forms of tenure except owner-occupation. Furthermore, the process of privatization of the relatively small narrow public rental sector started somewhat earlier than in any other European country given the favorable conditions offered to tenants, conditions which mean that nowadays this housing stock barely represents 1 per cent of the total housing stock. Spain led the way in Europe in raising the possibility of selling out public (social) housing stock and focusing on the production of owner-occupied social housing (Leal Maldonado, 2005) (see Figure 7.3). According to the Spanish Ministry of Promotion, in 2003 more than 86 per cent of landlords were private parties, scarcely 6.7 per cent were private companies and 7.2 per cent corresponded to public administrations or publicly-owned companies. This is evidently one of the main differences in comparison with other European countries such as The Netherlands or Germany, where private or publicly-owned companies are in the majority. This fact is clearly of most importance when implementing policies aimed at stimulating the supply side of the sector. The relevance of the private rental sector as an alternative to so-called ‘social housing’ in Europe is even larger if we take into account the fact Unregulated dwellings
Regulated dwellings
700 000 600 000 500 000 400 000 300 000 200 000 100 000
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
0
Figure 7.3
Regulated and unregulated housing, 1962–2007
Source: Spanish Ministry of Public Works (1960–1995), Spanish Ministry of Housing (1995–2007).
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 147 Percentage over finished regulated housing 70 60
61.0
50
52.1
40
42.9
39.4
40.0
2005
2006
30
36.5
20 10 0 2002
2003
2004
2007
Figure 7.4 ‘Disqualified’ regulated housing Source: Spanish Ministry of Housing and authors’ calculations.
that, until very recently, dwellings which qualified during their construction as ‘Regulated Housing’ could later be ‘deregulated’ and from then on participate as ‘deregulated housing’ in the market. This represented a considerable percentage (see Figure 7.4) of the total amount of completed public housing. According to available data, the total amount of ‘deregulated housing’ represented in 2000 and 2007 around 61 and 36.5 per cent of the total amount of completed Regulated Housing respectively. The downward trend noted may be related to the tougher conditions encountered in ‘deregulating’ Regulated Housing in the new Housing Programs introduced since 2005. A last element to be taken into consideration is the enormous fiscal support received by housing nowadays in Spain; the fiscal advantages involved in house buying are among the most generous in Europe and OECD countries to favor homeownership rather than renting. Indeed, the current tenure unbalance has been strongly determined by a complex framework of tax relief, which achieved its maximum level with the Law 48/1985 in which even the acquisition of second homes received fiscal benefits.6 The following section revises the main aspects of rent regulation in Spain, paying special attention to contract length and possible rent updates. 7.4.2 Historical revision of rent control and regulations in Spain Despite its integration in the general trend of housing policies and programs in Spain, the rental sector shows some particular measures and
148 Housing Market Challenges in Europe and the United States
regulations. Three periods may be distinguished according to the different direct regulations applied to the rental sector (Pareja-Eastaway and San Martín, 2002): 1. From 1946 until 1985: This period strictly corresponds to the ‘first generation’ contracts. It started in 1946, with the Urban Rent Act, a clear example of a populist measure that favored tenants by means of rent freezes and automatic extension of contracts. This law was reformulated later in 1964 (Urban Rent Act 4104/1964), making revision of contracts possible7 by maintaining the conditions of compulsory renewal. The updating of contracts was freely defined in the rental contract. All expired contracts were compulsorily renewed although there were some exceptions, mainly related to the possibility of the property owner’s relatives needing the dwelling. Compulsory renewals implied the absence of a rent revision clause. Rents were freely established and agreed by the landlord and the tenant when setting the rent contract, except in the case of rental Regulated Housing where policy measures legally fixed rents. Two direct consequences of the rigidity of this legislation were that tenants received few incentives to move out of their flats and owners, who were responsible for maintenance and repairs, were not encouraged to improve and modernize their properties. Consequently, although these measures were aimed at supporting tenants, in practice they were the basic cause of the continuous decrease in the rental sector, which had dropped from 54 per cent in 1950 to 23 per cent at the beginning of the 1980s. In conclusion, it is evident that rent control was undoubtedly a key factor in contributing to the downward trend of rental housing in Spain. 2. From 1985 until 1994: Rent controls lasted longer in Spain than in other European countries, as they were not abolished until 1985, which was a turning point in rental legislation in Spain. The Boyer Decree8 represented a milestone in rental regulation in Spain. Absolute liberalization in terms of defining the duration of the contract and the updating of rent prevailed in new contracts. The objective of this measure was to increase the amount of dwellings made available in the rental sector and thus the decree introduced the ‘removal of compulsory renewal which will stimulate the construction of residential dwellings and commercial offices aimed at the rental sector’ (Royal Decree 2/1985, p. 13.176 ). The absence of clear rules made for an unstable reference framework for new contracts and made the participation of the rental sector in
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 149
the market even less significant. At that time several types of contracts co-existed in Spain depending on when the rental agreement was signed, thus contributing to an unclear scenario for new agreements: contracts with ‘old rents’ and an unlimited duration along with recently established with a very short duration. Undoubtedly, a reform of the 1985 Act was needed from the very beginning of its application. One of the clearest consequences of the liberalization of the legal framework for the rental sector was the increase in new rents due to the instability created by short-term conditions (Faus, 2000). 3. From 1994 onwards: The last Urban Rent Act passed in Spain dates from 1994.9 The main objective of this law was to stimulate the rental sector through the creation of a more stable framework of relations between landlords and tenants, preserving freedom in the setting of contract duration and rents but with a minimum of five years required by law or if the contractual period is shorter, with a compulsory renewal clause. Rents are freely settled by landlords and tenants but must be updated in line with the CPI (Consumer Price Index). This type of regulation corresponds to ‘third generation’ contracts according to Arnott (1995), i.e., rent regulation of rents during the period of tenancy but not once the contract has been renewed. Any possible expectations regarding a possible turning point in the downward trend of rental units are not confirmed by the available data. Spain displays a wide range of contracts depending on the period the contract was signed and the legal framework in force at that time. The rental stock can therefore be classified depending on the existing regulation when the contract was signed. Table 7.3 shows the cost of the rental stock depending on the year the contract was signed. Table 7.3 shows the upward trend of rental costs after the change in legislation in 1985. The number of dwellings with rents lower than a100 decreased significantly after 1985. Therefore, almost half of the contracts signed from 2000 onwards have a rent fee of between a300 and 500. The negative effect of rent regulation on the rental sector is common in the majority of European countries, as previously mentioned in this chapter. Spain shows a remarkable difference given the lack of (social) alternatives in the housing market, particularly regarding the absence of a strong public rental sector. In Europe, the private rental sector represents an additional alternative (not the only solution) for problems of economic insolvency (Rodríguez Alonso, 2005).
150 Housing Market Challenges in Europe and the United States Table 7.3
Cost of rented dwellings according to the year of signature
%
Total
Less than a100
a101–300
a301–500
More than a500
Till 1984 1985–1994 1995–2000 After 2000 Weighted average
100 100 100 100 100
45.7 25.4 8.0 2.4 15.8
38.8 48.8 42.2 32.6 38.5
12.7 20.0 41.1 50.9 36.5
2.9 5.8 8.8 14.2 9.2
Source: Encuesta de Demanda de Alquiler, Ministerio de Fomento (2003).
7.4.3 Addressing a more balanced tenure system in Spain Several arguments can be made in favor of promoting a larger rental sector in Spain. Firstly, the need for a larger rental sector is linked to redistributive objectives, given the need to provide rental dwellings of a social nature. In those cases where the tenure system is completely unbalanced in favor of homeownership, it is inevitable to associate rental housing with social housing or housing aimed at low-income families or vulnerable social groups (Pareja-Eastaway and San Martín, 2002). In Spain, as mentioned previously, the lack of affordable alternatives to owner-occupation determines both the housing capacity and guarantees associated with the rest of the housing stock aimed at the more precarious demand segments. Secondly, a larger rental sector means more rationality in household choices as it offers a larger and more varied range of options. This applies especially to young people and newly-formed households. Moreover, a proper dimension of the rental sector guarantees the possibility of choice for geographically mobile people. This fact might even create positive externalities in the labor market as it facilitates labor mobility (Lundborg and Skedinger, 1999). Finally, a sufficient quantity of rental dwellings contribute to the stabilization of real estate cycles, acting as a shock absorber of sudden fluctuations in housing prices and contributing to a more efficient assignment of resources. This issue plays a particularly important role in a country such as Spain given the recent real estate boom and subsequent crisis. Nowadays there is a significant gap between supply and demand in rental dwellings. There is a considerable unsatisfied demand among certain social groups, especially young people, immigrants and low-income households who encounter many difficulties in finding affordable housing given the scarcity of Regulated Housing (RH) and rental housing.
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 151
With respect to this problem, there is a substantial debate about the need to enlarge the social housing sector in Spain, both rental and owneroccupied (Pareja-Eastaway and San Martín, 2002; Pareja-Eastaway, 2007). Nevertheless, one of the main challenges still facing housing policy is the need to devote enough resources to encourage the production of public and rental dwellings. New Regulated Housing implies a considerable problem, which lasts for decades: this problem is related to the relative ease with which subsidized housing is adapted into the nonregulated sector since a considerable quantity of this housing becomes privatized,10 even before the legal established period. This is undoubtedly profitable for owners of Regulated Housing, especially in boom periods, as the difference between the market price and price of acquisition of Regulated Housing becomes enormous.11 Furthermore, although considerations about the lack of efficiency of regressive fiscal measures such as tax relief on housing acquisition would recommend its abolition (Sánchez-Martínez, 2002); this would involve a high political cost due to its popularity among citizens. Therefore, governments are prevented from taking decisive action against this issue and even from implementing side measures aimed at fiscal benefits in the rental sector. In Spain nowadays, several policy measures aimed at stimulating the rental sector do not involve any changes in legislation (rent controls or length of contracts). Most of them take into account the following elements: • The public sector adopts the role of intermediary between landlords
and tenants of rental dwellings, promoting the creation of information networks available to both sides of the market (i.e., Xarxa d’Intermediació per al lloguer social – Intermediary Network for Social Rent – in Catalonia or the Bizigune Program in the Basque Country). • During the length of the contract, insurance with public guarantees will cover any risks including the possible non-payment of rent. The creation of the Public Company in the Rental Sector – Sociedad Pública de Alquiler – in 2005 has not only boosted networking and information in the rental sector but also improved eviction mechanisms through the creation of a new device, the so-called ‘Secure Rent’12 along with subsidizing the use of legal mechanisms to ensure eviction in case of non-payment. • Over the last few years, the Ministry of Housing has also encouraged demand through the establishment of a ‘voucher’13 aimed at young people (under 35 years of age) who are about to become independent
152 Housing Market Challenges in Europe and the United States
•
•
• •
•
from the parental home and direct subsidies to low-income families.14 However, the final output of these recent measures is still unclear given the negative effects that they might create in the market considering the scarcity of rental dwellings. The subsidizing of demand might finally be passed on or capitalized on the supply side by means of higher rents. Owners of rental dwellings receive incentives to upgrade their dwelling through grants and subsidies. An additional measure adopted at the beginning of the 1990s was the inclusion of fiscal deductions in income tax for rental dwellings. However, this measure disappeared after eight years15 and the only form of fiscal deduction that remained was for the acquisition of owner-occupied housing. As stated previously, Spain has a well-established policy framework of fiscal incentives for house buying. As García Montalvo (2003) points out, this strategy displays certain inconsistencies and contradictions in its objective of stimulating the rental sector.16 The development of a social housing sector (public rent) by means of assigning below-market rents to those who can prove they belong to certain vulnerable social groups. Taxing vacant dwellings becomes possible in order to stimulate their participation in the market. As private landlords represent more than 80 per cent of supply, other measures aimed at developing a professional sector, able to compete with the homeownership sector and other alternatives for savings, in terms of liquidity and returns, are under revision.17 Housing Plans have encouraged the rental sector since 1992 but as a side measure, i.e., the main subsidized tenure form by far was homeownership. However, the last Plan 2009–2012 promotes the rental sector with an option to buy after some years of tenancy.
The revitalization of the rental sector is urgently needed owing to the unsolved problem of housing access for certain social groups: measures aimed at enlarging the sector beyond mere short-run measures, involve, among others, the creation of a stable legal framework in order to facilitate its development.
7.5 Conclusions Many European countries have shown a growing unbalance towards owner-occupation since the mid-1990s. There are multiple reasons that have contributed to this situation and they are linked to macroeconomic
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 153
and financial variables such as the booming scenario of the real estate sector and the downward trend of interest rates. Nevertheless, housing policies and rent regulations have not been neutral factors in this market output. Despite the context-related definition of social housing, there is growing concern related to the unsolved housing problems of certain demand segments, especially young people, immigrants and low-income households. The current debate on housing policies focuses on the role played by the rental sector as a means of providing affordable and adequate housing for these collectives. There is an apparent consensus regarding the need to preserve and encourage the rental sector as it fulfils certain roles that cannot be achieved by other means. Therefore, the role played by rent regulations and other side-policy instruments has the reference point and the key issue in many policy discussions to identify the proper mechanisms to enlarge the rental sector, or at least stop its downward trend. According to many empirical and comparative studies, the traditional arguments favoring the regulation of the housing market and, in particular, the rental sector, given the lack of competitive conditions are counteracted by the lack of efficiency in terms of output of certain control mechanisms and the negative side-effects they may create. Supply-side measures (i.e., subsidies to developers, fiscal incentives to owners, taxes on vacant dwellings), demand-side measures (i.e., subsidies or vouchers) and market regulations (i.e. rent control) are being revisited by many European countries to stimulate rental tenancy. Historically, Spain has made several efforts to provide affordable/Regulated Housing for certain demand segments, most facilitating ownership access. As we have already pointed out, there is no scarcity of housing in Spain (indeed, there is just the opposite), particularly after the property boom that has taken place since the 1990s. However, according to the OECD (2005), Spanish housing policies have been unable to provide affordable accommodation for low-income households given the mismatch between supply characteristics and demand needs. Homeownership has been additionally favored to the detriment of the rental sector by the housing price inflation that has occurred since the end of the 1990s. A certain contradiction can be detected in the spirit of Spanish housing policies given the opposite targets proposed by Housing Plans and rental regulations. The downward trend of rental units in the market, partly supported by housing policies, has been the main reason to change the legal framework of the sector. Different legislations, ranging from strict
154 Housing Market Challenges in Europe and the United States
regulation to complete liberalization and back to a less flexible scheme, have been unable to reactivate a historical reluctance to rent. Both tenants and landlords have been negatively affected by the way the rental sector has been regulated and treated fiscally. Regulation and liberalization have both failed to provide a substantial increase in the rental sector, counteracting the trend established since the 1950s. The so-called ‘ownership culture’ usually associated with countries like Spain where homeownership remains the dominant tenancy should not be confused with the limits an almost negligible rental sector imposes on the population. Although reasons concerning better quality and location of owner-occupied dwellings may be acceptable in some cases, not all households are in a position to buy a house. Access to a decent home does not necessarily mean access to homeownership. In this chapter, we have mentioned several arguments (i.e., emancipation, mobility and excessive mortgage burdens) which deserve the attention of public authorities in order to guarantee a significant rental sector both in quality and in quantity. Nowadays, positive discrimination towards this sector is completely justified in order to avoid the lack of confidence among tenants and landlords which was created in the past.
Notes 1. Associate Professor, Economics Department, University of Barcelona, Spain. 2. Associate Professor, Applied Economics Department, University of Granada, Spain. 3. This characteristic is considered in Esping-Andersen (1990) as a ‘Traditionalist/ traditionalist-familiar Welfare State’ label assigned to Southern Europe. 4. Although official recommendations suggest 30 per cent as a maximum. 5. With public subsidies, both standard quality and luxury dwellings were built. 6. This measure lasted until 1991. 7. Two possibilities to modify rents: by changing the level (stabilization clause) or by means of periodical updating. 8. This is the name received by the Royal Decree 2/1985, of April 30th on Measures of Economic Policy as it was designed by Miguel Boyer, who was then the Spanish Minister of Economy and Public Finance. 9. Law 29/1994. 10. Legally, the beneficiaries of RH can sell their houses at a market price 15 years after they bought them. Some Autonomous Communities in Spain have tightened up the conditions and have extended the legal time period for selling to 30–50 years and have even totally forbidden their sale as has occurred in the Basque Country (Burón Cuadrado, 2008). 11. An additional problem is the fraud detected in selling prices (higher than the legal threshold) or even the transformation of RH into second homes. 12. This figure mainly guarantees payment of rent and dwelling conditions to the owner.
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 155 The amount stands at a210 per month during a maximum of four years. The subsidy may represent up to 40 per cent of the rent. Law 40/1998. Some Autonomous Communities have recently included this measure in their own legislation, comprising deductions of between 5 and 20 per cent in the rent paid by young people (between 32 and 35 years of age) with a certain maximum to be deducted. 17. In fact, Real Estate Investment Funds and Societies were regulated under Law 19/1992 in 1992. The aim of this law was to develop a strong professional sector managed by independent societies in the rental sector.
13. 14. 15. 16.
References Allen, J. (2006) Welfare Regimes, Welfare Systems and Housing in Southern Europe, European Journal of Housing Policy, 6 (3), 251–77. Allen, A., J. Barlow, J. Leal, T. Maloutas and L. Padovani (2004) Housing & Welfare in Southern Europe, Oxford: Blackwell Publishing. Arnott, R. (1995) Time for Revisionism on Rent Control? Journal of Economic Perspectives, 9, 99–120. Arnott, R. (2003) Tenancy Rent Control, Swedish Economic Policy Review, Vol. 10, pp. 89–121. Balchin, P. (1996) Housing Policy in Europe, London: Routledge. Boelhouwer, P. and H. van der Heijden (1992) Housing Systems in Europe: Part I: A Comparative Study of Housing Policy, Housing and Urban Policy Studies, No. l, Delft: Delft University Press. Burón Cuadrado, J. (2008) Una Política de Vivienda Alternativa, Ciudad y Territorio Estudios Territoriales, XL (155), 9–40. Clapham, D., J. Hegedüs, K. Kintrea and I. Tosics (eds) (1996) Housing Privatization in Eastern Europe, London: Greenwood Press. DiPasquale, D. and E.L. Glaeser (1999) Incentives and Social Capital: Are Homeowners Better Citizens? Journal of Urban Economics 45 (2), 354–84. European Central Bank (2003) Structural Factors in EU Housing Markets, Frankfurt: ECB. Ellingsen, T. and P. Englund (2003) Rent Regulation: An Introduction, Swedish Economic Policy Review, 19, 3–9. Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism, Cambridge: Policy Press. Faus, J. (2000) Libro Blanco del Sector Inmobiliario, IESE and Fundación de Estudios Inmobiliarios. García Montalvo, J. (2003) La vivienda en España: desgravaciones, burbujas y otras historias, Perspectivas del Sistema Financiero, No. 78, FUNCAS. García Montalvo, J. (2007) Algunas consideraciones sobre el problema de la vivienda en España, Papeles de Economía Española, No. 113, FUNCAS. Green R. and M. White (1997) Measuring the Benefits of Homeownership: Effects on Children, Journal Urban Economics, 41, 441–61. Glaeser E.L. (2003) Does Rent Control Reduce Segregation? Swedish Economic Policy Review, 10, 179–202.
156 Housing Market Challenges in Europe and the United States Haffner, M., M. Elsinga and J. Hoekstra (2008) Rent Regulation: The Balance between Private Landlords and Tenants in Six European Countries, European Journal of Housing Policy, 8 (2), 217–33. Harloe, M. (1985) Private Rental Housing in the United States and Europe, London and Sydney: Croom Helm. Kemeny, J. (1995) From Public Housing to the Social Market: Rental Policy Strategies in Comparative Perspective, London: Routledge. 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 (5), 59–75. Kleinman, M. (1996) Housing, Welfare and the State in Europe: A Comparative Analysis of Britain, France and Germany, Cheltenham: Edward Eldgar Publishing Ltd. Leal Maldonado, J. (2000) Comportamientos esidenciales y necesidades de vivienda, in Taltavull (Coord.), Vivienda y Familia, Madrid:Fundación Argentaria. Leal Maldonado, J. (2005) Desigualdad residencial y sistema de bienestar en España in: Ruiz-Huerta, J. (ed.): Políticas públicas y distribución de la renta, Madrid: Fundación BBVA. Lind, H. (2001) Rent Regulation: A Conceptual and Comparative Analysis, European Journal of Housing Policy, 1, 41–57. Lind, H. (2003) Rent Regulation and New Construction: With a Focus on Sweden 1995–2001, Swedish Economic Policy Review, 10, 135–67. Lundborg, P. and P. Skedinger (1999) Transaction Taxes in a Search Model of the Housing Market, Journal of Urban Economics, 45 (2), 385–99. Maclennan, D. and A. More (1997) The Future of Social Housing: Key Economic Questions, Housing Studies, 12 (4), 531–47. Mandic, S. (2000) Trends in Central East European Rental Sectors, Journal of Housing and the Built Environment, 15 (3), 217–31. McCrone, G. and M. Stephens (1995) Housing Policy in Britain and Europe, London: UCL Press. Malpass, P. (2008) Housing and the New Welfare State: Wobbly Pillar or Cornerstone? Housing Studies, 23 (1), 1–19. Navarro, V. (2004) El Estado de Bienestar en España, Barcelona: Tecnos. OECD (2005) Economic Survey of Spain 2005: Stabilising the Housing Market, OECD Publications, available at http://www.oecd.org/dataoecd/53/3/34586052.pdf Oxley, M. and J. Smith (1996) Housing Policy and Rental Housing in Europe, London: Taylor and Francis, Inc. Pareja-Eastaway, M. and I. San Martín (1999) General Trends in Financing Social Housing in Spain, Urban Studies, 36 (4), 699–714. Pareja-Eastaway, M. and I. San Martín (2002) The Tenure Imbalance in Spain: the Need for Social Housing Policy, Urban Studies, 39 (2), 283–95. Pareja-Eastaway, M. (2007) Residential Opportunities and Emancipation Strategies in an Owner-occupied Dominated Market, ACE: Arquitectura, Ciudad y Entorno, 2007, Year II, No. 5, available at http://hdl.handle.net/2099/3700 Priemus, H. and F. Dieleman (2002) Social Housing Policy in the European Union: Past, Present and Perspectives, Urban Studies, 39 (2), 191–200. Priemus H. and D. Maclennan (1998) The Different Faces of Private Rental Housing Journal of Housing and the Built Environment, 13 (3), 197–204.
Montserrat Pareja-Eastaway, Teresa Sánchez-Martínez 157 Priemus, H. and S. Mandic (2000) Rental Housing in Central and Eastern Europe as No Man’s Land, Journal of Housing and the Built Environment, 15 (3), 205–15(11). Rodríguez Alonso, R. (2005) Infrautilización del parque de viviendas en España: aparición de viviendas vacías y secundarias, Boletín CF+S 29/30, Notas para entender el mercado inmobiliario, available online at: http://habitat.aq.upm.es/ boletin/n29/lista.html Rodriguez-López, J. (2006) Los booms inmobiliarios en España. Un análisis de tres periodos, Papeles de Economía Española, 109, 76–90. Rossi, P. and E. Weber (1996) The Social Benefits of Homeownership: Empirical Evidence from National Surveys, Housing Policy Debate, 7, 1–35. Sánchez-Martínez, M.T. (2002) La política de vivienda en España, Análisis de sus efectos redistributivos, Granada: Ediciones de la Universidad de Granada. Scalon, K. and C. Whitehead (2004) International Trends in Housing Tenure and Mortgage Finance, The Council of Mortgage Lenders, available at: www.cml.org.uk/cml/filegrab/ Trilla, C. (2001) La política de vivienda en una perspectiva europea comparada, Colección Estudios Sociales, Barcelona: Fundación La Caixa. Turner B. and S. Malpezzi (2003) A Review of Empirical Evidence on the Costs and Benefits of Rent Control, Swedish Economic Policy Review, 10, 11–56. Van der Heijden, H. (2002) Social Rental Housing in Western Europe: Developments and Expectations, Urban Studies, 39 (2), 327–40. Van der Heijden, H. and P. Boelhouwer (1996) The Private Rental Sector in Western Europe, Housing Studies, 11 (1), 13–33. Whitehead, C. (2003) The Economics of Social Housing, in: O’Sullivan, T. and K. Gibb (ed.) Housing Economics and Public Policy, Oxford: Blackwell Science Ltd. Whitehead, C. and K. Scanlon (2007) Social Housing in Europe,London School of Economics and Political Science.
8 Fiscal Aspects of Housing in Europe Guido Wolswijk1
8.1 Introduction A large degree of volatility in housing markets and in mortgage markets may have disruptive effects on the economy. Large negative shocks on these markets may, for instance, adversely affect the economy and private consumption in particular via the wealth effect and via less mortgage equity withdrawal, but negative effects on housing investment may also arise.2 Disruptions in the housing sector can be quite severe as a result of the lagged responses in the supply of houses and because of households’ adaptive formation of expectations, extrapolating recent house price increases (and related capital gains) to the future. Large adverse shocks can also impair financial stability.3 Governments have various instruments at their disposal to influence housing markets including affecting demand via changes in tax rates of, for instance, the personal income tax, or more specific measures oriented towards the housing sector. Housing markets in the EU are largely national in nature, with a rather limited volume of cross-border activities: specific national instruments are therefore broadly as effective as before EMU.4 Nevertheless, in the event of housing market crises, it is often monetary policy or fiscal demand management that is called upon to boost domestic demand, with little attention being paid to specific fiscal instruments. This contribution considers fiscal policy instruments for influencing housing market developments in the euro area countries. This is in line with the call to increase attention to the role of micro policy measures to limit asset price fluctuations, as expressed by the G10 Contact Group on Asset Prices (2002). The focus in this chapter on the 15 euro area countries is motivated by the lack of national monetary policy options 158
Guido Wolswijk 159
for them, which potentially increases the value-added of national fiscal instruments in affecting housing market developments. Alternative policy options, such as monetary policy and prudential supervisory policy, fall outside the scope of this chapter. We focus specifically on the owner-occupied part of the housing market.5 Section 8.2 provides a description of main fiscal instruments available for national policymakers to affect owner-occupied housing market developments, and how they are currently used in euro area countries. Section 8.3 discusses how some specific tax instruments could be used – structurally or on an ad hoc basis – to correct housing and mortgage market developments that are deemed unwarranted. Section 8.4 concludes this chapter.
8.2 Fiscal policy instruments and housing markets There are noticeable differences in housing and mortgage market in the euro area. Mortgage debt levels range from somewhat above 10 per cent of GDP in Italy to more than 100 per cent of GDP in the Netherlands. Similarly, there has been a large degree of variation in price increases of houses in the last few years, ranging from slight nominal decreases to increases well above 15 per cent. Such diversity reflects the fact that housing market determinants are still very national in nature. Mortgage debt and house prices reflect a variety of factors, including social-cultural preferences, population growth, the inflation record, the timing and degree of financial deregulation, national regulations (e.g. regarding consumer protection), and governments’ past and current policies on housing and mortgage-financing. It also needs to be recognized that mortgage debt and house prices influence each other: Higher mortgage levels that reflect mortgage equity withdrawal may cause higher house prices if the additional resources are reflected in higher housing demand (e.g. for second homes), while in turn a high house price may induce higher mortgage demand as buying a new house will require additional financing. Fiscal policies affect housing market developments and thereby also housing and mortgage debt growth in various ways. At a general level, fiscally responsible behavior aimed at keeping public finances sustainable may decrease agents’ uncertainty and increase their willingness to make longer-term commitments such as buying real estate. Also, low government deficits and debt allow for relatively favorable financing conditions on financial markets, thus enabling cheaper financing of long-term mortgage loans.6 More specific fiscal instruments that affect housing markets
160 Housing Market Challenges in Europe and the United States
run from regulations on the use of land to interest rate deductions of mortgage interest payments from the income tax, and from subsidized mortgage loans from state-allied banks to property taxes raised when buying a new house. Governments often aim to promote home-ownership motivated by positive external effects of good housing, for instance improved public health and higher participation in the local community. Thus, Diamond and Lea (1992) report ‘Home-ownership is a reward in itself and may provide an incentive for work, effort and thrift in other activities’. Some empirical studies, focusing on the US, confirm such positive externalities (e.g. Glaeser and Shapiro, 2002). Related is the argument that households underestimate the importance of housing for themselves (merit-good). Finally, income distribution effects play a role, with measures aiming at making own housing also available for the entire population rather than only for the high-income groups which may be less credit-constrained. Besides these social goals, there is also a goal of stabilization of housing and mortgage markets, to which we will return in Section 8.3. Households face multiple choices regarding the housing market: They need to choose between housing consumption and consumption of other goods and services, between buying and renting a house, between buying a new house or an existing one, between living themselves in the house or renting it out, and – for some – between mortgage and equity financing of the house. Government instruments may affect all these choices. Given that most governments aim to favor home-ownership, taxes usually are non-neutral as regards home-ownership. As an own house reflects a financial investment, neutrality of taxes would imply equal treatment compared to other asset classes, such as bonds, implying deductibility of costs made (interest) and taxation of income in kind (imputed rent) and financial gains on selling the house (capital gains tax). At the same time, housing is a durable consumption good; equal fiscal treatment to other durables such as cars implies levying the normal VATrate when buying a new house. Actual tax regimes for owner-occupied housing in the euro area countries can be compared to this benchmark to indicate the degree of fiscal subsidization. Below, we consider in more detail the main tax instruments, in particular a) the taxation of the imputed rental value of the home, b) income tax deductibility of mortgage interest payments, c) capital gains taxes, and d) VAT applying to new houses.7 We do so for a recent year, 2007.8 Table 8.1 summarizes the situation for all euro area countries (and for the US), describing the situation for a ‘typical’ household owning and living in the house. As to the capital gains tax, the table describes the
Guido Wolswijk 161 Table 8.1 2007
Main tax categories affecting housing/mortgages in the euro area,
Austria Belgium Cyprus Finland France Germany Greece Ireland Italy Luxembourg Malta Netherlands Portugal Slovenia Spain PM1: US PM2: neutrality
Tax on imputed rent
Interest deductibility (max rate applying)
Capital gains tax
VAT on new homes as on other durables?
N Y N N N N N N Y Y N Y N N N N Y
Y (50%) Y (50%) N Y (29%) N N Y (40%) Y (42%) Y (45%) Y (38%) N Y (52%) Y (40%) N Y (45%) Y (35%) Y
n n Y n n n N n n n n N n n n n Y
N (exempted) Y Y N (exempted) Y N (exempted) Y N (13.5%) N (10%) Y N Y N (exempted) Y N (7%) Sales tax per state Y
Note: Numbers in brackets in the column on interest deductibility give the top marginal tax rate at which these payments can be deducted. In the column on capital gains taxes, N denotes absence of a capital gains tax, while n refers to a capital gains tax being in place but the gain on selling a house de facto often being exempted. Numbers in brackets in the column on VAT give the rate applying to house construction if deviating from the normal rate. Source: International Bureau of Fiscal Documentation (2007).
tax regime for an owner-occupier selling his/her house after having lived in it for a long period (more than ten years) and using the sale proceeds to buy another house. The VAT column shows whether new housing supply is treated analogous to other durable goods in VAT. Annexes Table A8.1–A8.4 provide more detail per instrument. A) Tax on imputed rent Equal treatment with other investment assets categories calls for imposing a tax on imputed rental income. Deviations from this principle can imply strong incentives to buy a house.9 In a minority of euro area countries (Belgium, Italy, Luxembourg, the Netherlands), owner-occupiers had to report imputed rent to their national tax offices in 2007. In these countries, the tax levy usually is relatively
162 Housing Market Challenges in Europe and the United States
low reflecting low rates and ceilings on the amount to be paid, and the use of out-of-date market values of houses as valuation base.10 The small number of countries implementing this kind of tax may reflect high enforcement costs and low revenue (Freeman et al., 1996). A typical problem in the implementation of this tax is that the income in kind derived from housing is not easily measurable, and therefore tax authorities need to resort to rough measures. Basing it on the market value effectively turns it into a wealth tax (Bourassa and Grigsby, 2000). Indeed, while taxation on imputed rent is not very common in euro area countries, all countries but Ireland impose a property tax, and some also a wealth tax.
B) Interest relief In a neutral fiscal system of housing, interest deductibility is justified as a means to deduct costs made from (taxed) earnings, following the principle that if benefits are taxed, costs made to earn the benefits can be deducted. Interest deductibility not only increases the affordability of principal dwellings, and thus the amount of mortgage debt taken up, but it also affects the type of mortgage loans taken up. In the Netherlands, for instance, which has the most unrestricted system of interest deductibility, taking up a mortgage loan coupled with a capital insurance is fiscally favored. Similarly, interest-only loans were introduced to take full advantage of the nearly unlimited deductibility of mortgage interest payments, now amounting to about half of outstanding mortgage debt. The preferential tax-status of housing is also reflected in the Netherlands being the country in the euro area with the lowest share of homeowners without mortgage (12.5 per cent).11 A large majority of euro area countries allows personal income tax deductibility of mortgage interest payments (and occasionally of reimbursement payments) subject to the house being the primary dwelling of the owner, while there are usually limits to the amount deductible and/or the time allowed for deduction. While in some countries it is eligible for all (e.g. Spain, Italy, Finland), in others the interest deductibility is targeted to specific households (Germany, France).12 The importance of this instrument has decreased somewhat over time as several EU countries have introduced or strengthened limitations on the deductibility of interest payments.13 In addition, the marginal tax rate at which interest payment can be deducted has fallen in many countries as a result of the tax reforms introduced around the turn of the
Guido Wolswijk 163
century. Lower inflation has been another factor contributing to this instrument having lost some of its importance.14 C) Tax on capital gains In a neutral fiscal system, capital gains on houses would be taxed just as the capital gains on other assets like equity or bonds. According to Hendershott and White (2000), the favorable tax treatment of capital gains, combined with the favorable treatment of imputed rents, constitutes the fundamental tax advantage of owner-occupied housing. On average, capital gains on the principal owner-occupied dwelling are exempted from capital gains taxes in the countries of the euro area in particular if the owner has occupied the dwelling for several years before selling it (‘anti-speculation clause’). In only few countries are the capital gains on housing treated more or less identically to other capital gains. Capital gains taxes are normally levied on the nominal increase in the value of a house. D) Value-added tax As housing is also partly consumption, in a neutral fiscal system buying a new house is subject to the same VAT rate as buying another durable good such as a car. In a minority of euro area countries, building a house is subject to the normal (high) VAT rate. However, most countries allow for reduced VAT rates on the supply, construction, renovation and alternation of housing or it even being fully exempted, a possibility included in the VAT Directives. Not taxing new homes or doing so only at a lower rate is motivated by producing incentives to build new houses, to keep the dwelling stock of high quality, and to promote movements along the housing ladder. The general picture emerging from this description and from Table 8.1 is that of a favorable fiscal treatment of housing and mortgage-financing in many countries as can be seen from the deviations from the neutral treatment. The table also shows that the tax treatment of owner-occupied houses in US taxes is not very different from the average of the euro area countries. Outside the euro area, Sweden explicitly aims at a fiscal neutral position as regards the choice of owning a house or renting one. The comparison of the fiscal treatment of housing as described above is not fully complete as it does not take into account some other taxes that are levied when constructing, buying, owning, selling or inheriting a house. Transactions costs, for instance, may play a role in these decisions. High costs of moving to another house or changing (conditions
164 Housing Market Challenges in Europe and the United States
% of total tax revenue 16 14 12 10 8 6 4 2 0 AT
BE
FI
FR
DE
GR
Recurrent house tax Figure 8.1
IR
IT
LX
Other house tax
NL
PT
ES
Total
Housing taxes, 2006
Note: For Belgium and Portugal, the distinction between recurrent and other taxes is not available. Malta, Cyprus and Slovenia are not OECD-members. Source: OECD Revenue Statistics (2007).
of) mortgages may act as a brake on moving/remortgaging while also hindering labor mobility. Comparable recent information on transactions costs, and in particular the tax-part of that, is however not readily available.15 Other taxes that could potentially play an important role in housing decisions are property taxes and wealth taxes, playing a role akin to the taxation of imputed rent. In addition, inheritance and gift taxes may affect the choice whether to rent or to buy, especially in countries where passing on a house from the parents to their children is common (e.g. Cyprus). Furthermore, specific government expenditures on housing (e.g. interest subsidies) also may have significant effects on housing market developments, but the large degree of diversity in this area does not allow for any straightforward comparison across countries. A broad idea of the importance of some of the taxes not dealt with explicitly may be derived from Figure 8.1, showing the share of taxes on immovable property (such as wealth taxation) in total tax receipts. This share differs widely, ranging from a 1/2 per cent in Greece to almost 14 per cent in Ireland. It must be noted though that in some countries certain housing-related receipts fall outside the scope of the OECDdefinition. The overall favorable treatment of housing in fiscal systems is broadly confirmed by some indicators for the overall subsidy for each country. Van den Noord (2003) quantifies governments’ involvement in housing
Guido Wolswijk 165 Table 8.2
Overall housing subsidy measures, 1999
The Netherlands Ireland Spain Finland Austria Italy Portugal Germany France Belgium
Difference preand after-tax mortgage interest rate
Government subsidy (% of mortgage interest payment)
2 0.9 0.9 0.9 0.6 0.5 0.2 0 0 0
34.9 37.3 25 – 7.9 17.5 – – 25.9 37.3
Source: Van den Noord (2003) (2nd column) and Neuteboom (2004) (3rd column).
markets for the euro area countries taking into account information on tax interest deductibility, tax credits and imputed income from housing. He calculated the 1999 tax wedges and the resulting real cost of financing a house (nominal interest rate plus tax wedge minus inflation). His results (see Table 8.2) show that euro area governments on balance subsidize housing, with few exceptions where it is broadly neutral (Belgium, France, Germany), and one (Greece) in which, on balance, a tax is levied on mortgage-financed housing. Neuteboom (2004) estimates for nine European countries a standardized net interest rate, taking into account costs and benefits directly linked to a mortgage such as stamp duties, mortgage interest deductibility and guarantees. Results for the year 1999 show relatively favorable financing conditions in the Netherlands and Belgium, mainly reflecting interest deductibility, and much less so in Austria, where the mortgage-related costs considered actually increase the net interest rate. While broadly confirming the net subsidization of housing markets, the results only refer to the year 1999, and therefore provide only limited guidance to the current situation.16 Non-neutrality of government interventions in the housing market is not without costs and risks. A preferential fiscal treatment of housing requires higher tax rates elsewhere, with distortionary effects on the economy. The practice of exempting gains on owner-occupied housing
166 Housing Market Challenges in Europe and the United States
for the capital gains tax after having owned it for a number of years locks in the investment for that period, making these funds not available for other, possibly more productive, purposes (‘crowding out’). Limited effectiveness of housing subsidies is another concern; as supply in the short term tends to be rather inelastic, increased housing demand following higher fiscal subsidization may be mainly reflected in higher housing prices: Fiscal subsidies thus may have a redistributive element, generating potential capital gains for current owners at the expense of newcomers on the market, thus actually hindering their access to the housing ladder.17 Freeman (1997) compared housing taxes and subsidies in 12 OECD countries with tenure patterns, and found that tenure patterns are associated with tax and subsidy frameworks, but there are many more factors operating. Fiscal factors are found to affect in particular the tenure choice of young people.
8.3 Influencing housing market developments via tax instruments The fiscal instruments that are available to national governments may not only be used to promote home-ownership but also as instruments to avoid or to correct housing market imbalances. This has become more relevant now as euro area countries no longer have national monetary policies at their disposal, and the macroeconomic importance of mortgage debt is growing with its size. The costs and risks outlined in the previous section increase the weight of the stabilization role of these fiscal instruments compared to its social goals. As mortgage developments are linked to underlying housing market developments, this section also includes measures that impact on the demand and supply of owner-occupied dwellings. In analyzing government measures, we will only present some general considerations, as its concrete use should ultimately depend on countryand situation-specific features. For house price increases, for instance, the elasticity of housing supply matters a great deal: if housing supply is very inelastic in the short and medium-term, increasing demand for houses on account of more generous fiscal subsidies is likely to be reflected to a large extent in higher house prices rather than in new supply. Thus, a thorough understanding of underlying causes of housing market imbalances is also required. Below, we first discuss some structural government policies to influence growth of mortgage demand and house prices and next some specific fine-tuning measures. The discussion focuses on measures most
Guido Wolswijk 167
commonly referred to, without however excluding the use of other tax instruments.
8.3.1 Structural reduction of housing market volatility A number of fiscal measures can be considered that counteract unwarranted high levels or sharp fluctuations in mortgage debt or house prices. Among them are the following: • High stamp duties can be used to stabilize housing markets: increasing
stamp duties raise the costs of moving, thus limiting housing speculation by reducing liquidity. However, it may distort the choice of (potential) house owners and negatively impact on labor mobility.18 Relatively high levels of stamp duties in a few euro area countries suggest little scope for further increases without adverse economic consequences.19 • Increasing the tax base in line with house prices (‘automatic stabilizers’) can also be considered, for example via a regular updating of market values of dwellings as the tax base for real estate tax, property tax, wealth tax and/or tax on imputed rent. Regular updating implies that taxes are levied on recent market-based values. Thus, house price booms would automatically raise tax liabilities, increasing the costs of housing and dampening demand. A progressive rate structure would strengthen this effect. However, the measure may also raise tax liabilities of lowerincome house owners in case of strong price increases, requiring them to cut on other consumption or even to give up the house as increasing taxes may make home-ownership less affordable. • The capital gains tax in most countries already includes an ‘antispeculation’ clause, linking exemption of house price gains from this tax to a minimum number of years of owning and occupying the house. Increasing this number provides fiscal incentives to keep dwellings longer. However, it also increases the cost of moving, which negatively affects labor mobility rates. • Reductions in mortgage interest tax relief, lowering demand at given interest rates, could also dampen housing market volatility: House price variability is highest in countries with the largest (negative) tax wedges according to Van den Noord (2003). At the same time, reducing mortgage interest deductibility will increase the interest-sensitivity of mortgage demand and – to the extent that this subsidy is capitalized – will cause capital losses for current owners. Being a regressive instrument and causing a bias towards debt-financing are additional
168 Housing Market Challenges in Europe and the United States
arguments sometimes put forward to restrict this type of subsidy.20 Also, the effectiveness of this instrument has been questioned.21 • Apart from the general reservations on government interventions in the housing market (higher tax rates, higher house prices if supply inelastic, crowding out of financing for other purposes), there are some specific disadvantages of using interest relief as a means to promote house ownership. Being a tax expenditure, its hidden character on the budget makes it less transparent, hindering the priority-setting in parliament. In addition, as its financial advantage increases with the marginal tax rate (if not deductible at a fixed rate), it operates regressive, favoring high-income earners in particular. However, capitalization of these benefits in house prices makes introducing limits on interest deductibility politically difficult. • The private and the social rental sector may act as shock absorbers, helping to contain housing market shocks. If rents do not react instantaneously to house price changes, the rental sector offers an alternative for those preferring to wait for a housing boom to pass over. Changes in the often unfavorable tax positions of landlords may be needed to increase rental housing supply. In 2003, Spain introduced tax measures including a tax rebate on the net return on rentals for the owners of rental houses. In implementing changes, the timing and size of measures is crucial to avoid large disruptions to housing markets. This may reduce large negative macroeconomic effects, and has the advantage of limiting the extent to which homeowners are hit twice: Via reduction of the subsidy and via lower house prices. The spreading out of measures reducing the favorable tax-status of house owners over a longer time-period, as in UK, gives a good example of how substantial changes to the fiscal treatment of housing can be introduced without causing macroeconomic or housing market disruptions.22 8.3.2 Fine-tuning housing markets While the above measures can be considered for reducing housing market volatility on a structural basis, governments could also consider finetuning measures. Such, for instance, has been proposed by Ahearne et al. (2008), having governments using housing tax systems in a countercyclical way. In particular, they argue that when the ECB monetary policy stance is too lax for a particular country, property taxes and capital gains taxes could be increased, and a tax could be levied on interest
Guido Wolswijk 169
payments for flexible rate mortgages.23 Measures suggested in general in the literature include: • Varying stamp duties according to the housing market conditions may
be considered. Raising rates in a housing boom increases transaction costs of moving, and therefore will have a negative effect on housing market activity. In the UK, a stamp duty ‘holiday’ was implemented in 1992 when the housing market was cooling off substantially.24 Its effectiveness may be limited, as costs can be spread out over the duration of the mortgage contract, as there are numerous other costs involved in changing residence, and such expenses can be included in the mortgage amount in a few countries. • Changes in property tax rates can be used to fine-tune housing market developments. In a housing boom, rates could be raised, increasing the costs of owning a house, and therefore reducing housing (finance) demand. An advantage over stamp duties is that it does not affect costs of (labor) mobility. For the UK, Maclennan et al. (1999) propose using this tax rate to reduce house price volatility, and making the central bank responsible for setting the property tax rate. • Government subsidies may be temporarily raised or lowered with a view to stabilizing housing markets. Portugal, for instance, tightened eligibility requirements for mortgage subsidies directed towards low-income families in 1999, with a view to avoiding overheating the housing market. Measures can also be taken on the supply side, extending or reducing temporary investment incentives. In general, the feasibility of effective fine-tuning of housing markets via budgetary instruments can be questioned. Apart from judging whether housing markets are far out of equilibrium and the question of whether government intervention is needed, effective fine-tuning requires good timing and appropriate calibration of the government response. A measure intended to prevent a housing bust may aggravate a downturn in the housing market if it comes too late, and thus act pro-cyclically. Very little empirical work has been done in this area but Stephens (1996), surveying the UK government responses during the housing market recession in the early 1990s, found three out of the eight government measures to be anti-cyclical (for instance, the increase in the stamp duty threshold), two neutral, and two pro-cyclical.25 It also needs to be taken into account that fine-tuning measures may create uncertainty in the market and if anticipated, may become a
170 Housing Market Challenges in Europe and the United States
source of volatility. Expectations of forthcoming restrictions of housing subsidies may induce a strong increase in housing transactions now, and a steep drop after its actual introduction. Finally, budgetary effects may be asymmetric as it is politically easier to lower tax rates in housing market upturns than to raise them in downturns. While structural measures to reduce volatility therefore as a rule appear preferable, exactly which measures to use in a given situation and country very much depends on national characteristics of the mortgage and housing markets, and the fiscal instruments available. In applying them, due account has to be given to other considerations, notably effects on labor mobility and possible distortionary effects in housing markets. In addition, disruptive experiences, as in Sweden in the early 1990s with large tax changes (Jonung et al., 1996) and favorable experiences in the UK with gradually phasing out interest relief for home-owners, suggest that some spreading out of major fiscal reforms over time is preferable to avoid large disruptions to housing markets, which could have adverse effects on the economy and on financial stability. In some cases, applying stricter rules only to new cases (‘grandfathering’) can be considered.
8.4 Conclusions Fiscal aspects of mortgage debt and house price changes are sometimes underexposed in discussing measures to limit housing market volatility. Assuming an intervention in the housing market to be required, either for preventing or for correcting some disequilibrium, fiscal instruments may be useful as housing markets still are predominantly national in nature. Such measures need to be targeted to the specific situation and national circumstances. Also given large differences in the tax treatment of housing in the euro area, there is not a single recipe for all countries. Structural fiscal measures such as reducing mortgage interest relief and increasing reliance on quick adjustments of tax bases to market price developments appear to be useful options for governments for limiting growth rates of mortgages and house prices. Additional considerations may support such actions such as the question of whether housing (still) needs to be fiscally favored, limited effectiveness of some instruments, and possible negative by-effects in the form of high tax rates or unwarranted income distribution effects. In general, it seems preferable
Guido Wolswijk 171
to spread out significant structural changes in fiscal instruments over time to avoid market disturbances which could have large adverse consequences for macroeconomic and financial stability. More reservations are needed for applying fiscal fine-tuning measures on the housing market, as doubts about identification of a bubble, and about timing and calibration of measures prevail. Such operations therefore may be more appropriate in exceptional circumstances.
Annex: Main tax instruments affecting housing markets in euro area countries, 2007 Table A8.1
Tax on imputed rents
Country
Yes/No
Remark
Austria
N
Belgium
Y
Cyprus
N
Finland
N
France
N
Abolished in 1965
Germany
N
Abolished in 1987
Greece
N
Abolished in 2003
Ireland
N
Italy
Y
But owner-occupied dwelling largely exempted
Luxembourg
Y
But tax amount based on outdated values
Malta
N
Netherlands
Y
Portugal
N
Slovenia
N
Spain
N
PM: US
N
On the basis of average net rent values
Up to 0.55% of market value, with a maximum
Unless not occupied by owner, then imputed rent due
Source: International Bureau of Fiscal Documentation (2007).
172 Table A8.2
Mortgage interest payments tax deductibility
Country
Yes/No
Maximum tax rate
Remarks
Austria
Y
50%
Tax deductible as special expense up to a limit that goes to 0 as annual income increases
Belgium
Y
50%
Tax deductible up to a limit for first 10 years, and lower limit for years after, provided mortgage maturity is at least 10 years
Cyprus
N
–
Finland
Y
28%
France
N
–
Germany
N
–
Greece
Y
40%
Tax credit of 20% for annual mortgage interest for 1 principal home in lifetime, subject to limits.
Ireland
Y
42%
Italy
Y
45%
Luxembourg
Y
38%
‘Tax credit’ at 20% standard income tax rate with (low) limit, which increases a bit for first-time buyers for first 7 years Tax credit of 19% for annual mortgage interest on owner-occupied dwelling, with limit Tax deductible, with (low) limits, also depending on length of time of owneroccupation
Malta
N
–
Netherlands
Y
52%
Tax deductible without limit for mortgage for owner-occupied dwelling, for at most 30 years
Portugal
Y
42%
Tax credit of 30% of mortgage interest and amortisation for permanent home, with a limit
Slovenia Spain
N Y
– 43%
PM: US
Y
35%
Normally deductible from capital income that is subject to income tax if mortgage for owner-occupied dwelling Gradually abolished over the period 1991– 2000. Only tax credit in very special cases for loans before 1998 Abolished in 1987 with the introduction of a subsidy scheme
Tax credit of 15% of principal and interest paid to offset against tax liabilities with a limit if owner-occupied for at least 3 years Tax deductible up to limit (1 mil US$)
Source: International Bureau of Fiscal Documentation (2007).
173 Table A8.3
Tax on housing capital gains
Country
Tax in place?
Tax rate (max) applicable*
Remark
Austria
Y
50%
No taxation if sold more than 10 years after acquisition
Belgium
Y
16.5%
No taxation if sold more than 5 years after acquisition
Cyprus
Y
20%
Exemption if owner-occupied for at least 5 years, with cumulative lifetime limit
Finland
Y
28%
No taxation if used 2 years or more as primary residence
France
Y
16%
No taxation if principal residence
Germany
Y
45%
No taxation if sold more than 10 years after acquisition
Greece
N
–
Ireland
Y
20%
No taxation on gains from sale of private residence unless due to development
Italy
Y
43%
Principal dwelling of owner excluded and else excluded if dwelling held for 5 years or more
Luxembourg
Y
9.7%
Main residence exempted from taxation
Malta
Y
35%
Exemption if owner-occupied for at least 3 years
Netherlands
N
–
Portugal
Y
42%
Exemption for permanent dwelling if total gain reinvested in another permanent residence within 2 years
Slovenia
Y
20%
Exempted if acquired before 2002 or if used more than 3 years as permanent residence
Spain
Y
18%
No taxation if reinvested in new primary residence
PM: US
Y
25%
No taxation if owned-occupied during 2 of the last 5 years, with upper limit
–
Taxed at a presumptive return rate (4%)
Note: * Tax rates are not well comparable as they disregard thresholds, possible deductions, tax progression, the number of years the dwelling has been owned and occupied, and local surcharges. Source: International Bureau of Fiscal Documentation (2007).
174 Housing Market Challenges in Europe and the United States Table A8.4
Indirect taxes
Country
Austria Belgium Cyprus Finland Franceα Germany Greece Ireland Italy Luxembourg Malta Netherlandsβ Portugal Slovenia Spain PM: US
VAT on house supply
Taxed as other durables?
– 21% 15% – 19.6%/– 19% 13.5% 10% 15% – 19%/– 20% 7% Sales tax
N Y Y N Y/N N Y N N Y N Y N Y N Varies per State
PM: standard VAT rate 20% 21% 15% 22% 19.6% 19% 19% 21% 20% 15% 18% 19% 21% 20% 16% –
– denotes VAT exemption. Notes: α Exemption for first transfer of dwelling after 5 years of completion, otherwise taxed. β Exemption if newly constructed dwelling is sold at least 2 years after first actual use. Source: International Bureau of Fiscal Documentation (2007).
Notes 1. European Central Bank, DG-Economics. Opinions expressed in this chapter reflect the views of the author and not necessarily those of the ECB or the Eurosystem. 2. Housing market developments also have an effect on governments’ budgetary positions. See Schuknecht and Eschenbach (2002) for a description of the channels of impact of asset prices on fiscal variables. 3. Finland and Sweden experienced large negative shocks in the early 1990s, which were partly fiscally-induced, with substantial effects on the national economy and on financial stability. See Jonung et al. (1996) for details. 4. Housing (finance) markets are still predominantly national, despite some internationalization of the funding of mortgage loans. As (national) tax systems have a major influence on housing loan characteristics, the effectiveness of national instruments remained broadly unchanged. The European Commission is working on proposals to speed up the integration of EU mortgage markets, see European Commission (2007). 5. The contribution of M. Pareja-Eastaway and T. Sánchez-Martínez to this volume (Chapter 7) deals with the rental sector in more detail.
Guido Wolswijk 175 6. Maclennan et al. (1999) mention crowding out by the high government debt in Italy as a factor that could explain low mortgage debt to income ratios in that country. 7. The situation described applies to purely domestic transactions. Cross-border issues are detailed in EMF (2004). 8. A more recent but less detailed description of taxation of housing is included in ECB (2009). 9. For the US, Rosen and Rosen (1980) estimated that one-quarter of the US postSecond World War increase in home-ownership can be attributed to fiscal subsidization of housing, with non-taxation of the rental value comprising the main element. 10. The most notable example is Luxembourg, basing its taxation on 1941 values, even for houses that had not been built by that time. 11. Data taken from Van der Hoek and Radloff (2007). 12. See ECB (2009) annex 4 for more details about interest deductibility of mortgage interest rates in the euro area countries. 13. As to non-mortgage interest deductibility, Maki (2001) provides evidence for the US that abolishment of the tax deductibility of interest payments on consumer credit increased loan demand with houses as collateral. 14. The benefit of the interest deductibility reduces with falling inflation, even assuming real interest rate remaining equal, as deductibility refers to the nominal rather than the real interest rate. 15. Total costs of buying a house financed with a mortgage for a selection of euro area countries (Belgium, Germany, Greece, Spain, France, Italy) have been estimated to range between 5 (Germany) and 17 per cent (Belgium) in 2004, with property taxation usually explaining the bulk of the costs (EMF, 2006). 16. Greece, for instance, in 2003 abolished taxation of imputed rent, bringing it more in line with the favorable tax status of housing in other countries. 17. For the Netherlands, a recent study concludes that almost 75 per cent of the fiscal subsidy given via mortgage deductibility is reflected in house prices (Brounen and Neuteboom, 2008). 18. Estimates for the Netherlands indicate that a 1 per cent point increase in transaction costs (stamp duties, capital gains tax and brokerage fees) relative to the house price decreases homeowners’ mobility by 8 per cent (Van Ommeren and Van Leuvensteijn, 2003). 19. In Belgium, total property transfer taxes and registration fees added up to 121/2 per cent of the property price in the early 2000s, while in Italy it was 11 per cent (see Belot and Ederveen, 2005). 20. In this context, the French Constitutional Council in 2007 declared a planned introduction of mortgage interest deductibility as unconstitutional, saying it would create a tax advantage beyond what is needed to encourage nonhomeowners to buy homes, noting that the deduction would also apply to current homeowners as well. 21. Literature on this, focusing on the US, is inconclusive. For instance, Follain and Dunsky (1997) conclude for the US that mortgage debt demand is highly responsive to the rate at which interest on mortgage debt can be deducted; but Glaeser and Shapiro (2002) conclude only a small impact of mortgage interest deductibility on home ownership. For Italy, Jappelli and Pistaferri (2007) studied effects of the decoupling of the marginal tax rate and the
176 Housing Market Challenges in Europe and the United States
22. 23.
24. 25.
after-tax mortgage rate in the 1990s, and concluded that mortgage debt demand of high-income taxpayers was not affected. The UK gradually phased out mortgage interest relief in the period 1974– 1999, among others, by not adjusting relief ceilings to inflation. It is not clear from the Report whether measures working in the opposite direction should be taken when ECB monetary policy would be considered too tight for a country. More recently, in September 2008, the UK government decided to raise the exemption level for stamp duties below which no tax needs to be paid. He also considered policy changes by lenders, building societies and insurance companies, and assessed most of their measures to be pro-cyclical, e.g. tightening capital adequacy requirements during housing recessions, widening margins to cover for provisions and write-offs, and lowering loan-to-value ratios.
References Ahearne, A., J. Delgado and J. von Weizsacker (2008) A Tail of Two Countries, Bruegel Policy Brief, No. 4, June. Belot, M. and S. Ederveen (2005) Indicators of Cultural and Institutional Barriers in OECD Countries, CPB Memorandum, The Hague. Bourassa, S. and W. Grigsby (2000) Income Tax Concessions for Owner-Occupied Housing, Housing Policy Debate 11, 521–46. Brounen, D. and P. Neuteboom (2008) De Effectiviteit van Hypotheekrenteaftrek (Effectivity of Mortgage Interest Deductibility), ESB (Economic-Statistical Bulletin), 120–21. Contact Group on Asset Prices (2002) Turbulence in Asset Markets: The Role of Micro Policies. Diamond, D. and M. Lea (1992) Housing Finance in Developed Countries: An International Comparison of Efficiency, Journal of Housing Research 2 (1). ECB (2003) Structural Factors in the EU Housing Markets, March. ECB (2009) Housing Finance in the Euro Area, ECB Occasional Paper, No. 101. EMF (2004) Tax & Subsidy Related Problems When Taking Out a Mortgage Loan Across an EU Border, Brussels. EMF (2006) Study on the Cost of Housing in Europe, Brussels. European Commission (2007) White Paper on the Integration of EU Mortgage Credit Markets. Follain, J. and R. Dunsky (1997) The Demand for Mortgage Debt and the Income Tax, Journal of Housing Research, 8(2), 155–99. Freeman, A., A. Holmand and C. Whitehead (1996), Is the UK Different? International Comparisons of Tenure Patterns, Council of Mortgage Lenders. Freeman, A. (1997) A Cross National Study of Tenure Patterns, Housing Costs and Taxation and Subsidy Patterns, Scandinavian Housing and Planning Research, Issue 4, 159–74. Glaeser, E. and J. Shapiro (2002) The Benefits of the Home Mortgage Interest Deduction, NBER Working Paper, No. 9284.
Guido Wolswijk 177 Hendershott, P. and M. White (2000) Taxing and Subsidizing Housing Investment: the Rise and Fall of Housing’s Favoured Status, Aberdeen Papers in Land Economy, No. 2000–5. International Bureau of Fiscal Documentation (2007) European Tax Handbook 2007, Amsterdam. Jappelli, T. and L. Pistaferri (2007) Do people respond to tax incentives? An Analysis of the Italian Reform of the Deductibility of Home Mortgage Interests, European Economic Review, 51 (2), 247–71. Jonung, L., H. Söderström and J. Stymne (1996) Depression in the North. Boom and Bust in Sweden and Finland, 1985–1993, Finnish Economic Papers, 9, 55–71. Maclennan, D., J. Muellbauer and M. Stephens (1999) Asymmetries in Housing and Financial Market Institutions and EMU, CEPR Discussion Paper, No. 2062. Maki, D. (2001) Household Debt and the Tax Reform Act of 1986, American Economic Review, 91, 305–19. Neuteboom, P. (2004) A Comparative Analysis of the Net Costs of a Mortgage for Homeowner in Europe, Journal of Housing and the Built Environment, 19(2), 169–86. OECD (2007), Revenue Statistics 1965–2006, Paris. Rosen, H. and K. Rosen (1980) Federal Taxes and Homeownership: Evidence from Time Series, Journal of Political Economy, 88 (1), 59–75. Schuknecht, L. and F. Eschenbach (2002) Asset Prices and Fiscal Balances, ECB Working Paper, No. 14. Stephens, M. (1996) Institutional Responses to the UK Housing Market Recession, Urban Studies, 33 (2), 337–51. Van den Noord, P. (2003) Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence, OECD Economics Department Working Paper, No. 356. Van der Hoek, M. and S. Radloff (2007) Taxing Owner-Occupied Housing: Comparing the Netherlands to other European Union Countries, Public Finance and Management, 7 (4), 393–421. Van Ommeren, J. and M. van Leuvensteijn (2003) New Evidence of the Effects of Transaction Costs on Residential Mobility, CPB Discussion Paper, No. 18. Wolswijk, G. (2006) Determinants of Mortgage Debt Growth, European Journal of Housing Studies, 6.
9 Towards a New Housing System in Transitional Countries: The Case of Hungary1 József Hegedüs
9.1 Introduction: Dissolution of the East European Housing Regimes The soviet bloc collapsed in 1989/1990, since when the political map of the region has changed; new independent countries have emerged, not always in the same peaceful way as the Czech Republic and Slovakia. Political systems were transformed, multi-party systems were introduced everywhere in the region, but the one-party systems were not automatically replaced by modern democracies. The political changes were followed by a huge economic recession, but after 5–10 years Central European countries recovered, reached the pre-transition level, and the economies of the region started to be integrated into the world economy (as part of the globalization process). Societies have also undergone transformation: income inequalities have increased substantially, slowly a middle class has formed, and a new elite has emerged partly from old party nomenclatures. The social and economic conflicts of the transition coerced governments into introducing changes in the welfare system, as the old soviet type of ‘social safety net’ collapsed and new measures needed to manage the challenges of the transition. The analysis of the development of the post-socialist housing system touches upon the huge welfare regime literature, which started with the famous Esping-Andersen typology (1990). Firstly, in what way is housing part of the welfare regime? As in the original Esping-Andersen typology housing did not play a role, though later there were several attempts to integrate housing in the typology. However, it proved to be quite difficult because of the special characteristics of housing as a commodity. Housing was considered to be a ‘wobbly pillar’ in the welfare regime theory (Torgersen, 1987). Recent literature 178
József Hegedüs 179
on relations between housing and the welfare system showed the significance of housing as a form of social security, a way of redistributing incomes between social groups and generations, a means to reproduce the system of social care based on intra-family relations (Poggio, 2008; Stephens and Fitzpatrick, 2007; Malpass, 2008). Secondly, what kind of welfare regimes do transitional countries move to? Is there a new model which combines the elements of the modern European welfare regimes with the ‘socialist tradition’? There is no agreement among researchers about whether it is possible to draw any general conclusions related to the development of welfare systems in transition countries. The problem is that public sector reforms (education, health-care system, etc.) are inchoate, the measures and institutions in different areas of welfare sectors follow different principles even within one country, and there is a big gap between the rhetoric of programs and their actual implementation. Most of the research is non-conclusive, which is a clear sign of the ongoing conflicting processes in welfare politics (Tomka, 2005; Ferge, 2002; Cerami, 2005; Tausz, 2009). We tend to share the view of Kasza (2002, p. 284), who argues that ‘it might be a better strategy to focus comparison on particular policy fields . . . and it remains to be seen that a manageable typology will emerge from such policy-specific research’. There are too many inconsistencies among the different policy areas to make it possible to put them under the same regime category. The chapter aims to contribute to the research on the role of the housing sector in transition countries. There are several studies and publications dealing with changes in the housing sector,2 but there is no answer to the question of whether these changes point towards a specific post-socialist housing system. The aim of this chapter is to give an overview of the housing market development in Hungary in the context of transition countries and to attempt to describe the main characteristics of post-socialist housing regimes. The East European Housing Model (Hegedüs and Tosics, 1996) summarizes the main characteristics and elements of the housing system in the centrally planned economy, which was a social-economic system with high job security,3 low – highly subsidized – housing costs, and small income differences. In this housing system, a vast majority of services was provided ‘in kind’ or at an under-cost/market price allocated according to ‘merits’ (Kornai, 2000). The main characteristics of this model were single party political control over the housing sector, the subordinate role of market mechanisms, lack of competition among housing agencies (bureaucratic coordination), and broad control over the allocation of housing services (huge, non-transparent subsidies). However,
180 Housing Market Challenges in Europe and the United States
under this model several ‘sub-models’ (versions) emerged as responses of the individual countries to challenges in the process of developing the socialist economy (Turner et al., 1992). While the main characteristics of the model could be explained as a consequence of the centrally-planned economy (priority of industrialization, socialist capital accumulation, controlled urbanization, etc.) through a hard structuralist approach, the divergences of the model were considered theoretically as ‘policy options’ taken by individual societies through the interaction of agencies under the soft control of the communist party.4 The structural conflicts (‘cracks’ in the model, such as housing shortage, inefficiency of central planning, quasi-market processes, existence of informal economy) were managed by different methods, e.g., introducing strict control mechanisms (Bulgaria, Russia, East Germany) or allowing quasi-market processes (Yugoslavia, Hungary). These variants of the model were different from each other in terms of tenure structure (state-owned rental, cooperative housing, private rental and owner-occupied sector) or in terms of housing finance schemes working under the state-controlled bank sector. Despite differences, the common structural features have left their imprint upon the housing systems (e.g., housing estates, poorlymaintained public housing, and rationed ‘elite’ houses for the nomenklatura), which justifies the use of the term ‘East European housing model (EEHM)’. The crucial question is how EEHM has been changed after the transition. Is it possible to define a post-socialist housing model or will the emerging new housing model be just similar to one of existing housing models in the developed world? To give an answer to this question we describe the development of the housing sector (investment, inflation, house prices, etc.) in Hungary in the context of macroeconomic changes (Section 9.2). Then we offer an overview of the main stages of the housing policy in Hungary after 1990 (Section 9.3). The next two parts will analyze the controversial effects of two housing programs started in 2000, firstly the mortgage market program (Section 9.4), secondly the social rental program (Section 9.5). The fifth part of the chapter will show the importance of overlapping policy areas (safety net and income benefit program) in housing policy. In the concluding part of the chapter we will summarize the main elements of the post-socialist housing systems.
9.2 Macroeconomic recession and the Hungarian housing sector The political transition had an enormous effect on macroeconomic trends. As a result of the transitional recession the GDP fell by 15 per cent
József Hegedüs 181
40
GDP ( % change on previous year) CPI ( % change on previous year)
30
Interest rate on housing loans %
20 10 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
⫺10 ⫺20
Figure 9.1 Macroeconomic trends: GDP, CPI, and interest on housing Source: Hungarian National Bank, Central Statistical Office.
in the first part of the 1990s. Structural reforms (privatization and bank reform) were complemented with a strong fiscal stabilization package (1995–98) and the maintenance of sound macroeconomic policies. In the 1990s, the labor market changed, owing to the closing down of many previously state-owned companies and to the restructuring of the production sector, the employment rate decreased dramatically and more than 1 million employees left the job market, many of them pushed into early retirement or into forced entrepreneurship with close ties to the informal economy. The informal economy was estimated to be as large as 25–33 per cent of GDP between 1990 and 1997 (Laczkó, 2000), and informal transactions are widely accepted today as well by consumers (VAT tax evasion), employees (as wages paid directly into ‘pocket’, 25 per cent of all employees are affected), service providers (tax evasion), etc. (Semjén et al., 2008). During the 1990s poverty became one of the most important social issues. While the average income decreased, income inequalities increased dramatically in the first part of the 1990s, and remained stable over the last decade. The ratio of average income in the lowest percentile to the average in the highest percentile increased from 4.6 in 1987 to 7.6 in 2004 (Keszthelyiné Rédei and Szabó, 2006). The economic recession associated with transition increased regional inequality in the country: the eastern part of the country was hit much more by the economic decline than were other areas. The net income per capita is 50 per cent higher in the most-developed region than in the two less-developed regions.
182 Housing Market Challenges in Europe and the United States Table 9.1
1970 1980 1990 2000
Basic indicators of housing conditions in Hungary, 1970–2000
Number of units (in millions)
Persons per 100 units
Number of rooms per 100 units
Units with 3 or more rooms ( %)
Units without comfort (%)
3.14 3.55 3.86 4.06
327 302 274 274
164 199 237 263
10.8 24.3 40.5 45.9
65.9 37.7 18.7 15
Share of public rental housing (%) 26.2 26 19 3.7
Source: Central Statistical Office.
Hungary’s population has been constantly decreasing since the early 1980s. Today its population is slightly above 10 million. The average household size has been stable around 2.65, which is slightly larger than the European average (around 2.53). The total fertility has decreased from 1.92 in 1980 to 1.32 in 2007, which is among the lowest levels in Europe and exacerbates the ageing of the Hungarian society to a large extent. Despite generous child benefits,5 paid maternity leave, etc., the figure has been constantly decreasing from the 1980s with a speed-up in the second half of the 1990s. The Hungarian demographic situation can be characterized with low fertility, high mortality, and increasing pace of ageing. Besides natural demographic movements, the nature of regional migration has also changed in Hungary: both direction and scale have altered since the transition (1990), which mostly meant a flow of residents from villages to towns and large migration to the capital city. Since the 1990s there has been no significant demographic pressure on the housing market; the quantitative shortage of housing had ceased to exist by the end of 1980s. Urbanization slowed down and inner-migration trends did not increase the demand for housing substantially. The housing situation improved in the 1990s partly because there was no demographic pressure on the housing sector, and fulfilling the demand for housing had been postponed (see Table 9.1). Housing units with three or more groups did not increase, and the large share (15 per cent) of the obsolete stock is considered to be critical. As a consequence of the housing privatization, 15–20 per cent of the total housing stock moved from state ownership to the owner-occupied sector. In the Hungarian housing system the state rental sector had a 25 per cent share before the transition. Its role was decisive in urban areas, where privatization caused a dramatic change in the tenure structure.
József Hegedüs 183
70 000 60 000 50 000 40 000 30 000 20 000 10 000
New construction
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
0
Building permit
Figure 9.2 New construction and building permits between 1989 and 2008 Source: Housing Statistics, 1988–2008, CSO.
150 140 130 120 110 100
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
90
Estimated house price increase (previous year ⫽ 100) Consumer price index (previous year ⫽ 100) Figure 9.3 House price changes Source: CSO, HNB and author’s calculations.
Privatization was basically a ‘give away’ privatization where the actual price was around 10–15 per cent of the market value. As a consequence, in 2000 the estimated share of the rental sector was 8 per cent, of which 4 per cent was public.
184 Housing Market Challenges in Europe and the United States
Because of the economic depression, housing investment decreased and housing output decreased by more than 50 per cent as well. The change in the subsidy system in 1994 had a short-term effect on the housing output, but by 1998 the number of the new units had dropped even to a lower level than before. Summarizing the trends in the 1990s, we can point out that as a consequence of the recession households’ demand for housing decreased, which was supported by the easing demographic pressure on the market. Parallel to this trend, real house prices decreased also (see Figure 9.3).
9.3 Housing policy regimes After the political changes at the end of the 1980s, three stages of the housing policy can be outlined. In the first period (1989–94), the Hungarian government tried to manage the housing crisis related to the economic decline and the unsustainable subsidy system of the socialist period. The government ‘moved out from the housing sector’ decreasing the subsidies and diminishing its direct role. Decentralization was part of this process as the local governments were assigned to manage the housing allowance program partly financed from their own resources. The housing policy of this period could be characterized basically as crisis management: the two major programs were the privatization of the state rental sector and the consolidation of the ‘old loans’. Both measures had a regressive social effect: The financial gain of privatization and early repayment of the loans were proportional to the households’ wealth, thus low-income households were trapped in the residualized social rental sector or were not able to pay back their mortgage at a discount price. The Law on Rental sector (1993) and the Social Law (1993) made it clear that the government does not take responsibility for housing, but leaves it open for a future intervention. In the 1990s the officially measured housing subsidies reached 3.7 per cent of the GDP, and of the total homeowner subsidies in 1990 more than two-thirds went to the interest subsidy of the ‘old loans’. In 1993/94 the subsidies related to borrowing were reduced substantially, under 1 per cent of the GDP. In the second period (1995–2000) new institutions were set up and the legal background improved. Meanwhile, the level of subsidies gradually decreased as a consequence of the decreasing housing output. Two basic housing financial institutions were established: contract savings banks and mortgage banks. The law on contract savings banks was very controversial as the subsidies given to savers made the housing subsidy
József Hegedüs 185
system even more regressive, and there was no direct relation between the subsidies and the increase in housing investments. The change of the subsidy system to compensate the big cuts in 1994 had a temporary effect on housing output6 (see Figure 9.2). The government’s housing policy actions took on the recommendations of the World Bank and other international agencies, and laid down the principles of separation of the subsidy system from housing finance and the principle of targeting subsidies to households in need. From 1998 (when a new right-wing government came into office), a new rhetoric was introduced into housing policy, namely the need for supporting middle-income households, but for two years nothing important had happened. In spite of different approaches in housing policy actions, no real changes in the subsidy system and housing programs took place in the period. The third period started after 2000, when the Hungarian government started an active program backed by the positive macroeconomic changes. The government housing program of 2000 targeted three housing areas: 1) emergence of the mortgage market; 2) expanding the local government social housing, and 3) renewal of the housing stock (especially the urban housing estates built in the 1970s and 1980s). A new subsidy system was introduced for housing mortgage (‘dual’ interest rate subsidy, tax exemption after mortgage payment, etc.), which resulted in an unprecedented increase of housing loans from 2001. A grant program was initiated for local government social housing investment, which covered 75 per cent of the costs for social housing investment. Preferential loans were offered for renovation of housing estates as well. Because of the fiscal cost of the program, the new socialist-liberal government elected in 2002 tried to slow down these programs, but it took almost three years to make radical changes in the subsidy schemes. From 2005 the interest rate subsidy was cut, the grant program for social housing stopped, but the housing estate renovation expanded. Against expectation the mortgage market did not collapse because of these changes, as competing banks offered cheap unsubsidized foreign currency loans, so the growth of the mortgage market has not stopped. The investment grant for social housing programs was replaced by a rent allowance program, which did not work. However, renewal of housing estates, the less expensive program for the government, became very successful, and in the years 2005–08 15 per cent of the stock was involved in it. As a consequence of the 2008 financial crisis, the housing market was hit as well. Borrowing stopped, the quality of loans deteriorated, mortgage payment arrears appeared, etc. The government issued announcements and initiated a program to help households. But it is yet to be seen
186 Housing Market Challenges in Europe and the United States
how deep the crisis will be, and how effective the government assistance package will prove.
9.4 Housing program from 2000 – a ‘great leap’ and corrections As a result of the austerity program, economic performance improved. The real GDP grew at an average rate of 4.4 per cent over the period of 1997–2001, and by 2001 the GDP had reached the level of the pretransition period. Unemployment had been on the decrease from the middle of the 1990s, and stabilized at 5–7 per cent at the turn of the century. By the end of the 1990s, the sign of a housing crisis presented itself clearly in contrast to the economic recovery. The housing output reached its minimum in 1998, the public housing stock was privatized and local governments had no means for social housing; housing subsidies were cut and no housing finance system was set up; the deterioration of the multi-unit urban housing stock (housing estates) speeded up; increasing housing costs (energy, water and other housing services) caused serious affordability problems. In the new housing policy of the right-wing government (1998–2002) the most significant element was the support to housing mortgage introduced in 2000. There were two different types of interest rate subsidies: (a) an interest rate subsidy to mortgage bonds, and (b) an interest rate subsidy for loans connected to new construction. (See the details of the programs in Hegedüs and Somogyi, 2005.) The third element in the subsidy program was the Personal Income Tax (PIT) mortgage payment allowance. From 1994, the maximum amount that could be deducted from tax payment was 20 per cent of the mortgage payment (interest and capital), maximum HUF 35 000 per year.7 Its effect was not all that influential, as the ceiling remained low until 2000. In 2001, the maximum deduction was increased to 40 per cent, and HUF 240 000 per year (around a1000 per year) in case of new construction. In the original proposal (in 2000) both mortgage subsidy schemes and the PIT allowance was at a low level, which gave a manageable impetus to the development of mortgage finance. This policy would have helped the upper middle and middle-income groups to have access to housing loans with a relatively ‘shallow’ subsidy system. In this way it would have been possible to target the remaining part of the subsidies to the needy social groups. It seemed to be a reasonable price for building up a modern housing finance system. However, the government had been under constant pressure from lobby groups, and the conditions and eligibility
József Hegedüs 187
Total housing loan (billion HUF)
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
18% 16% 14% 12% 10% 8% 6% 4% 2% 0% 1989
4500 4000 3500 3000 2500 2000 1500 1000 500 0
Housing loans (%ofGDP)
Figure 9.4 Housing loans, 1989–2008 Source: Hungarian National Bank.
criteria for the two schemes had relaxed and in spring of 2002 (before the election) the personal income tax deduction was extended to loans for buying existing housing units as well. By the end of 2002 it had become clear that the volume of housing loans increased very quickly as a consequence of the subsidies. In 2002 the outstanding loans more than doubled and reached 4.5 per cent of GDP (see Figure 9.4). However, the fiscal effects of the mortgage programs were not projected correctly. In October 2002, one of the prominent economic research institutes estimated the future cost of the two mortgage subsidy programs at HUF 42 billion for the year 2005, while in fact it had reached HUF 140 billion (Molnár and Pichovsky, 2002).8 The government – for political reasons – kept postponing decisions.9 Moreover, in 2002 the government increased the subsidy for contract savings and housing policy allowances to demonstrate that they wanted to support the sector. After a long political discussion the government changed the conditions of the mortgage program by the end of 2003. The changes in the subsidy programs were influenced by the ‘construction lobby’, which argued that most of the subsidies had gone to the existing housing market, and had not helped housing investments (Varga, 2003). Tax exemptions in PIT for mortgage repayment were also cut severely. The maximum tax deduction was decreased to HUF 120 000 from HUF 240 000, and borrowers were eligible for a maximum of four years. In the case of buying existing units, the maximum deduction was equal to
188 Housing Market Challenges in Europe and the United States
1.6% 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% 0.0% 1998
1999
2000
2001
2002
2003
2004
2005
2006
Housing allowances
Rental programs
Home improvement programs
Home ownership - loan subsidies
2007
Home ownership - capital grant
Figure 9.5
Housing subsidies 1998–2007 as per cent of the GDP
Source: Hungarian government and author’s calculations.
30 per cent of the mortgage payment, while in the case of new units it remained 40 per cent. In 2007, the PIT allowance was abolished totally. Housing policy was one element of the extremely loose fiscal policy between 2001 and 2006, which resulted in a budget deficit above 9 per cent of the GDP in 2005. Housing policy continued to remain in the forefront of politics. At the beginning of 2005 the government started a new program that will have important effects on mortgage finance. A government guarantee program was introduced for young couples and public servants (‘key workers’). The government increased again the housing construction subsidy and extended its use to the market of existing units at 50 per cent of the subsidy level. These program elements were a reaction to criticisms, namely that the mortgage subsidy program had helped the relatively richer households, as its loan-to-value ratio was around 50–60 per cent, and households without intergeneration transfer and substantial savings had been left out of the program. The macroeconomic effect of the housing program from 2000 has been debated among experts. The mainstream approach (adopted by official
József Hegedüs 189
government policy in 2000) argued that the overall demand-side subsidies will generate economic growth and, after a while, will increase government revenues, which will more than compensate for the fiscal cost (Barta et al., 2003). The increase of indebtedness of households was not considered to be a problem as the household loans to GDP ratio was far from the European average. However, there were other approaches pointing out the negative macroeconomic effects, such as the increasing deficit, decreasing household savings and vulnerability of the economy because of the increasing level of indebtedness (Palocz, 2003). The indebtedness of Hungarian households is not too high in international comparison, but because of the interest spread the payment burden is quite high by international standards. The increase of housing borrowing in 2002 and 2003 was a reaction to the new housing policy, which gave a generous subsidy to middleclass homebuyers. The net value of the mortgage subsidy for two years was 50–70 per cent of the loan (taking into calculation the two interest rate subsidies and the PIT allowance). The interest rate subsidy system in the late 1980s and early 1990s had the same level of subsidy (Hegedüs and Várhegyi, 2000, p. 1637), which was not sustainable not only at the beginning of 1990s, but in 2003 also. The equity effect of the mortgage program was clearly regressive. On the basis of the PIT allowance in 2004, we estimated that the upper 20 per cent of households in income distribution get 60 per cent of the total subsidy, and the upper 40 per cent get 80 per cent of it (Hegedüs and Somogyi, 2005, pp. 199–202). Housing credit grew very fast in Hungary between 2000 and 2004, and the cut in subsidies did not stop the expansion of the market As a consequence of subsidy cuts, the issuance of mortgage loans in HUF decreased, but the total decrease was less than expected as a new foreign exchange dominated (FX) mortgage loan was introduced (see Figure 9.6). Lending in foreign currency went from close to zero in 2004 to 90 per cent in 2007. The cost of Swiss Franc loans was much cheaper and as Hungary was a member of the EU and the majority of the banks were in European ownership, concerns about the financing were dismissed.10 Until 2008 it was considered highly improbable that the value of the HUF would decrease so much that the financial advantages of the FX loan would be lost. There were no direct fiscal burdens in relation to the loans issued in FX. And even in early 2008 there were plans to introduce an incentive to convert HUF subsidized loans to FX loans in order to decrease the fiscal pressure on the government. Households preferred housing loans till 2004, but later consumer loans surpassed housing loans. Thus there were two trends in household
190 Housing Market Challenges in Europe and the United States Change of the stock of loans in billion HUF quarterly 500 400
Household loans
300
FX household loans
200 100 08Q4
08Q1
07Q2
06Q3
05Q4
05Q1
04Q2
03Q3
02Q4
02Q1
01Q2
00Q3
99Q4
99Q1
98Q2
97Q3
96Q4
96Q1
0 ⫺100
300 250
Housing loans
200
Consumer loans
150 100 50
Figure 9.6
08Q3
08Q1
07Q3
07Q1
06Q3
06Q1
05Q3
05Q1
04Q3
04Q1
03Q3
03Q1
02Q3
02Q1
01Q3
01Q1
00Q3
00Q1
99Q3
99Q1
98Q3
98Q1
97Q3
97Q1
96Q3
96Q1
0 ⫺50
Household borrowing, 1996–2008
Source: Hungarian National Bank.
borrowing: Firstly the share of consumer loans became higher in household borrowing, secondly households preferred FX loans. Both had a negative effect on the economy (see Figure 9.6). FX housing loans crowded out HUF subsidized housing loans, but at the same time there was a shift from housing loans towards consumption loans as well. There is a debate on whether these trends could be considered sustainable. The rate of the growth of retail credit responds to changing macroeconomic conditions and to expected income growth, which had increased consumption, and decreased saving rates causing imbalances in current account deficits. The competition among banks and the poor achievement of the Bank Supervision Authority contributed to loose underwriting procedures. For example, in order to adjust to the existence of the informal sector, banks did not keep the 30 per cent payment-to-income ratio, and they advertise this practice widely. Interestingly enough, home prices were not affected by the expansion of the mortgage market as much as was expected. Right before the fast mortgage market expansion price bubble took place in the Hungarian housing market, the average estimated house prices increased by 200–220 per cent between 1998 and 2001 (see Figure 9.3). This can be explained by changes in macroeconomic conditions (such as household income, income expectation, and interest rates) and the collapse of the
József Hegedüs 191
stock market in 1998. In the years of fast mortgage expansion (2002 and 2003) the price increase was less intensive. Although we can find a positive price-mortgage correlation, it was less significant than was expected by experts (Valkovsky, 2000). The housing finance sector revived after the macroeconomic stabilization, and in almost every country the market-based housing finance system with different institutional frameworks started to develop. By 2007, the share of the outstanding housing loans to GDP had increased above 10 per cent of the GDP in Poland, Czech Republic, Slovakia, and Slovenia without major mortgage subsidies. This fact questioned the rationality of the Hungarian mortgage program: the mortgage market could have probably expanded without huge subsidies, even if the development were slower. The house price increase with flexible mortgage banks led to extreme increase in outstanding loans in the Baltic States, especially Latvia and Estonia, where the loan to GDP ratios surpass 30 per cent. The share of FX loans, which very much increase the vulnerability of the countries, was typically about 40 per cent in 2006, but there were some important exceptions (e.g. Czech Republic). The crisis of 2008 will slow down the development, but financial institutions will continue with mortgage finance in a more prudent way. The Hungarian housing market did not have a price bubble before the 2008 crisis. Therefore we cannot expect a huge decrease of prices due to the supply shock caused by the increasing number of foreclosed flats on the market. When the 2008 worldwide economic crisis hit Hungary it had to take emergency loans from multi-national institutions. The Hungarian currency has fallen about 40 per cent against the EUR, which increased the payments for FX loans sharply, and the mass defaults of FX loans have become a real danger. The number of clients in the Central Adverse Credit Database (KHR)11 increased from 498 000 to 609 000 (by 20 per cent) in 2008. The number of defaults is expected to increase rapidly in 2009. In Hungary the mortgage market changed after September/October 2008. Banks started to tighten the underwriting criteria; some banks stopped issuing loans in underdeveloped areas (where house prices and numbers of transactions are low). Banks typically stopped issuing FX loans from November 2008 onwards. However, only one bank announced that it was not going to issue FX loans, but because of the bank competition, they withdrew the announcement. The conditions of the different types of loans have been changed; HUF-based loans with a mortgage rate subsidy have once more become the cheapest.
192 Housing Market Challenges in Europe and the United States
9.5 Failure of social rental programs As a result of give-away privatization, the percentage of housing stock in the public rental sector fell from 20 per cent in 1989 to 4 per cent in 2003. Households trapped in the public sector were typically the neediest, those who could not afford to pay even the discounted price or who did not consider it worthwhile buying their apartments because they were of such low quality. In cities with more than 50 000 inhabitants the proportion of dwellings owned by local governments is 4–10 per cent. In total, 58 per cent of municipal flats are in cities with populations over 100 000, whereas only 20 per cent are in towns with populations of less than 50 000 inhabitants. Local authorities can freely determine rents, define allocation procedures – whether to use waiting lists or case-by-case tenders for vacant units – and decide on the organizational form of the housing maintenance company. They can choose what type of rental contract to use – fixed or open term, social or cost-rent or market rent. After 2001 they could also freely decide when and whether to sell dwellings. Publicsector rents are very low and they cover approximately 30–40 per cent of actual costs. Moreover, the vast majority of tenants have difficulty paying even these rents, and many are in arrears. According to a housing survey of 2003 (CSO, 2004), 22 per cent of households in the public sector are in arrears, while in the private sector the figure is 5 per cent. Owing to financial disincentives and their own limited resources, local governments have underinvested in the maintenance of their housing stock, which continues to deteriorate. It is estimated that the public housing sector requires HUF 300 billion for renewal and rehabilitation of the stock – 30 times the yearly rent revenue of 2004. Local governments can only allocate vacated and newly-built rental units, which represents only 4–5 per cent of all housing market transactions including buying and renting. The need for social housing is high; according to our need assessment study of 2001, the stock of potential unmet demand for rental housing is around 750 000 units, including about 500 000 social units – approximately 2.5 times the stock at the time (MRI, 2001). The social rental sector has shrunk in most of the transitional countries.12 It has become a residual sector, which concentrates the most vulnerable groups of society. In almost every country in the region, politics (and housing policy) realized the need for social housing after mass privatization and the recovery of the economy at the end of 1990s.
József Hegedüs 193
The housing policy launched in 2000 included a grant program for local authorities supporting five housing areas: the rental sector, energysaving renewal, rehabilitation, land development, and renovation of housing owned by churches.13 The most important element was support for the public rental sector. Local authorities were eligible for a grant of up to 75 per cent of the cost of investments for social rental, cost-based rental (see below), housing for young families and the elderly. Between 2000 and 2004, several hundred local governments took part in the program. The total investment amounted to HUF 60 billion and close to 13 000 units were established. The cost-based option aimed at ensuring long-term cost recovery in the sector, with rent levels higher than existing social rents but lower than market rents. The regulations set the minimum annual rent at 2 per cent of the construction cost. Although this cost rent approach did not guarantee long-term cost recovery, in the first years the actual operational and maintenance cost of the units was considered to be lower than the rent. The cost rent is about 40–60 per cent of the market rent. The high level of interest in the rental sector program is an indication of the commitment of local authorities to solving the housing problem. Before the launch of the program, the Hungarian Government’s Housing Policy Committee was concerned that local authorities would not be able to participate because most of them could not afford the 25 per cent own contribution required, but in fact only 45 per cent of the amount requested by local authorities could be funded. The program had several weaknesses. Average costs were considered to be very high, although one of the most important selection criteria was the average cost per square meter. Allocation criteria for new tenancies were not regulated, and local politics played a role in discretionary allocation. The cost rent was considered to be too high for poor people, but not enough for long-term cost recovery. Moreover, during the period of operation of the scheme, the privatization process continued, thus local authorities privatized 25 000 units but built, bought or renewed only 8800 rental units. In 2004, the government stopped the social rental program because of fiscal pressure, citing the high cost per unit. It proposed a rent allowance program as a substitute for the budget-financed rental program. The proposal was that local governments would enter into long-term contracts with private investors to use newly built rental units for social use. Central and local government would jointly provide a rent subsidy that would bridge the gap between affordable and market rent – cost plus profit. The proposal failed because the guaranteed rent level required by investors was unacceptably high – twice the actual market rent. However,
194 Housing Market Challenges in Europe and the United States
the importance of the social rental sector was never questioned in government documents. In 2005 a new rent allowance program was finally introduced, which aimed to use the private rental sector for social purposes. Local governments could apply for rent allowance for low-income families with children who had private rental contracts. The central government would pay a maximum of 30 per cent of the rent or a28 per month,14 and the local government would also contribute at least as much. The program was a failure: very few local authorities put forward a proposal. One reason for this was that the program required that landlords be registered with the Tax Authority. The majority of private landlords do not pay tax, and they did not change their approach for the sake of participating in this program. A second problem was that the income limit, under about a180 per capita per month, was so low that eligible households were not able to pay the rent. In 2006 another new loan program was launched giving local governments access to subsidized loans from the Hungarian Development Bank for investment in the public rental sector. However, the interest shown by local governments was very limited. Homeownership programs, which have dominated the housing subsidy system (see Figure 9.5) were much more costly for the budget. The real reason for the lack of social housing programs is not only the direct fiscal effect,15 but the lack of long-term interest, as the social and financial sustainability of the new social housing stock has been very weak. The pure operation of the sector needed huge subsidies (to bridge between the difference between market rent and ‘social’ rent), extensive tenure rights remained, allocation principles were not transparent, down-payment requirement as a pre-condition to enter the sector made an indirect regressive selection among possible tenants, and a number of negative elements (e.g. non-payment, deterioration, etc.) emerged. After 2004 modification of the subsidy system, the rehabilitation of housing estates became a successful program, relatively cheap, effective and socially acceptable. The program started in 2000, but picked up speed only after 2004. Under this program, a third of rehabilitation costs were paid by the central government, a third by the local government, and a third by owner-occupiers themselves. Owners also had access to special loans and even to another indirect subsidy through housing saving banks. In the six years from 2000 to 2007, 22 per cent of the stock of this kind was renewed at least partially. The success of the program was related to several important factors. Partly, MPs and local politicians favored it because offering grants for condominiums gave them political advantages. The other factor was the ‘construction lobbies’ (specialized
József Hegedüs 195
in different elements of rehabilitation), which supported the program in order to create a market for their product. Moreover, the contract saving banks developed a product (a kind of bridge loan for saving contracts) that rechanneled subsidies to the typically low-middle income households living in housing estates.
9.6 Housing affordability – safety net as last resort Housing affordability became an important issue after the transition. The ratio of housing expenditure to household income increased from 11 per cent to 21 per cent between 1990 and 2003 while real per capita household income decreased by 30–35 per cent between 1989 and 1997 and did not reach the pre-transition level again until 2000 (CSO, 2004). The housing allowance could not bridge the increasing gap between the housing cost and the income: According to the 1997 household survey, 15.4 per cent of households had real difficulty paying utility costs and rent (HHP, 1998). In Hungary, the housing allowance scheme was introduced in 1993 as part of the new Social Act. Local authorities had to introduce housing allowances for households whose housing expenditure was higher than 35 per cent of household income. The detailed conditions were to be defined by local authorities in their local ordinances: the size of the allowance, maximum household income, and allowable housing consumption – the source of the funding was the budget of the local government. In 2003, the cost of the housing allowance program was HUF 3.5 billion, accounting for 1–2 per cent of the social benefit programs, and the take-up was 5 per cent of households although surveys indicated that some 10–15 per cent of households had difficulty paying their housing costs. Realizing the deficiencies of the local governmentrun housing allowance program in 2004, the government introduced a central housing allowance program (Hegedüs and Teller, 2005). This program was administered by local authorities and 90 per cent of the cost was covered by the central government. As a consequence of the new program, the number of beneficiaries increased from 175 000 households to 350 000, and the cost from 3.5 billion to HUF 17.8 billion. In 2007 the government, having postponed an increase in the gas price for several years, increased it to the world market price. The flat-rate subsidy applied between 2004 and 2006 was replaced by a means-tested benefit program, with different income groups receiving different levels of price subsidy between 15–30 per cent; the highest income group was ineligible. The cost of the program was HUF 110 billion in 2007,16 seven
196 Housing Market Challenges in Europe and the United States Table 9.2
2000 2001 2002 2003 2004 2005 2006 2007
Housing allowance, 2000–2007 Housing allowance outlay (in thousand HUF)
Number of benefit recipients of housing allowance
3 550 882 3 586 817 3 762 148 3 538 595 5 767 570 12 062 534 16 668 368 17 809 754
197 032 183 220 175 055 148 232 196 619 302 893 349 088 357 057
Source: Central Statistical Office.
times more than the housing allowance, and the number of beneficiaries was 2.1 million. The program is managed by the treasury, not through local governments. The existence of a gas price compensation program demonstrates the significance of the effect of public policies (for example, energy price regulation) on safety net programs In spite of the housing allowance and gas price compensation programs, there are huge deficiencies in the safety net. The basic problem of the housing allowance system is that the means-tested support does not make it possible for households to live at a socially accepted standard. Poor households have to employ different survival strategies, e.g. combining different (fragmented) benefit programs, moving to the informal economy, using their housing equity, etc.17 Homeowners’ households facing economic hardship paying housing related costs are forced to use their equity to pay their accumulated debt and move down to a less valuable unit typically located far from urban centers (small ‘dead end’ villages or blind alleys) – a phenomenon which is called social suburbanization. This mobility trend leads to a higher spatial concentration of poverty and segregation of poor households (dominated by Roma families).
9.7 Conclusion: Is there a post-socialist housing regime? This chapter has argued that housing policies in the region have a common point of origin, but the emerging housing policies after the transition reflect the various historical, economic, and political realities of the particular country. The Hungarian housing system (similar to other countries in the region) faced three basic challenges: 1) rebuilding the
József Hegedüs 197
housing finance system, 2) restructuring social housing, and 3) renewing the urban, multi-unit housing stock. The specific attributes of each new housing system can be explained by the combined influences of existing institutions (‘path dependence’), efficiency of ‘knowledge transfer’, and the role of local politics. The ‘new model’ of the emerging housing system depends on the economic, institutional and policy responses to the basic challenges. After the political turn the governments in the region worked under constant fiscal pressure caused by the social and economic costs of the bankrupted socialist economy. As a consequence, the state had to ‘withdraw’ from the housing sector, which resulted in cutting the subsidies for new construction (both public and private), privatization of the building industry and industry for building materials, price liberalization in housing services, privatization of public housing, privatization of the banking sector, etc. Market processes started to dominate the housing sector generating new types of conflicts and problems, e.g. affordability, social segregation, management of multi-unit buildings, hardship in paying housing related service costs (heating, water and sewage, waste, etc.). The housing regime has been changed not only by market forces but by the emergence of some elements of a new social policy (e.g. the means-tested housing allowance, which was a sign of the developing new welfare regime). However, because of the ‘weak state’ (limited public administration capacity to manage safety net programs, inability to tax the society efficiently, lack of reliable information on households’ income and the significance of the informal economy) the efficiency of the income benefit programs was very low. The retaining factors (social forces interested in maintaining the old institutional ways) have not lost their function either, as the story of gas price compensation demonstrates. In housing, which is liable to social conflict, the temptation to maintain the traditional structure (using implicit ‘across the board’ type subsidy through inefficient public companies) was very high. Maintaining the rent level below the actual cost of rent, setting up public management companies, introducing price control of energy costs, and forgiving ‘non-payments’ (eviction evasions) were typical answers given by the state to the social conflicts caused by the transition. The development of the Hungarian housing system and housing policy showed some common elements of the ‘post-socialist’ housing regime based on the global challenges countries in the region faced. One of the common elements of the housing policy in transition countries was the privatization and restitution, which led to a special tenure structure
198 Housing Market Challenges in Europe and the United States
(‘super-homeownership’). However, the specific form and condition of the privatization varied greatly across the countries. However, the size of the public sector in some of the countries (Czech Republic, Poland and Latvia) is substantial even today. At the end of the 1990s, most countries in the region had recovered from the transitional recession and realized that housing was an important and neglected area of public policy. National housing programs were prepared (Slovakia 1999, Hungary 2000; Romania 2000; etc.), which emphasized the role of the rental sector. However, social housing programs, due to fiscal and institutional constraints, had limited results. Another general trend after 2000 is the development of the mortgage finance sector and the attempts to revitalize the social housing sector. The development of housing finance institutions in the region did not follow the same pattern (e.g., mortgage banks in Hungary, commercial banks in Poland, a state-guaranteed housing fund in Slovenia), but market-based housing finance has become a dominant structure everywhere. Housing loans have increased all over the region, but few countries stimulated borrowing with huge subsidies like Hungary. Moreover, the economic crisis of 2008 will have an effect on housing regimes in the region. The housing output will decline, house prices will decrease, housing policy programs will be subordinate to austerity programs, and the possibility of social housing programs will be more limited. The main elements of the new institutional structure are already in place, but the process is unfinished and existing institutions do not form a ‘post socialist housing regime.’
Notes 1. The findings of this chapter are based on research conducted in the EU 7th Framework Program DEMHOW Demographic Change and Housing Wealth (Grant Agreement Number 216865) and partly draw on studies Hegedüs and Somogyi (2005), and Hegedüs and Teller (2005, 2009) also. The author wishes to thank Nóra Teller for her valuable comments and help in preparation of this chapter. 2. Major studies dealing with housing systems in transition countries are Turner et al., 1992; Struyk, 1996; Hegedüs et al., 1996; Buckley and Tsenkova, 2001; Pichler-Milanovich 2001; Lux, 2003; Tsenkova, 2003, Hegedüs and Struyk, 2005). 3. To be unemployed was considered a ‘crime’, which led to a high ‘inside unemployment’ (meaning many jobs were kept in the firms with low salary and almost ‘no work’).
József Hegedüs 199 4. This approach could be conceived as a ‘soft structuralist’ approach, which combines ‘rational choice’ (policy choice or agency choice) with structural elements. In our earlier work we followed this argumentation, for example, in the explanation of ‘self-help’ housing in Hungary (Hegedüs, 1992). 5. The national expenditure on child and family related benefits that reach approx 2.5 per cent of the GDP. 6. Interestingly enough, it was a result of a regulatory mistake, when the construction subsidy for families with three or more children was increased so much that it covered the total cost of the construction in less-developed regions. Small entrepreneurs stepped into the market and organized the mass use of this subsidy scheme, which had a very controversial social impact. 7. This was equal to 187 USD/year in 1997. 8. The Hungarian National Bank report in 2004 (HNB, 2004) evaluated the changes of the mortgage subsidy program in the spring of 2002 as unsustainable. However, this was not indicated in the financial report of 2002 (HNB, 2002), which had a section on the role of housing. 9. After 1998 housing became a hot political issue. The leading political parties got into a ‘game’ promising more and more support to the housing sector without understanding the fiscal and social consequences of the proposed programs. Think tanks often supported by different interest groups are indirectly interested in justifying the programs. (They are typically ‘captured’; the culture of independent think tank policy research is still immature.) Probably this phenomenon is related to the ‘immature’ public policy in these countries. It is partly related to the fact that the government does not have the capacity to predict the possible effects of its policy, but a lack of real interest in monitoring the effectiveness may even be more important. 10. Valkovszky (2000), an analyst of the Hungarian National Bank, argued that the growth of housing credit will increase the vulnerability of the Hungarian economy, even the HUF loans, because of low savings the government deficit has to be financed on the foreign markets. 11. Borrowers who are late with a loan payment, which is higher than the minimal wage (HUF/month 69 000), for more than 90 days. But people who cheated in the underwriting process can be put on a ‘blacklist’ of the Central Adverse Credit Database. They are on the database for five years. 12. It is not clear whether the exceptions (e.g. Czech Republic, Poland, and Russia) represent another model, or as a result of the slower pace of privatization in those countries. 13. Other countries in the region have also launched social housing programs but none seems to have led to a real breakthrough in this area, with the exception of Poland’s TBS (Bondarczuk and Muziol-Weclawowicz, 2008). 14. In 2005, a1 was equal to HUF 248. 15. The average cost for a social rental unit was approximately the same (around HUF 5 million = a20 000) as the present value of the homeownership subsidy between 2002 and 2004. 16. In 2006 the cost of the price subsidies to the energy suppliers was HUF 160 billion. 17. The poorest households easily become victims of illegal usury loans or housing mafia, and are forced get involved in survival crimes.
200 Housing Market Challenges in Europe and the United States
References Barta, J., L. Molnár, R. Petz, K. Pichovszky and L. Skultéty, L. (2003) Public Revenues Generated by Housing Subsidy Programs (Lakásberuházások támogatásából származó államháztartási bevételek), manuscript, GKI Economic Research Institute. Bondarczuk, J. and A. Muzioł-Wecławowicz, (2008) Polish Social Rental Housing. Dilemmas of the Program, paper presented at the workshop on Options for Social Rental Housing in Central and Eastern Europe, Budapest, October. Buckley, R. and S. Tsenkova, (2001) Housing Market Systems in Reforming Socialist Economies: Comparative Indicators of Performance and Policy, European Journal of Housing Policy, 1(2), 257–89 . Cerami, A. (2005) Social Policy in Central and Eastern Europe, The Emergence of a New European Model of Solidarity (Dissertation zur Erlangung des Grades eines Doktors der Sozialwissenschaft der Universität Erfurt), manuscript. CSO (2004) Housing Conditions 2003, Central Statistical Office, 2004. Esping-Anderson, G. (1990) The Three Worlds of Welfare Capitalism, New Jersey: Princeton University Press. Ferge, Z. (2002) Social Structure and Inequalities in Old Socialism and New Capitalism in Hungary, Review of Sociology of the Hungarian Sociological Association 8(4), 9–34. Hegedüs, J. (1992) Self Help Housing in Hungary, in: Beyond Self-Help Housing, K. Matey (ed.), Profil Verlag, pp. 217–31. Hegedüs, J. and E. Somogyi, (2005) Evaluation of the Hungarian Mortgage Program 2000–2004, in: J. Hegedüs, and R.J. Struyk (eds) Housing Finance: New and Old Models in Central Europe, Russia and Kazakhstan, LGI Books, Open Society Institute, 2005, pp. 177–208. Hegedüs, J. and R. Struyk, (2005) Divergences and Convergences in Restructuring Housing Finance in Transition Countries, in: J. Hegedüs, and R.J. Struyk (eds) Housing Finance: New and Old Models in Central Europe, Russia and Kazakhstan LGI Books, Open Society Institute, 2005, pp. 3–41. Hegedüs, J. and N. Teller, (2005) Development of the Housing Allowance Programs in Hungary in the Context of CEE Transitional Countries, European Journal of Housing Policy, 5(2), 187–209. Hegedüs, J. and N. Teller, (2009) Past and Future Development in Public Policies towards Homeownership and the Use of Housing Wealth, The Case of Hungary, manuscript DEMHOW Demographic Change and Housing Wealth (EU 7th Framework Program, Grant Agreement Number 216865) February, 2009. Hegedüs J. and Tosics I. (1996) Disintegration of East-European Housing Model, in: D. Clapham, J. Hegedüs, K. Kintrea, and I. Tosics (eds) Housing Privatization in Eastern Europe, Greenwood, pp. 15–39. Hegedüs, J. and É. Várhegyi (2000) The Crisis in Housing Financing in the 1990s in Hungary, Urban Studies, 37(9), 1610–41. Hegedüs, J., S. Mayo, and I. Tosics, (1996) Transition of the Housing Sector in the East Central European Countries, Review of Urban & Regional Development Studies, 8(2), 101–36. HHP (1998) Hungarian Household Panel Survey 1992–1997 (TÁRKI Social Research Institute).
József Hegedüs 201 HNB (2002) Report on Financial Stability, Hungarian National Bank, December. HNB (2004) Report on Financial Stability, Hungarian National Bank, December. Kasza, G. (2002) The Illusion of Welfare Regimes, Journal of Social Policy 31(2), 271–87. Keszthelyiné Rédei M. and Z. Szabó, (2006) The Level and Distribution of Incomes in the Last One and Half Decades in Hungary (A jövedelmek színvonalának és elosztásának alakulása Magyarországon az elmúlt másfél évtízedben), in: Incomes and Income Distribution, 2004 (Selected studies) Central Statistical Office, Budapest 2006. Kornai, J. (2000) What the Change of System from Socialism to Capitalism Does and Does Not Mean, Journal of Economic Perspectives 14(1), 27–42. Laczkó, M. (2000) The Hidden Economy and Its Effect on the Post-socialist Economy on the Basis of the Household Energy Consumption. An Analysis of the Role of the Hidden Economy in Hungary, Joint Research Program of the Economic Institute of Hungarian Academy of Science and TARKI, 2000 (Egy rázós szektor: a rejtett gazdaság és hatásai a poszt-szocialista országokban a háztartási áramfogyasztásra épül˝ o becslések alapján. Elemzés a rejtett gazdaság magyarországi szerepér˝ ol. MTA KTK és Tárki közös kutatási program.) Lux, M. (2003) State and Local Government: How to Improve the Partnership, in: M. Lux, (ed.) Housing Policy: an End or a New Beginning?, LGI Books, Open Society Institute, 2003. Malpass, P. (2008) Housing and the New Welfare State: Wobbly Pillar or Cornerstone?, Housing Studies 23(1), 1–19. Molnár, L. and K. Pichovsky, (2002) The Financial Cost of the Mortgage Subsidy (manuscript) (A kamattámogatási rendszer költségei és finanszírozhatósága a lakáshitelezésben) GKI Economic Research Institute, October. MRI (2001) Social housing sector (Metropolitan Research Institute) manuscript. MRI (2006) Alternative models of social housing (Metropolitan Research Institute) manuscript. Palocz, É. (2003) The Effect of Mortgage Loans on Household’s Savings, manuscript (A lakáshitelek hatása a lakossági megtakarításokra), Kopint-Datorg. Pichler-Milanovich, N. (2001) Urban Housing Markets in Central and Eastern Europe: Convergence, Divergence or Policy Collapse, European Journal of Housing Policy, 1(2), 145–87. Poggio, T. (2008) The Housing Pillar of the Mediterranean Welfare Regime: Family, State and Market in the Social Production of Home Ownership in Italy, ENHR Working Group Home Ownership and Globalisation Conference: Building on Home Ownership: Housing Policies and Social Strategies OTB Research Institute for Housing, Urban and Mobility Studies Delft University of Technology, the Netherlands, 13–14 November, 2008. Semjén, A., I.J. Tóth, M. Medgyesi, and Á. Czibik (2008) Tax Evasion and Corruption: Population Involvement and Acceptance (Adócsalás és korrupció: lakossági érintettség és elfogadottság), Discussion Papers Institute of Economics, Hungarian Academy of Sciences, Budapest, 2008/13. Stephens, M. and S. Fitzpatrick, (2007) Welfare Regimes, Housing Systems and Homelessness: How Are They Linked?, European Journal of Homelessness, 1(1), 201–12. Struyk, R. (ed.) Economic Restructuring in the Former Soviet Block: The Case of Housing, Urban Institute Press, Washington D.C. 1996.
202 Housing Market Challenges in Europe and the United States Szalai, J. (2005) Poverty and the Traps of Post-communist Welfare Reforms in Hungary: A Fourth World of Welfare Capitalism on the Rise?, paper prepared for the Annual Conference of RC19, ISA, 8–10 September 2005, Northwestern University, Chicago, Illinois, USA. Tausz, K. (2009) From State Socialism to a Hybrid Welfare State: Hungary, in: K. Schubert, S. Hegelich, and U. Bazant, (eds) The Handbook of European Welfare Systems, Routledge, 2009. Tomka, B. (2005) Determinants of East Central European Welfare Systems: A Comparative Perspective, 1945–2000, paper presented at the 20th ICHS Conference, Sydney, 3–9 July 2005. Torgersen, U. (1987) Housing: the Wobbly Pillar under the Welfare State, in: B. Turner, J. Kemeny, and L. Lundqvist, (eds) Between State and Market: Housing in the Post-Industrial Era, Stockholm: Almqvist and Wiksell. Tsenkova, S. (2003) Housing Policy Matters: The Reform Path in Central and Eastern Europe: Policy Convergence? in: S. Tsenkova and S. Lowe (eds) Housing Change in Central and Eastern Europe: Integration or Fragmentation, Aldershot: Ashgate Publishing Limited, pp.193–205. Turner, B., J. Hegedüs, and I. Tosics, (eds) (1992) The Reform of Housing in Eastern Europe and the Soviet Union, Routledge, 1992. Valkovszky, S. (2000) Hungarian Housing Market, Working Paper, Hungarian National Bank 2000/3. Varga, D. (2003) Who is Right in Housing Policy? (Kinek van igaza a lakáspolitikában?) ÉS 48(31).
10 House Price and Other Housing Market Data: A User’s Perspective Anthony Murphy1
10.1 Introduction House prices, housing markets and credit markets are attracting a great deal of attention at the moment because of the credit crunch and large falls in house prices in many countries, including the USA, the UK, Ireland and Spain (Mishkin, 2007; Taylor, 2007). Figure 10.1 shows some of the channels of transmission of the mortgage and housing crisis to consumption, investment, output and unemployment. This chapter discusses the data needs of policy-makers and others who want to analyze and model house prices and the housing market using annual or quarterly data. Good quality house price data are Mortgage, housing, and leverage crisis
Lower demand for housing
Less home construction
Lower capital of financial firms
↓ Home prices and wealth, slower consumption
Credit standards tightened on all loans
↑ Counter-party risk, money and bond mkts hit
Slower GDP growth Figure 10.1 The channels of transmission of the mortgage and housing crisis 203
204 Housing Market Challenges in Europe and the United States
obviously essential for monitoring housing market developments, as a general macroeconomic indicator, for measuring inflation and household wealth, etc. (Arthur, 2006; Fenwick, 2006). A range of mix adjusted and other house price indices are available for some countries, notably the UK and the USA. However, in many other countries house price data are very limited, both in terms of the time period covered and the regions, cities or type of buyers for which they are available. The outline of this chapter is as follows. Section 10.2 discusses the need for long runs of mix adjusted and disaggregated house price data. Unfortunately, the data for many European countries are rather limited. Some of the technical and practical issues involved in constructing mix adjusted house prices series are briefly reviewed in Section 10.3. The UK evidence in Section 10.4 suggests that the main house price indices display the same medium and long-run trends, since they are cointegrated with each other. In other words, real house prices are I(1) and composition or mix bias is either I(0) or trend stationary. The implication is that, for modeling purposes, it does not matter much which house price series you use. Section 10.5 reviews various approaches to modeling house prices, including the inverted housing demand and house price to rent approaches. Based on this review Section 10.6 concludes with a brief summary and wish list of the housing market and other data that are required for modeling European house prices.
10.2 Mix adjusted and disaggregated house price indices Mix adjusted house price data are clearly preferable to average or median house price data since, inter alia, houses are heterogeneous and do not change hands that often (Thwaites and Wood, 2003). The two simple examples in Table 10.1 show that the average houses can be misleading when the composition or mix of houses sold changes over time. Ideally, one would like long runs (25 years plus) of data disaggregated by area or region, type of property and type of buyer (first-time buyer or former owner-occupier). Unfortunately, the data for many European countries are rather limited, as shown in Table 10.2, which is based on the BIS database and the Girouard et al. (2006) and FT European house price datasets. Disaggregated data are extremely useful for many reasons. Regional and other spatial data vary idiosyncratically and are thus more informative about the determinants of house prices than national data so, providing good data exists, regional house price models tend to yield more accurately estimated parameters. A regional house price model is
Anthony Murphy Table 10.1 Year
205
Simple mix adjusted house price examples Type of home
Stock of homes
Sale price
Quantity sold
Average price
Mix adjusted
Example A 0 1
‘Small’ ‘Large’ ‘Small’ ‘Large’
1200 800 1200 800
£80 000 £100 000 £84 000 £105 000
105 95 95 105
% Price
£89 500
£88 000
£95 025
£92 400
6.17%
5.00%
£89 500
£88 000
£95 050
£94 400
6.20%
7.27%
Example B 0 1 % Price
‘Small’ ‘Large’ ‘Small’ ‘Large’
1200 800 1200 800
£80 000 £100 000 £84 000 £110 000
105 95 115 85
Note: In the two examples, houses are either ‘small’ or ‘large’ and the stock of houses is adj used to compile the mix adjusted houses price index, Pt = s=size wst Pts , where wst is the share of ‘small’ / ‘large’ houses in the housing stock and t = 0, 1. In Example A, both types of houses rise in price year-on-year by 5%. In Example B, ‘large’ houses appreciate by 10% while small houses appreciate by 5%. However, fewer ‘large’ houses are sold.
also necessary to address regional issues such as the house price ‘ripple effect’ in the UK, whereby house prices in London tend to lead prices in the South East and, with longer lags, the rest of Britain (Meen, 2001; Cameron et al., 2006). Disaggregated data for first-time buyers are useful since we know that credit conditions are an important determinant of house prices (and consumption, inter alia) and in both the UK and the USA, one may infer changes in credit conditions from changes in the distribution of loan to value (LTV) and loan to income (LTI) ratios, as well as changes in unsecured credit and from the responses to surveys of loan officers (Fernandez-Corregudo and Muellbauer, 2006; Duca et al., 2009). Fernandez-Corregudo and Muellbauer (2006), for example, examine quarterly UK micro data from the Survey of Mortgage Lenders to learn about changes in credit conditions from LTV and LTI ratios of first-time buyers classified by region and age group. They combine data on the proportions of high LTV and high LTI loans with aggregate information on consumer credit and mortgage debt to give ten quarterly credit
206 Housing Market Challenges in Europe and the United States Table 10.2
Start dates of some European house price data
Country
Start date & frequency
Reference
Notes
Austria Belgium Denmark Finland France Germany Greece Ireland Italy Luxembourg Netherlands Norway Poland Portugal Spain Sweden Switzerland UK
2001 (Q) / 1988 (Q) 1971 (A) 1988 (Q) 1992 (Q) 1983 (Q) 1997 (Q) 1994 (A), 2005(Q) 1994 (A) 1971 (Q) / 1997 (Q) 1989 (Q) 1983 (A), 2005 (Q) 1976 (M) 1992 (Q) 2001 (A) 2003 (Q) 1987 (Q) 1975 (A), 1986 (Q) 1970 (A), 1980 (Q) 1968 (Q)
FT FT OECD OECD OECD OECD, FT FT FT FT FT OECD OECD FT FT OECD FT
National / Vienna One-Family-Homes
OECD
Mix Adjusted
Metropolitan Areas, per m2
Average / Mix Adjusted Average, 13 Urban areas
Average per m2
Note: A = Annual, Q = Quarterly and M = Monthly. References: OECD = Table A1, Girouard et al. (2006). FT (Financial Times) = FT European House Prices – Notes and Metadata (online at ft.com).
measures for 1975–2001. These are modeled in a ten-equation system. By controlling for relevant economic and demographic influences on the demand and supply of credit, Fernandez-Corregudo and Muellbauer are able to estimate a single time-varying index of credit conditions that captures the common variation in the ten credit indicators that cannot be explained by the economic and demographic controls. The credit conditions index (CCI) increases in the 1980s, peaking towards the end of the decade. It retraces part of this rise in the early 1990s, before increasing again to a level that exceeds the previous peak (see Figure 10.4 below). The index is useful when modeling both house prices and consumption, and when interpreting monetary conditions.
10.3 Construction of mix adjusted house price indices The construction of mix adjusted house price indices involves many technical and practical choices including mix adjustment procedures;
Anthony Murphy
207
data sources; data cleaning, and weighting and linking. For example, see Diewert (2007) and Wood (2005). (a) Mix adjustment procedures: Are mix adjusted house prices indices based on repeat sales regressions more accurate than those based on stratification, administrative records and/or hedonic regressions? Hedonic regressions are data intensive since we need to know the physical and neighborhood characteristics of each property. Repeat sales regressions are less data intensive but require large datasets. Repeat sales regressions may also overestimate the growth in house prices, since improvements are ignored. The repeat sales regression approach is very common in the USA, where it is the basis of the OFHEO,2 S&P/Case-Shiller and Freddie Mac house price indices. The limited evidence suggests that the exact adjustment procedure does not matter that much. (b) Data sources: Should house price indices be compiled using a sample or all mortgage approvals, completions or registrations at the government land registry? Trade-offs between comprehensiveness and speed of publication can be circumvented. For example, in the UK, the Financial Times house price index is based on a combination of lagging Land Registry data (published with a one- to threemonth lag on all registered transactions) and leading Halifax loan approvals data. (c) Data cleaning: The data used to construct the house price indices should exclude remortgages and further advances, transactions at non-market prices (e.g. sales of council houses to sitting tenants in the UK) as well as obvious data entry errors. Many indices exclude properties bought without a mortgage. A priori, it is unclear whether or not buy-to-let transactions should be included. (d) Weighting and linking: Should the mix adjusted house price results be grossed up to reflect the stock of owner-occupied housing or the flow of transactions in some base year(s) or more recent years? The mix adjusted series should probably be weighted to the owneroccupied housing stock (since this is the relevant quantity when looking at the demand for housing) and chain-linked over time. In practice, the results are often weighted to reflect flows in some base year. The base year weights should, of course, be updated regularly. There is no unique, best mix adjusted house price index.3 However, the UK and USA evidence is that the various house price indices display the same medium and long-run trends, since they are cointegrated with each
208 Housing Market Challenges in Europe and the United States
other. In other words, real house prices are I(1) and composition or mix bias is either I(0) or trend stationary. For modeling purposes, it does not matter much which UK or USA house price series are used. When modeling European housing markets, the important issue is not the choice of exact house price index but the shortage of long run, disaggregated average and/or mix adjusted house prices data, as well as other housing related data. This hinders research.
10.4 Example: UK mix adjusted house price indices Since the British are obsessed with house prices and have experienced a large number of house price booms and busts, it should be no surprise that a large number of mix-adjusted house price indices are compiled for the UK. We will look at some of the main house price indices, all of which take some account of changes in the mix (type of property, number of bedrooms, etc.) and ‘quality’ (central heating, floor area, etc.) of houses that are sold. The four indices we consider are produced by the following bodies: • DCLG (Department of Communities & Local Government) – The
DCLG index was originally constructed using a simple stratification based method of mix adjustment using region and a small number of physical characteristics to form the cells and varying sample of quarterly mortgage completion data from 1968 on. In recent years, a hedonic regression approach has been used (Fenwick and Duff, 2002).4 • Nationwide Building Society. The Nationwide index is constructed using a hedonic regression with many characteristics and mortgage approvals. The index is produced quarterly from 1989 onwards. Average house price data are available back to the 1950s. • Halifax Bank of Scotland. This index is also constructed using a hedonic regression with many characteristics and mortgage approvals. The index is available monthly from 1983 onwards. • Land Registry. The index is based on a repeat sales regression for all property registrations (including cash sales but excluding refinances) in England and Wales. It is seasonally adjusted, and available monthly from 1995 Q2 on. All of the indices are disaggregated by region and the first three are disaggregated by type of buyer. Land Registry house price indices are available for fairly small areas e.g. 3-digit postcode. Figures 10.2 and 10.3 show that average (unadjusted) house prices and all four mix-adjusted house price series for the UK and Greater London
209
Log UK house price (2000 Q1 = 100) 5.6 5.2 4.8 4.4 4.0 3.6 DCLG average NSA DCLG mix adjusted NSA Land registry SA (England and Wales Halifax mix adjusted SA Nationwide mix adjusted SA
3.2 2.8 2.4 1975
1980
1985
1990
1995
2000
2005
Figure 10.2 Various measures of UK house prices Source: Author’s calculations.
Log London house prices (2000 Q1 ⫽ 100) 5.5 5.0 4.5 4.0 3.5
DCLG average NSA DCLG mix adjusted NSA
3.0
Land registry mix adjusted SA Halifax mix adjusted SA Nationwide mix adjusted SA
2.5 2.0 1975
1980
1985
1990
1995
2000
Figure 10.3 Various measures of Greater London house prices Source: Author’s calculations.
2005
210 Housing Market Challenges in Europe and the United States Table 10.3
Various measures of quarterly ln house prices for the UK 1992 Q2 – 2008 Q2
2000:2 – 2008:2
Mean
Std Dev
Mean
Std Dev
UK house prices DCLG Average DCLG Mix Adjusted Land Registry Mix Adjusted Halifax Mix Adjusted Nationwide Mix Adjusted
2.00% 1.90% – 1.70% 1.90%
3.70% 2.40% – 2.20% 2.30%
2.50% 2.10% 2.40% 2.50% 2.40%
4.00% 2.50% 1.70% 2.40% 2.50%
London house prices DCLG Average DCLG Mix Adjusted Land Registry Mix Adjusted Halifax Mix Adjusted Nationwide Mix Adjusted
2.30% 2.20% – 2.00% 2.10%
4.40% 4.00% – 2.80% 3.00%
2.40% 2.00% 2.20% 2.20% 2.10%
4.40% 3.70% 1.60% 2.80% 2.60%
Note: The DCLG average and mix-adjusted data are not seasonally adjusted. The land registry data refer to England and Wales only.
move together over the medium run, so mix adjustment does not matter too much. The only analogy is the Halifax UK house price index in the first half of the 1990s, which may reflect the different regional mortgage market penetration of the Halifax. Of course, the quarter to quarter or short-run variation in the indices is greater as shown in Table 10.3. Nevertheless, the contemporaneous correlations in the quarterly log changes in the seasonally adjusted indices (Halifax, Nationwide and Land Registry) are between 0.8 and 0.9 in the case of the UK and 0.6 and 0.7 for Greater London.
10.5 Modeling house price: What data are required? Apart from house prices, what other data are needed to analyze and / or model the housing market? To some extent this depends on the models you are using: (a) inverted housing demand equations, which condition on the housing stock; (b) house price to rent equations; (c) reduced form equations; (d) ad hoc equations and vector autoregressive equations where, for example, house prices are driven by income and interest rates only.
Anthony Murphy
211
The most basic theory of what determines house prices is just a story of supply and demand, where the supply – the stock of houses – is given in the short run. Then prices depend on the stock of housing and the factors driving demand. In the USA, where rental markets are well-developed and with rents generally market determined in contrast to the heavily regulated rental markets of the UK, the most popular model of house prices is the house price to rent model. The standard supply and demand model of the housing market is summarized in Cameron et al. (2006) and Muellbauer and Murphy (2008).5 House prices are given by the inverted demand curve, i.e. by the existing stock of housing, income, the user cost of housing and other factors driving demand.6 This approach is well grounded in theory. In addition, we have strong priors regarding the values of the key long-run elasticities, corresponding to the ‘central estimates’ set out in Meen (2001) and Meen and Andrew (1998), inter alia. Reduced form models or models relating house prices to rents (Ayuso and Restroy, 2006; Weeken, 2004) are generally less useful in economies such as the UK where the long-run supply of new housing is inelastic and the rental sector is small, regulated and not representative of the private housing sector as a whole.7 The house price to rent approach to modeling house prices assumes that, absent substantial frictions and credit restrictions, arbitrage between owner-occupied and rental housing markets implies the house rent-to-price ratio equals the real user cost of housing uc, defined as the tax-adjusted mortgage rate plus depreciation and taxes minus the expected appreciation of house prices. Inverting and taking logs implies that the log house price-to-rent ratio equals the log of the inverse of real user cost: ln hp rent = − ln (uc) , where the elasticity of the price-to-rent ratio equals -1 and the price-torent ratio is invariant to the housing stock and deviations of income from trend. This model is very simple, attractive and misleading. As Kim (2007) has recently shown, in an equilibrium model with binding, maximum LTV ratios on marginal home buyers, the equilibrium log house price-to-rent ratio is more complicated: ln hp rent = f uc, LTV max , ydev , where ydev is the deviation of actual income from permanent income, and the negative real user cost elasticity can be smaller than 1 in line with empirical results.
212 Housing Market Challenges in Europe and the United States
Ad hoc models, which ignore the supply of housing or condition on the average mortgage advance, a highly endogenous variable, are also potentially misleading. For example, it is difficult to explain the prolonged boom and recent bust in Irish house prices without taking account of the supply of new housing, which was initially restricted by a lack of zoned and serviced land. Returning to the inverted demand house price equation, the basic model is obtained by letting log housing demand be given by ln h = −α ln hp + β ln y + z where hp = real house price, y = real income, and z = other demand shifters. The own-price elasticity of demand is –α and the income elasticity is β. Solving for hp yields ln hp = (β ln y − ln h + z)/α which expresses house prices as a function of the housing stock (quantity) and the other factors shifting demand. Many estimates of the income elasticity of demand suggest that β is in the region of 1, in which case the income and housing-stock terms in the above equation simplify to log income per house, i.e. ln y − ln h. Indeed, Cameron et al. (2006) find that log income per house is a major component of long-run equilibrium house prices in the UK. The demand shifters included in z cover a range of other drivers. Since housing is a durable good (with an investment component), intertemporal considerations imply that expected or ‘permanent’ income and the ‘user cost’ should be important drivers. The user cost takes into account that durable goods deteriorate, but may appreciate in price and incur an interest cost of financing as well as tax. The usual approxima tion is that the real user cost, expressed as a rate, is uc = r +δ+t −hpe hp, where r is the real after-tax interest rate of borrowing, possibly adjusted for risk, δ is the deterioration rate, t is the property tax rate, and hpe hp is the expected real rate of capital appreciation. The ex post user cost can take on negative values as rates of capital appreciation in house-price booms have sometimes exceeded interest and other costs of owning a home. An important practical issue for the modeler is how to measure expected house-price appreciation, since only in recent years have surveys begun to ask households about this. A reduced form approach in which expected appreciation is assumed to
Anthony Murphy
213
be a function of lagged appreciation, interest rates, and of log income, log housing stock is most often used. Then log real house prices are explained by this same set of variables.8 Almost the entire empirical literature on house-price determination agrees that the housing market is not efficient: systematic mispricing can persist. Most empirical models find that lagged house price (capital gains or losses) are an important determinant of current house prices, termed the ‘bubble-builder’ by Abraham and Hendershott (1996). This is consistent with an important extrapolative element in expectations. The deviation of prices from long-run ‘fundamentals’ is then the ‘bubble-burster’.9 For example, a series of positive shocks to fundamentals can lead to rising prices and the expectation of further appreciation leading to greater and greater overvaluation. In due course, the increasing negative pull from fundamentals and additional new house supply reduce the rate of appreciation. When prices eventually fall, the falls can then be exaggerated by expectations of further falls. Many mortgage borrowers face limits on their borrowing and may be risk averse. As a result, nominal as well as real interest rates, demography (such as the proportion of households in the under-35 age group where many first-time buyers are to be found), and proxies for downside risk or mortgage default are important drivers of house prices, at least in the UK (Cameron et al., 2006). The subprime crisis and credit crunch have highlighted the importance of credit conditions for house prices (and savings), particularly in the USA and UK. However, measures of credit supply conditions have often been omitted from empirical house price models.10 Various approaches have been adopted for modeling credit conditions, since they are not directly observable. We discussed the credit conditions index (CCI) estimated by Fernandez-Corugedo and Muellbauer (2006), which is designed to measure shifts in the credit-supply function to UK households, especially since 1980. For the UK, Cameron et al. (2006) estimate a dynamic equilibriumcorrection system of house-price equations for nine regions of Britain using annual house prices for 1972 to 2003. We show that credit-market liberalization, as measured by CCI, had a large direct demand-shift role on house prices, but also altered the relative role of real and nominal interest rates, inter alia, with interesting implications for monetary policy. Given recent international concerns with credit conditions in mortgage markets, our approach provides a conceptual and modeling framework for understanding the effects of such shifts.
214 Housing Market Challenges in Europe and the United States
In Cameron et al. (2006), the long-run equilibrium or solution for ln hpr , the log of real house prices in region r, is: ln hpr = β0r + β1r t + 1.6 1/3 ln yr /hsr + 2/3 ln yGB /hsGB + 1.1 CCI − 0.6 (rmr + sdr) − 22.4 CCI × demeaned rmr − (0.3 + 0.5 CCI) × demeaned ln mr
Apart from region specific constants and trends (β0r and β1r t), the righthand side variables in the long-run solution are log real income per house in region r and nationally, ln yr /hsr and ln yGB /hsGB ; the credit conditions index CCI; the log of the nominal after tax mortgage rate mr interacted with CCI; the real after tax mortgage rate rmr interacted with CCI and the sum of rmr and sdr, the average Stamp Duty (transaction tax) rate. The long-run elasticity of real house prices with respect to real income per house term is 1.6, which is near the central estimates in the literature. Population or the number of households is implicit in our formulation of real income per house since income and the housing stock can both be put on a per capita basis, but the number of people or households just cancels out. The credit condition variable CCI picks up direct credit liberalization effects on house prices. The value of CCI, which is approximately zero up to 1980, climbs to about 0.22 in 1991, dips somewhat after this, and then reaches a maximum of about 0.25 in 2001. See Figure 10.4 which shows the estimated long-run effects of the credit conditions index CCI and interest rates, interacted with CCI, on the level of log real UK house prices. The inclusion of the real interest rate (rmr) in the long-run equilibrium reflects its role in the user cost of housing. When modeling British house prices for periods including the 1970s, the inclusion of some measure of credit availability is necessary in order to obtain a correctly signed real interest rate effect. The nominal interest rate (mr) is included because, as Kearl (1979) argued, when nominal interest rates rise with inflation, leaving real rates unchanged, the real interest burden under standard mortgage contracts is more heavily tilted to or loaded on the first few years. However, as the liberalization of credit markets made refinancing easier, households have been better able to get round these near-term cash-flow constraints. We therefore interacted the log nominal mortgage rate with CCI and found a positive coefficient, so the negative effect of the nominal rate on house prices weakens as CCI rises. The same reasoning also suggests that real
Anthony Murphy
215
Estimated long run effects of changes in the credit conditions index CCI on log real UK house prices 0.3 0.2 Credit conditions CCI (Excluding interaction effects)
0.1 0.0 ⫺0.1 ⫺0.2
Interest rates (Including CCI interactions)
⫺0.3 5
10
15
20
25
30
Figure 10.4 Estimated long-run effects of changes in the credit conditions index Source: Author’s calculations.
rate effects strengthen as CCI rises, strongly confirmed by our empirical result. For the USA, Duca et al. (2009) use time series data on LTV ratios for first-time home-buyers to control for mortgage liberalization. We show that the inclusion of cyclically adjusted credit standards measures in both inverted housing demand and house price to rent equations (Kim, 2007) are essential in order to obtain robust and sensible long-run econometric relationships as well as well-fitting and robust models of short-run changes in real house prices. Abstracting from the short-run dynamics, our estimated long-run rent and house price to rent equations for the USA are: ln (rent) = 0.3 ln hp + 0.2 uc − 0.01 ur ln hp rent = 0.9 ln LTV adj − 0.13 ln uc where hp = real house prices, rent = real rents, LTV adj = the cyclically adjusted LTV ratio for first-time buyers, our proxy for mortgage market liberalization, uc = a backward looking, four-year horizon measure of the user cost of housing and ur = unemployment rate. The estimated speed
216 Housing Market Challenges in Europe and the United States
Log real US house prices actual and simulated (2009 Q1 onwards) 0.9 0.8 – 2008 Q4 value – 0.7 0.6 0.5 Actual Ln real HP Simulated Ln real HP
0.4 0.3 90
92
94
96
98
00
02
04
06
08
10
12
14
16
Figure 10.5 Simulated real US house prices Source: Author’s calculations.
of adjustment to the long run is slow since rents are very sticky. If our model is correct, the US house price bust may last quite a long time. The simulated path of future real US house prices is shown in Figure 10.5, assuming the US economy recovers slowly, mortgage credit conditions revert to their end 1999 value and our model is correct.
10.6 Some conclusions and a wish list The review of house price models in the previous section is useful when specifying a wish list of the data one would like when modeling European house prices, as well as consumption and monetary conditions. Table 10.4 lists these types of data. Unfortunately, when applying the inverted housing demand and house price to rent models to European housing markets, one is immediately confronted by the large gaps in the house price and other housing market time series data. From a user’s perspective, the major issue hindering research is the lack of long runs of consistent time series data. A great deal of resources is being devoted into compiling consistent Euro area mix adjusted house price indices going forward in time. The available evidence suggests that the exact choice of mix adjustment procedure does not matter that much when examining medium-term or
Anthony Murphy Table 10.4
217
Data for modeling aggregate house prices
House price
Mix adjusted house prices by type of dwelling, borrower (first-time buyer and existing owner-occupiers) and region etc.
Rents
House price to rent ratios
Income
Household disposable income and the distribution of individual earnings
Interest rates, mortgage credit, etc.
After tax new and average mortgage interest rates; gross and net mortgage lending by type of borrower; stock of outstanding mortgages; housing equity withdrawal
Affordability measures
House price to income / earnings ratios and after tax repayment to income ratios for first time borrowers; the share of housing costs in income etc.
Credit conditions
The distribution of loan to value and loan to income ratios for first time buyers and existing owner occupiers
Housing supply
New housing supply (starts/completions) by type of dwelling and region; housing stock, by tenure; vacancy rates; construction and land costs
Demography
Population level and change (including net migration); population structure (e.g. the share of the population aged 25 to 34); household formation (headship rates)
Household wealth
Disaggregated household sector wealth = housing wealth + liquid assets + net illiquid assets – mortgage debt – consumer debt
long-term house price trends. From a practitioner’s point of view, what would greatly help research is if some resources were devoted to filling in the gaps in the historic series, possibly using unadjusted, administrative data on house prices. At the macro level, one would also like consistent, medium- to longrun time series data on the housing stock (including vacancy rates) and the main components of household wealth. In addition, the distribution of loan-to-value (LTV) and loan-to-income (LTI) data for first-time and other borrowers would be very useful for tracking changes in credit conditions. At the micro level, one would like survey measures of house
218 Housing Market Challenges in Europe and the United States
price expectations and lending, as well as more information on housing finance and wealth in private household and panel surveys. The desire for more complete historic data should come as no surprise since, from a user’s perspective, the major constraint on research is just the patchiness of the data.
Notes 1. Hertford College, Catte St., Oxford OX1 3BW, UK. E-mail:anthony.murphy @economics.ox.ac.uk 2. Office of Federal Housing Enterprise Oversight (http://www.ofheo.gov/). The OFHEO website provides a comprehensive listing of US house price data sources, as well as a reconciliation of the OFHEO and S&P/Case-Shiller house price indices. For the US as whole, the S&P/Case-Shiller house price index is more volatile than the corresponding OFHEO index and displays a large fall in prices, because the S&P/Case-Shiller index includes more ‘non-conforming’ loans and therefore gives greater weight to areas hit by the subprime crisis and credit crunch. 3. Diewert (2007) summarizing the conclusions of the 2006 Paris OECD–IMF Workshop on Real Estate Price Indices noted the demand for official house price indices that are at least roughly comparable across countries and suggested that the needs of users cannot be met by a single house price index. 4. The rationale for using the hedonic regression approach is that the number of ‘cells’ (that is, possible combinations of different sets of values for the main effects considered) is so large that it is not possible to estimate accurately, or at all, average house prices for each cell individually from the data available. The DCLG hedonic model used includes the following seven main effects: location (local authority district or London borough); local authority cluster; type of neighbourhood; dwelling type; number of habitable rooms (or bedrooms); old/new; type of buyer – first-time buyer or former owneroccupier (FTB/FOO) and three interactions: neighbourhood type × dwelling type; neighbourhood type by FTB/FOO and dwelling type × old/new. 5. The standard model is used by Buckley and Ermisch (1982), Mankiw and Weil (1989), Meen (1996 and 2001), Muellbauer and Murphy (1994, 1997), Muth (1989) and Poterba (1984, 1991) inter alia. 6. Most empirical models of new housing supply are relatively simple – supply depends on current and lagged house prices and construction costs. Land prices are normally excluded from the supply equation, since house and urban land prices are simultaneously determined and urban land price data are often poor quality. The change in the stock of housing equals new supply less demolitions, etc. 7. In addition, the available rent data are poor. Furthermore, even with good data on rents, demand shocks will shift price to rent ratios because rents are far stickier than house prices. 8. Many commentators suggest that this, in a nutshell, is the story of US house prices since 2000. Financial innovations in securitization and changes in procedures by rating agencies resulted in the subprime revolution, extending loans to borrowers with poor credit histories previously denied loans. Many
Anthony Murphy
219
of these loans were for adjustable rate mortgages which particularly benefited from the lowest interest rates for decades in 2001–3. The house-price rises, set in train by these credit-supply and interest-rate changes, fooled many people into thinking that such rises would be sustained. Fundamentals changed in 2003 as interest rates returned to more ‘normal’ levels, and high rates of building expanded the housing stock, while house prices became increasingly overvalued. As the extent of bad loans gradually became clear, the fundamentals changed again, as the supply of credit for all types of mortgages contracted. 9. Of course, as Muellbauer and Murphy (2008) acknowledge, it can be quite difficult to separate temporary from permanent shifts in the fundamental determinants of house prices, including income, interest rates, and credit supply. How does one judge whether the ‘fundamentals’ are at long-run sustainable levels, for example, whether an increase in credit availability is temporary or permanent? In 2005 very few commentators fully appreciated how flimsy were the foundations of the subprime revolution. 10. See Aron et al. (2009) and Muellbauer (2007), for example.
References Abraham, J.M. and P.H. Hendershott (1996) Bubbles in Metropolitan Housing Markets, Journal of Housing Research, 7, 191–207. Aron, J., J. Duca, J. Muellbauer, K. Murata, and A. Murphy, (2009) Credit, Housing Collateral and Consumption in the UK, US and Japan, mimeo. Arthur, S.V. (2006) Residential Property Prices – What Has Been Achieved Since 2003?, paper presented at the OECD–IMF Workshop on Real Estate Price Indexes held in Paris, 6–7 November 2006. Ayuso, J. and F. Restoy, (2006) House Prices and Rents: An Equilibrium Asset Pricing Approach, Journal of Empirical Finance, 13, 371–88. Buckley, R. and J. Ermisch, (1982) Government Policy and House Prices in the United Kingdom: An Econometric Analysis, Oxford Bulletin of Economics and Statistics, 47, 273–304. Cameron, G., J. Muellbauer, and A. Murphy, (2006) Was There a British House Price Bubble? Evidence from a Regional Panel, Discussion Paper 5619, London, Centre for Economic Policy Research. Diewert, E. (2007) The Paris OECD–IMF Workshop on Real Estate Price Indexes: Conclusions and Future Directions, University of British Columbia, Department of Economics discussion paper 07-01. Duca, J.V., J. Muellbauer, and A. Murphy, (2009) Credit Constraints and House Prices: Making Sense of the US Experience, mimeo. Fenwick, D. (2006) Real Estate Prices: the Need for a Strategic Approach to the Development of Statistics to Meet User Needs, paper presented at the OECD/IMF Workshop on Real Estate Price Indexes held in Paris, 6–7 November 2006. Fenwick, D. and H. Duff, (2002) An Improved House Price Index – Update on Developments, Economics Trends, 588, 29–31. Fernandez-Corugedo, E. and J. Muellbauer, (2006) Consumer Credit Conditions in the UK, Bank of England Working Paper 314.
220 Housing Market Challenges in Europe and the United States Girouard, N., M. Kennedy, P. van den Noord, and C. André (2006) Recent House Price Developments: The Role of Fundamentals, OECD Economics Department Working Paper No. 475, OECD, Paris. Kearl, J.R. (1979) Inflation, Mortgages and Housing, Journal of Political Economy, 87, 5(1), 1115–38. Kim, Y. (2007) Rent-Price Ratios and the Earnings Yield on Housing, mimeo, University of Southern California. Mankiw, G. and D. Weil (1989) The Baby Boom, the Baby Bust and the Housing Market, Regional Science and Urban Economics, 19, 325–46. Meen, G. (1996) Ten Propositions in UK Housing Macroeconomics: An Overview of the 1980s and Early 1990s, Urban Studies, 33(3), 425–44. Meen, G. (2001) Modelling Spatial Housing Markets: Theory, Analysis, and Policy. Norwell, Massachusetts: Kluwer Academic Publishers. Meen, G. and Andrew, M. (1998) Modelling Regional House Prices: A Review of the Literature, report prepared for the Department of the Environment, Transport and the Regions, Centre for Spatial and Real Estate Economics, University of Reading. Mishkin, F.S. (2007) Housing and the Monetary Transmission Mechanism, in Housing, Housing Finance, and Monetary Policy, Symposium, sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, 30 August– 1 September, 2007, pp. 359–413. Muellbauer, J. (2007) Housing, Credit and Consumer Expenditure, in: Housing, Housing Finance, and Monetary Policy, Symposium Sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, 30 August–1 September, 2007, 267–334. Muellbauer, J. and A. Murphy (1994) Explaining Regional House Prices in the UK, Department of Economics Working Paper 94/29, University College Dublin. Muellbauer, J. and A. Murphy (1997) Booms and Busts in the UK Housing Market, Economic Journal, 107 (November), 1701–27. Muth, R.F. (1989) Theoretical Issues in Housing Market Research, in: R.F. Muth and J.C. Goodman, The Economics of the Housing Market, Chur: Harwood Academic Publishers. Poterba, J.M. (1984) Tax Subsidies to Owner-Occupied Housing: An Asset Market Approach, Quarterly Journal of Economics, 99, 729–52. Poterba, J.M. (1991) House Price Dynamics: the Role of Tax Policy and Demography, Brookings Papers on Economic Activity, 2, 143–203. Taylor, J. (2007) Housing and Monetary Policy, in: Housing, Housing Finance, and Monetary Policy, symposium sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, 30 August–1 September, 2007, 463–76. Thwaites, G. and R. Wood (2003) The Measurement of House Prices, Bank of England Quarterly Bulletin, Spring 2003, 38–46. Weeken, O. (2004) Asset Pricing and the Housing Market, Bank of England Quarterly Bulletin, Spring, 32–41. Wood, R. (2005) A Comparison of UK Residential House Price Indices, BIS Papers No 21 – Real Estate Indicators and Financial Stability, 212–22.
11 Residential Property Price Statistics for the Euro Area and the European Union1 Martin Eiglsperger
11.1 Introduction This chapter deals with the residential property price indices for the euro area and Member States of the European Union (EU) collected and compiled by the European Central Bank (ECB) in co-operation with the EU national central banks. The raw data stem from various national sources. Since 2001, the ECB has compiled an aggregate for the euro area by weighting together changes in prices for houses and flats for the euro area countries. Since then, the statistical features of both the indicators for the countries and the euro area aggregate have improved, but still remain below the standards of other economic statistics and price indicators for the euro area. Dwelling price statistics for non-euro area EU countries have been collected since early 2005. Section 11.2 outlines the relevance of residential property price indices for the ECB, followed in Section 11.3 by a presentation of the national sources used and the statistical features of the price statistics currently compiled for the euro area. Section 11.4 describes the compilation of an aggregated euro area indicator for residential property prices. A summary, some conclusions and an outlook are presented in Section 11.5.
11.2 The role of residential property prices for ECB analyses The buying or selling of a dwelling is typically the largest transaction a private household enters into. Changes in residential property prices are therefore likely to influence substantially the budget plans and saving decisions of the potential buyers and sellers. Such price changes will also have an impact on the wealth of owners of dwellings given that it is 221
222 Housing Market Challenges in Europe and the United States
the largest asset in their portfolio. Houses and flats are purchased as an investment, which creates returns in the form of rental income in the case the dwelling is being let. In this case, changes in house prices may impact on rents. This is for the time being the main channel through which house price inflation may impact on the rates of change of the Harmonized Index of Consumer Prices (HICP) – the ECB’s measure to define price stability in the euro area. The impact materializes via rents and not directly via house prices, since the HICP covers rental payments by households (2008 weight in the euro area HICP: 6 per cent), but excludes – for the time being – the acquisition of a house or flat for living in it. Also expenditures by homeowners on maintenance or major repairs of owner-occupied dwellings are not yet covered by the HICP. House price developments have an effect on housing wealth, mortgage markets and residential construction investment and can provide important insights for financial stability analysis, since sharp changes in house prices can have a detrimental impact on financial sector health and soundness, by affecting credit quality and the value of collateral. Given these important uses of changes in dwelling prices, price indicators of good statistical quality are required. Since the analysis of the ECB focuses on euro area-wide developments, an aggregated residential property price indicator for the euro area is of particular interest. Owing to the divergent developments of dwelling prices across euro area countries, results for individual euro area countries are also essential. In addition, it would be desirable to obtain at national and at the euro area level a distinction in the price development between: • urban areas, or capital cities, and non-urban areas; • different housing types, most importantly between new and existing
dwellings, and houses and flats. As euro area statistics are compiled from national results, meaningful euro area aggregates call for a sufficient degree of comparability between the national data in statistical terms. In order to be useful in the monitoring of price developments, a quarterly frequency of the results is essential. The degree to which residential property price indicators are able to eliminate the effect of quality differences between different dwellings compared over time is crucial for analyzing price changes. For the time being, the ECB’s and EU central banks’ data collection and compilation of indicators for changes in residential property prices primarily aims at providing insights into changes in transaction prices. Besides that, statistics on the value of the housing stock can also
Martin Eiglsperger 223
contribute an important piece of information, e.g. for analyzing wealth effects. In this context it should be mentioned that different uses of residential property price indices require different concepts.2 Whereas an indicator which is designed for tracking price changes over time has to be adjusted for changes in quality over time, housing appreciation or depreciation of the housing stock as a whole due to quality changes will show up in indicators of housing wealth.3 Other conceptual differences are related to the type of price and weight information to be used for compiling indices.
11.3 ECB dataset for euro area and selected EU countries: Statistical properties 11.3.1 Available data The data set on residential property prices collected and compiled by the EU central banks and the ECB comprises indicators for all EU countries. Figures 11.1 and 11.2 show year-on-year rates of change in the price indicators which have been identified as being most representative of house price inflation in the respective country.4 An overview table of the development of these ‘headline’ residential property prices in euro area countries is shown in the annex. These data have been collected by national central banks and the ECB from various sources. In order to understand and interpret these non-harmonized data it is essential to take into account their very heterogeneous statistical features and quality. Aspects such as frequency, timeliness, coverage and the methods of calculation are presented in the following sections. 11.3.2 Data sources and compilation of price indicators In Bulgaria, the Czech Republic, Estonia, Ireland, Spain, France, Latvia, Luxembourg and Poland data on residential property price are collected by statistical institutes or ministries. The source of official residential property price indicators in Denmark, Lithuania, in the Netherlands, Finland, Slovenia, Sweden and the UK is information gathered for registration or taxation purposes. In Germany, the statistical institute collects prices from the local expert committees for property valuation. The statistical institutes in Spain and France calculate price indicators that make use of information provided by notaries. In Belgium, Germany, Greece, France, Italy, Portugal and Slovakia real estate agencies and associations, research institutes or property consultancies are used as data sources.
224
Belgium
Germany
Ireland
Greece
Spain
France
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
40 30 20 10 0 ⫺10 ⫺20
Italy
Luxembourg
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
40 30 20 10 0 ⫺10 ⫺20
Netherlands
Malta
Austria
Slovakia
Portugal
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
40 30 20 10 0 ⫺10 ⫺20
Finland
Figure 11.1 changes
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
40 30 20 10 0 ⫺10 ⫺20
Residential property prices for euro area countries, annual percentage
Note: Due to confidentiality reasons data for Cyprus and Slovenia are not shown. Source: National data and ECB calculations.
Martin Eiglsperger 225
Bulgaria
Denmark
Czech republic
60 40 20 0
Estonia
2008
2007
2006
2005
2004
2003
2002
2001
2000
⫺20
Hungary
Lithuania
60 40 20
Poland
2008
2007
2006
2005
2004
2003
2002
2001
⫺20
2000
0
United Kingdom
Sweden
60 40 20
2008
2007
2006
2005
2004
2003
2002
2001
2000
0 ⫺20
Figure 11.2 Residential property prices for non-euro area EU countries, annual percentage changes Note: Data for Latvia are not shown due to their high volatility (see annex). Due to confidentiality reasons data for Romania are not shown. Source: National data and ECB calculations.
Data from newspapers or websites are collected for the compilation of residential property price indicators for Malta, Hungary (‘Origo’), Austria (‘Austria Immobilienbörse’) and Romania; the last are merged by the central bank of Romania with expert data from the Chamber of Notaries. The national central bank of Cyprus combines data from the private sector (real estate agencies and commercial banks) and data from the Department of Lands and Surveys. In Ireland and the UK residential property price data are, inter alia, provided by mortgage lenders. The indicator
226 Housing Market Challenges in Europe and the United States
compiled by the UK’s Office for Communities and Local Government is based on a mortgage survey conducted by the Council of Mortgage Lenders. In the majority of the EU countries, residential property price indicators are calculated by national statistical institutes, namely in Belgium, Bulgaria, the Czech Republic, Denmark, Germany, Estonia, Spain, France, Latvia, Malta, Hungary, the Netherlands, Poland, Slovenia, Sweden and Finland.5 National central banks compile dwelling price indicators in Belgium, Germany, Greece, Italy, Cyprus, Luxembourg and Slovakia. In Austria, the national central bank co-operates with the Vienna University of Technology, while the price index compiled by the central bank of Luxembourg is based on the data from the statistical institute. In Ireland, Spain, France and the UK dwelling price indicators are also compiled by ministries. 11.3.3 Frequency Monthly statistics are available for Ireland, the Netherlands, Portugal and the UK. Quarterly data are provided for Belgium, Bulgaria, Denmark, Germany, Estonia, Greece, Spain, France, Lithuania, Malta, Austria, Finland, Slovenia, Slovakia and Sweden. Data from Italy are reported twice per year. For Germany, the Czech Republic, Cyprus, Luxemburg and Poland either only annual information is available or annual data have been chosen because of their much higher statistical quality compared with other data sets that are reported with a higher frequency. Data on changes in dwelling prices published at a lower than quarterly frequency do not satisfy the ECB’s requirements for short-term economic analysis. It is, however, acknowledged, in particular for smaller countries, that higher periodicity of the data implies fewer covered transactions and this may reduce the representativity of higher frequency results and increase their volatility. 11.3.4 Period covered For euro area countries, the original data have – with a few exceptions – a sufficient series length for economic analysis, that is, ten years or more in terms of index levels. For non-euro-area EU member states, the longest time series are available for Denmark, starting in 1971. Indicators from Sweden and the UK both cover more than 20 years. Time series for Hungary, Latvia, Poland and Slovakia start in 2002 or later. It should be noted that long time series covering several decades often require the linking of data sources that may differ in definition
Martin Eiglsperger 227
and coverage. Therefore, international comparisons of non-harmonized national sources have to take into account that the quality of the data is usually lower for periods before the 1990s.6 11.3.5 Timeliness The timeliness of the available data differs considerably. The first published figures are the monthly data from Ireland, the Netherlands, Portugal and the UK, for which data become available between one month and two months after the end of the reporting period. Most other data are provided in a time span of two to eight months after the reporting period. For Luxembourg, the annual price indicator becomes available only about a year and a half after the end of the reporting year. The timeliness of the data can also vary among different indicators for the same country. It should be noted that timeliness should not be considered separately, but in connection with the frequency of the data. 11.3.6 Coverage Most of the data sources refer to certain segments of the housing market. In most cases, the intrinsic characteristics of the data source imply that only a subset of the market is covered, since, for example, the databases are collections of the purchases financed by a certain mortgage lender or cover the transactions of real estate agencies in which they are involved. These samples may not be representative for the whole national residential property market. Other limitations in coverage are due to the sampling design, for example the dwelling type and specification of the objects and the geographical areas for which prices are observed over time. Indicators for Bulgaria, Estonia, Hungary and Romania refer to prices collected in the capital cities or in major cities. Available price indicators for Bulgaria, the Czech Republic, Estonia, Latvia, Hungary, Romania and Slovakia cover existing dwellings, mainly or exclusively flats. Geographical limitations to the capital city are unlikely to be representative of country-wide price development. However, it is acknowledged that it is not straightforward to create representative price indices that also cover very small towns, villages and rural areas. The exclusion of new dwellings may also have a systematic effect on the results. To what extent the limitations to certain quality categories (e.g. price indices for ‘good quality dwellings’) or the restriction to purchased financed by mortgage banks affect the results is not known and may also vary over time.
228 Housing Market Challenges in Europe and the United States
11.3.7 Prices The type of price information required for a residential property price index depends on the main use the index is constructed for. Whereas price dynamics of housing transactions are best measured by collecting transaction prices, valuing the housing stock requires either that the sample of transacted dwellings is representative of the characteristics of the stock – which is sometimes questionable (‘sample selection bias’) – or else the use of price information of non-transacted houses and flats. Appraisals of house values might be used in the latter case. However, care has to be taken in considering whether or not such appraisals might be biased,7 which is more likely if the appraised value is used when a mortgage loan is refinanced than in cases where it is used for taxation purposes.8 The latter situation applies to appraisal-based indices for Denmark and the Netherlands. Other aspects that might imply that appraised values are not fully representative for current house price dynamics are that appraisers might react with a delay to market developments or that changes in appraised values are only reported after actual price changes have reached a certain threshold. Transaction prices are typically available to mortgage lenders, notaries, land registry offices and tax authorities. Other sources are based on price data of different types: In Spain, for example, the valuation of dwellings serves as a basis for calculating property values (‘open-market appraised housing’). For Germany, the data source used by the national central bank comprises typical values quantified by real estate experts who refer to price data of various types, including also non-transaction prices. The Austrian and the Portuguese price index make use of offer prices provided on websites of real estate agencies or consultants; the Austrian indicator uses a broad data set which also includes transaction prices. The index of Malta is an example of an indicator which is based on asking prices advertised in newspapers. Given that the transaction price of houses and flats is usually fixed in a bargaining process, the price changes of offer prices might not always properly reflect the dynamics of transaction prices. In addition, different timing can harm comparability across countries. 11.3.8 Sample composition, quality adjustment, aggregation and weighting The concept of measuring pure price changes over time is usually operationalized by quantifying the price dynamics for goods or services keeping their product characteristics constant (‘matched models’).
Martin Eiglsperger 229
Whenever a product’s quality changes over time, the impact of this change on the price development has to be excluded in order to isolate the pure change in prices. However, it is not possible to follow this conceptual idea for dwelling price statistics for two main reasons: In the first place, no two dwellings are identical and even standardized terraced houses or apartments differ in important details such as location and equipment. Secondly, even if a sample of houses and flats could be designed, so that the physical attributes are almost constant over time, too few of them may be sold in a certain period, thereby making the calculation of reliable high frequency data difficult. Additionally, the location and the size of the land vary across sampled houses. In practice, changes in dwelling prices are often derived from average prices, i.e., the arithmetic or geometric mean or the median of recorded price data. However, such averages do not take into account that the price-determining characteristics of the houses and flats entering the sample at different points in time may differ. Therefore, the meaningfulness of such average prices for measuring house price inflation depends on the way it is controlled for differences in the composition of the sample with respect to the characteristics of the houses or flats. A detailed specification of the dwellings’ physical attributes and the location sets tight limits to changes in the composition and may therefore come close to pure price changes, whereas average prices of broader defined categories may be significantly influenced by non-price factors and their changes. However, the narrower the specification of the categories is, the fewer the transactions per category that can normally be recorded. Since the number of transactions can vary significantly over time, one might face the situation where in a tightly specified market segment no transaction takes place in a certain period. Overall, there is a trade-off between the homogeneity of the sample or sub-sample which would lead to a more accurate and higher frequency of measurement of pure price changes, and the representativity of the sample or sub-sample for the respective housing market. Most of the EU countries have applied the average-price concept in one way or another for the compiling of their indicators. Since the size of the usable floor area is one of the most important physical attributes of a dwelling, averages are typically expressed as price per square meter. Prices are often collected for typical houses and flats usually defined on the basis of very general criteria such as ‘good quality’, ‘good location’ or ‘medium size’. The criteria used for defining the categories and the typical units and their application in practice can vary greatly across indicators. Other
230 Housing Market Challenges in Europe and the United States
indicators refer to certain, more broadly specified dwelling types such as single-family houses, terraced houses or medium-sized flats. Combining quality and location assessments and house types may help to come closer to a pure price index. Quality-adjusted price indices can be compiled by applying hedonic regressions which use the information about the physical attributes of houses and flats and their location. Generally, hedonic regression analysis of house prices usually requires a well-defined set of data about housing characteristics. The national statistical institutes in France and Finland and the Austrian University of Technology calculate such hedonic price indices for residential properties. It should be mentioned that if the same set of dwelling characteristics is applied for specifying strata, weighting together average prices for these strata may produce very similar quality-adjusted results as obtained by calculating hedonic indices. For the Netherlands, the data of the Land Registry Office (‘Kadaster’) were formerly used to compile a so-called ‘repeat sales’-index. By comparing purchase prices for the same dwelling over different points in time, a ‘repeat sales’-index controls for differences in the physical attributes and the location. However, the overall condition of a house or flat might have deteriorated between two sales; the location might have become more attractive, e.g. because of better connection to public transport, and then the ‘repeat sales’-method does not necessarily measure pure price changes. Additionally, since only prices for dwellings are taken into account which are sold more than once, this might not be representative of the whole housing market. If house price values are appraised on a frequent basis, the latter problem could be overcome by covering also appraisal valuations.9 In co-operation with the Delft University of Technology, the national statistical institute of the Netherlands recently developed a new residential property price index which combines the ‘Kadaster’ data and government appraisals.10 Price changes are derived from the change in the average ratio of sales prices and appraisal values in a base period and a comparison period (the so-called SPAR approach). Whereas the sales prices refer to the base period and the comparison period, respectively, the appraised values in both ratios are those of the base period, so that the price change for one dwelling sold twice – in the base period and in the comparison period – would be purely determined by a change in its two sales prices. However, in contrast to the ‘repeat sales’-approach, a SPAR-index is not restricted to dwellings that are sold at least twice, since the set of dwellings for which the average ratio of sales prices and base period’s appraisal values is calculated can
Martin Eiglsperger 231
differ across periods. In order to reflect changes in the prices for the housing stock the SPAR-index for the Netherlands combines these ratios as a weighted arithmetic average.11 Aggregate price indices for residential properties can be created by specifying several distinct categories whose price data are weighted together. If the segments are sufficiently homogeneous, within-stratum prices could be combined by calculating average prices. Regressionbased indices may be preferable in cases in which the dwellings within a stratum are more heterogeneous, which might be unavoidable in a sample design in order to obtain a sufficiently large number of observations for each stratum. When regression approaches are used, aggregate indices can also be obtained directly, that is, by running the regression with the pooled data set. Dummies can be used for distinguishing between categories, which is usually done with respect to the regional dimension (‘location dummies’). While the results of such a pooled regression may differ from aggregate indices obtained by weighting together price data by categories, there is no statistical reason for a general preference of a pooled approach or a weighted-index concept. When aggregating indicators across regions, house types or quality categories, an appropriate weighting structure must be chosen. Since the theoretical foundation in this area is still incomplete, various approaches to weighting are applied in practice. Applying weights based on housing stock data usually implies a high degree of stability, so that the indicator’s variation over time is almost entirely driven by the changes in prices. If reliable information about the housing stock is not available, it is common practice to use population weights as a proxy. The potential volatility of transaction weights can have a significant impact on the index measures, in particular if the weights are frequently updated and chain-linked indices are compiled.12 A strong correlation between price increases and transactions might even result in index values that deviate substantially from expected outcomes, if the weights differ substantially in the course of a house price cycle. In this context, it should be mentioned that, from a practical point of view, the lack of up-to-date information on transaction values will in practice anyway not allow a high frequency of updating the weights used for residential property price indices. For a transaction-weighted index, new dwellings are a much more significant component than in the case of a stock-weighted index. A way to limit the volatility stemming from concurrent transaction-value weights for new dwellings is to apply a more stable weighting structure, which
232 Housing Market Challenges in Europe and the United States
reflects, e.g., the average number of purchases over several years.13 However, this might imply that in periods in which only a few houses are purchased in a certain segment of the market, the price changes might still get a high weight in the overall index resulting from former periods’ high transaction values or volumes. The dwelling price indices for Germany are compiled by using population weights for aggregating average prices for the cities. The indicators for Greece, Italy, and the Netherlands are weighted by using housing stock data. For Spain, price data are weighted together on the basis of the number of valuations, whereas the French national statistical institute aggregates its sub-indices across geographical areas by applying the share of the sales value in the index base period. For most of the other indicators no explicit weighting is applied. 11.3.9 Specification and compilation of headline indicators for euro area countries Since for some of the euro area countries more than one source exists, the ECB specified headline indicators for each of the euro area countries. The following criteria have been applied for selection as the country’s headline indicator to be used for compiling the euro area aggregate: • Coverage: should be broad, if possible, similar with other euro area
countries. • Frequency: subject to satisfactory coverage, the higher frequent source
has been selected. • Timeliness: subject to satisfactory coverage and frequency, the source
has been selected which shows the best timeliness. • Appropriate isolation of pure price changes
11.3.10 General assessment of the available data In general, the data set used by the ECB can give useful information on the trend of the prices for transactions of the residential housing stock in the euro area and EU countries. The data can partially fill a gap in the provision of statistics by official sources. However, the available statistics for each individual country are, though to varying degree, only rough approximations of an accurate and representative statistical measure of actual changes in prices in the housing market. Moreover, most of the available data refer only to a certain segment of the national market, which is not necessarily representative for the national housing market.
Martin Eiglsperger 233
Additionally, the methodological differences between the national indicators are significant with the consequence that differences in price developments between countries do not necessarily reflect actual differences in dwelling price inflation, but may be caused by statistical factors, such as for example the coverage, timing or inadequate quality adjustment. This is a common characteristic of all non-harmonized data and limits its usefulness for cross-country comparison or euro area aggregation. However, it has to be stressed that the statistical differences between national dwelling price statistics are much more significant than differences between most other non-harmonized national data used for economic analysis.
11.4 Compilation of a euro area residential property price indicator An aggregated euro area residential property price indicator is calculated by the ECB as an arithmetic average of the rates of change of the available national price indicators, weighted by shares in the euro area gross domestic product (GDP). GDP data have been used for weighting mainly due to availability and comparability of these data across EU countries. Timely data on transactions or housing stock which would provide more specific information about the structure of the housing market are not available for some countries. In order to get quantitative insights in the impact of alternative weighting schemes, the ECB conducted test calculations, for which existing data gaps were filled by referring to the best approximation.14 The results demonstrated that the general price trend was not affected by the choice of the weights. The headline euro indicator is calculated at annual and semi-annual frequency (see Figure 11.3). Indicators for Germany and Luxembourg which are only available at annual frequency are transformed into semiannual data by interpolating the respective annual series. Currently the ECB does not compile a quarterly estimate of the euro area indicator since, in addition to Germany and Luxembourg, also for Italy data are not available at quarterly frequency. Research planned for 2009 will show whether the application of disaggregation techniques would allow deriving a quarterly euro area time series. The euro area aggregates are calculated only when more than 80 per cent of the euro area country coverage is achieved. In the first step, the euro area aggregate is compiled at semi-annual frequency, using interpolated data for Germany and Luxembourg. Country values not available at the most recent end of the respective series are replaced, if appropriate,
234 Housing Market Challenges in Europe and the United States
Annual series
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
14 12 10 8 6 4 2 0
Semi-annual series
2008 H1
2007 H1
2006 H1
2005 H1
2004 H1
2003 H1
2002 H1
2001 H1
2000 H1
1999 H1
1998 H1
1997 H1
14 12 10 8 6 4 2 0
Figure 11.3 Residential property prices for the euro area at annual and semi-annual frequency, annual percentage changes Source: ECB calculations, based on national data.
by the latest reported rates of change (‘carry-forward’). In the second step, the annual series is converted from the semi-annual euro area aggregate. A long time series for the euro area has been compiled, closing gaps in the back data for the bigger euro area countries by referring to data for which the statistical quality is lower compared to the headline indicator, but sufficient for building a long time series for the euro area aggregate.
11.5 Summary, some conclusions and potential future improvements The European System of Central Banks is collecting a set of residential property price indicators which covers all countries of the European
Martin Eiglsperger 235
Union. While these data generally provide useful insights for analyzing housing market developments in the euro area and their implications on other areas in euro area economies, it has to be acknowledged that the majority of available dwelling price indicators is of limited statistical quality. Since statistical concepts differ across indicators, their comparability may be significantly affected by different types of price data used, differences in the coverage by regions and dwelling types and the way prices are adjusted for compositional differences in samples over time. The limited comparability of residential property price indicators across countries also impacts on the statistical features of the aggregate index for the euro area, compiled by the ECB, while the target criteria applied for the compilation of the euro area index have proven to reach a sufficient amount of robustness in terms of medium-term price trends. Overall, existing shortcomings in statistical terms indicate that the set of residential property price indicators must be improved significantly in several dimensions in order to meet user requirements sufficiently. The two highest priorities for improving EU national and euro area residential property price statistics are, first, work towards a set of quarterly results for all countries and, second, work towards higher quality in terms of more comparable coverage and methods. Quarterly results would also help improve significantly the low timeliness of the current semi-annual euro area estimates (currently up to six or seven months after the end of the reporting period). The ECB and national central banks will continue to jointly develop the common data set on residential property price statistics. Several minor and major improvements have been achieved over recent years. However, the fact that most national central banks are not directly involved in the collection or compilation of residential property price information limits the further improvements that can be achieved. The most promising EU-wide project that can be expected to contribute to the improvement of EU residential property prices is Eurostat’s pilot group on estimating price indices on owner-occupied housing as a potential component of the Harmonized Index of Consumer Prices. In this context, experimental dwelling price indices are to be compiled by all EU national statistical institutes. The ECB has expressed its strong interest in stand-alone residential property price indices, since these data would be an important step towards higher quality and more harmonized EU house price statistics.
236
Annex Table A11
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Overview table of residential property prices in EU countries, annual percentage changes*
Belgium Germany Ireland
Greece
Spain
Italy
France
Luxem- Malta bourg
Netherlands
Austria Portugal Slovakia Finland Euro area
New and existing houses whole country
New and existing houses and flats; whole country
New and existing houses and flats; whole country
New and existing houses and flats; whole country
New and existing houses and flats; whole country
Existing houses and flats; whole country
New and existing houses and flats; whole country
New and existing houses and flats; whole country
Existing houses and flats; whole country
New and existing houses and flats; whole country
New and existing houses and flats; whole country
Existing houses and flats; whole country
Existing houses and flats; whole country
New and existing houses and flats; whole euro area
– – 22.6 22.5 20.5 14.0 6.1 14.3 11.5 7.2 13.4 0.9 –9.4
– – 14.4 8.9 10.6 14.4 13.9 5.4 2.3 10.9 12.2 3.6 –
1.4 2.8 5.8 7.7 8.6 9.9 15.7 17.6 17.4 13.9 10.4 5.8 0.7
0.1 1.9 7.1 8.8 7.9 8.3 11.8 15.2 15.2 12.1 6.6 –
3.6 3.2 –0.4 0.8 3.9 6.0 12.6 7.2 7.0 8.6 5.8 4.9 4.2
4.0 3.6 3.6 6.4 7.3 11.4 10.8 11.5 14.0 11.7 – – –
8.8 8.5 4.9 3.2 8.4 5.0 8.7 13.3 20.3 9.8 3.5 1.1 –2.7
10.8 12.0 10.9 16.3 18.2 11.1 6.4 3.6 4.3 3.9 4.6 4.2 2.9
0.4 0.8 –5.0 –1.9 –1.2 2.2 0.2 0.3 –2.2 5.1 4.0 4.1 –
1.7 3.6 4.5 9.0 7.7 5.4 0.6 1.1 0.6 2.3 2.1 1.3 3.9
– – – – – – – 39.6 15.4 –10.3 16.8 23.9 –
– – – – – –0.5 7.4 6.3 7.3 6.1 7.4 5.9 0.9
2.0 2.2 2.7 4.9 6.0 5.5 6.8 6.5 7.2 7.6 6.5 4.3 –
4.5 3.6 6.7 7.8 7.1 6.2 7.8 7.1 12.0 16.7 11.1 9.2 –
New and existing houses and flats; whole country –1.1 –1.9 –1.7 1.4 0.2 0.2 –1.9 –1.2 –1.5 –1.5 0.2 0.7 0.2
10.1057/9780230246980 - Housing Market Challenges in Europe and the United States, Edited by Philip Arestis, Peter Mooslechner and Karin Wagner
2000 2001 2002 2003 2004 2005 2006 2007 2008
Bulgaria
Czech Republic
Denmark
Estonia
Latvia
Lithuania
Hungary
Poland
Sweden
United Kingdom
Existing flats; large cities
Existing houses and flats; whole country
New and existing houses; whole country
Existing flats; capital city
Existing flats; whole country
New and existing houses and flats; whole country
Existing houses and flats; capital city
New and existing flats; whole country
Existing houses and flats; whole country
New and existing houses and flats; whole country
1.6 34.3 29.5 12.9 27.8 31.0 51.7 10.1 –7.5
– – – 2.7 2.3 20.0 159.3 45.1 –
– – – 10.9 9.1 0.8 –0.8 1.7 –
– – –
11.2 7.9 6.3 6.6 9.3 9.0 12.2 10.4 –
9.8 8.0 21.1 18.9 18.8 5.2 8.5 9.3 –8.4
0.3 1.9 12.2 47.6 36.5 14.7 28.9 24.9
13.5 9.5 13.1 11.4 –0.8 0.6 6.5 21.1 –
6.5 5.8 3.6 3.2 8.9 17.6 21.6 4.6 –
–9.7 23.8 9.5 18.1 9.9 51.7 39.2 33.5 –
–6.1 20.0 3.8 45.3 –
Notes: * Data for Cyprus, Romania and Slovenia are confidential. Source: National central banks and ECB calculations.
237
10.1057/9780230246980 - Housing Market Challenges in Europe and the United States, Edited by Philip Arestis, Peter Mooslechner and Karin Wagner
238 Housing Market Challenges in Europe and the United States
Notes 1. Input and comments by Adrian Page and Daniela Schackis are gratefully acknowledged. The views expressed in this chapter are those of the author and do not necessarily reflect the views of the European Central Bank. 2. See Fenwick (2006, p. 6). 3. It must be noted that keeping the sample representative for the housing stock may require adjusting the composition of the sample for shifts in the dwelling characteristics. 4. Section 11.3.9 presents the criteria used for the identification of the most representative country indicators. 5. The national statistical institute of Estonia calculates a pure average transaction price per square meter for a specific type of flat in Tallinn. 6. In several cases it has not been possible to identify the sources of the national data that are used in published long-term studies on residential property prices. As a result of lacking metadata and the existence of various alternative non-official sources for residential property prices, the estimates used for long-term international comparisons may differ significantly. 7. For details on a potential bias in appraised values see Leventis (2006). 8. For this and the following see de Vries et al. (2007, pp. 45). 9. See, for example, Leventis (2006). 10. For this and the following see de Vries et al. (2007). 11. See de Vries et al. (2007, p. 4). 12. For this and the following see Hoffmann and Lorenz (2006, pp. 5–6 and p. 23). 13. See Muzzicato et al. (2008, p. 7) 14. See Ahnert and Page (2005, p. 299) and European Central Bank (2005, p. 57).
References Ahnert, H. and A. Page (2005) Euro Area Residential Property Prices: the Aggregation of Non-harmonised National Data, in: Bank for International Settlements, Real Estate Indicators and Financial Stability – Proceedings of a Joint Conference Organised by the BIS and the IMF in Washington DC, 27–28 October 2003, Basel, April 2005, pp. 288–307. De Vries, P., G. Mariën, J. de Haan, and E. van der Wal (2007) A House Price Index for the Netherlands: A Review of the SPAR Method, ENHR International Conference, Rotterdam, 25–28 June. European Central Bank (2005) Availability of Key Non-financial Housing Market Indicators, box in the February 2005 issue of the Monthly Bulletin, p. 57. Fenwick, D. (2006) Statistics on Real Estate Prices: the Need for a Strategic Approach, OECD–IMF workshop Real Estate Price Indexes, Paris, 6–7 November. Hoffmann, J. and A. Lorenz (2006) Real Estate Price Indices in Germany: Past, Present and Future, OECD–IMF workshop Real Estate Price Indexes, 6–7 November, Paris. Leventis, A. (2006) Removing Appraisal Bias from a Repeat Transaction House Price Index: a Basic Approach, OECD–IMF workshop Real Estate Price Indexes, Paris, 6–7 November. Muzzicato, S., R. Sabbatini and F. Zollino (2008) Prices of Residential Property in Italy: Constructing a New Indicator, Occasional Paper, Banca d’Italia, August.
12 Housing and Financial Wealth in Austria: What Can Survey Data Tell Us for the Analysis of Financial Stability Issues? Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner
12.1 Introduction Microeconomic data on private households are providing increasingly important information for many economic policy issues. The chapter examines how micro data can be used to improve the analysis of financial stability.1 In recent years, the level of private household debt and debt–income ratios has significantly increased in many countries (see for instance Girouard et al., 2006). This trend can be observed over both the short and the long term. While in many countries the latest upsurge in household debt is primarily a result of rising real estate prices, the long-term trend is unfolding against the backdrop of the broadening range of financing opportunities available to households. We deal with five aspects related to financial stability: How is housing wealth in Austria related to financial stability issues? Do households save adequately? Do investors take too high risks? Is over-indebtedness a threat for financial stability? Does financial literacy help to ensure financial stability? We use data of two Austrian surveys – the 2008 Household Survey on Housing Wealth (HSHW) and the 2004 Survey on Financial Household Wealth (SFHW), both conducted by Oesterreichische Nationalbank. We will base our approach on cluster analyses and on logit estimates. The chapter has the following structure: First, in Section 12.2 we provide a brief overview of the recent literature on survey data on household finance and consumption with a focus on financial stability. In Section 12.3 we describe the relevant available data sets on financial and housing wealth for Austria. Section 12.4 deals with the question of which
239
240 Housing Market Challenges in Europe and the United States
factors drive housing wealth. Section 12.5 focuses on private households’ financial investments and risks they bear. In Section 12.6 we take a closer look on households’ savings and pose the question whether they are adequate. Thereafter, in Section 12.7 we turn to the liability side of households’ portfolios and possible financial stability risks. In Section 12.8 we examine the role of financial literacy concerning financial stability and Section 12.9 concludes.
12.2 Literature overview The availability of micro data for understanding the impact of shocks, policies and institutional changes is particularly important in view of the extremely large heterogeneity in economic behavior of households. The recent financial crisis has demonstrated that a relatively small fraction of private households can have important effects on market outcomes. Housing assets are the most important form of collateral and the value of housing property affects households’ expenditure by improving access to credit for liquidity-constrained households. High-income households tend to have more debt than low-income households; however, the latter are more burdened by their debt. The dynamics of economic aggregates are determined not only by macroeconomic variables, but also by household-specific factors. This is particularly true for household consumption, savings and balance sheets, which are to a large extent driven by expectations about future individual income and demographic and social characteristics. Monitoring further changes in portfolio behavior is particularly relevant for an assessment of the impact of financial innovation. The recent financial turmoil has shown that numerous private households and lenders underestimated the risks associated with high indebtedness, such that they face severe financial distress once the macroeconomic conditions become less favorable. In his literature survey on ‘Household Finance’ Campbell (2006) stresses that a small group of wealthy persons may distort the information value of macroeconomic data if the wealthy exhibit an investment behavior that is different from the average. He concludes that wealthy people are much more willing to take on financial risks and therefore make more investments in stocks. The literature gives a number of different explanations for why households might opt for portfolios that deviate from what is considered an efficient household portfolio. The most prominent explanatory factors are lack of confidence, a low level of investor protection and
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 241
insufficient financial education (Guiso et al., 2005). Benartzi and Thaler (2001) provide empirical evidence for the hypothesis that households’ portfolio decisions are by far less rational and future-oriented than economic theory might suggest. In particular, they underpin their proposition that investors typically pursue a 1/n strategy with regard to their defined contribution plans – i.e., the contributions paid are equally distributed across the n funds covered by the savings plan. This means that the extent of stock holdings not only depends on individual investment behavior, but also on the behavior of the financial service provider. Guiso and Jappelli (2002) review the theoretical literature on optimal portfolio composition. Focusing on five OECD countries (USA, Germany, Italy, UK and the Netherlands), the authors show empirically that investment in risky assets goes up hand in hand with financial wealth. Basically, small investors’ possibilities of diversifying their stock portfolios are rather limited. As their volume of securities is usually relatively small, a diversification of stock holdings is only possible within tight limits. Moreover, for them, entry costs are a higher barrier than for wealthier investors. King and Leape (1987) assume that people’s readiness to hold risky assets also depends on their age, arguing that older people have acquired more know-how on stock prices, dividends and risks in the course of their lives. Surprisingly often, however, empirical data show that the relation between investment decisions and investors’ age is relatively weak (Guiso and Jappelli, 2002). The standard portfolio theory suggests that investors hold a combination of a ‘riskless asset’ and a ‘market’ portfolio, i.e., an optimal composition of risky assets (Gollier, 2002). The share of assets held in the market portfolio depends on the degree of risk aversion. According to this theory, every household should in fact hold risky assets. However, this is not the case. In fact, the share of households that have invested at least parts of their financial assets in risky assets continues to be rather low in many countries. On the basis of data from 150 countries, Honohan (2006) shows that the share of risky assets in private households’ total financial assets is highly concentrated and negligible for all the groups except the 5 per cent most wealthy households. The household debt situation has been studied by several central banks in recent years, including the Banca d’Italia (2008), the Banco de España (2005), the Bank of England (Barnes and Young, 2003; Barwell et al., 2006), the Deutsche Bundesbank (Bartzsch and Stöß, 2007), Suomen Pankki (Herrala, 2006; Herrala and Kauka, 2007), the European Central
242 Housing Market Challenges in Europe and the United States
Bank (ECB, 2005; Rinaldi and Sanchis-Arellano, 2006) and Sveriges Riksbank (Johansson and Persson, 2006). In addition, both the OECD and the BIS have also recently produced their own economic policy reports on this subject (see Debelle, 2004; Girouard et al., 2006). How does the share of households with repayment burdens in excess of the threshold value change in case of income loss due to unemployment or if repayment liabilities increase because of rising interest rates? This type of analysis has recently been carried out by a number of central banks (e.g. Herrala, 2006; Johansson and Persson, 2006; van Rooij, 2002; ˙ Vatne, 2006; Zochowski and Zaja˛czkowski, 2006). The study conducted by Herrala and Kauka (2007) is one of the few that goes beyond the descriptive presentation and modeling of scenarios. Herrala and Kauka project the development in the share of vulnerable borrowers by linking forecast values for determinants influencing the emergence of repayment problems with microdata. Furthermore, a central finding of empirical studies in behavioral economics that analyze financial education is that people may consider a certain behavior generally adequate, but may still postpone any corresponding changes in their own actual behavior (Benartzi and Thaler, 2001).
12.3 Data on housing and financial wealth in Austria A growing number of central banks has recognized the importance of household micro data. Among others, the Federal Reserve Board (Bucks et al., 2006), the Banca d’Italia (Brandolini et al., 2004), and the Banco de España (Bover, 2004) conduct such surveys regularly to collect such data. The OeNB’s Survey on Financial Household Wealth 2004 (SFHW) provides detailed microeconomic data on the financial portfolios of Austrian private households and the survey also covered a number of questions regarding people’s attitude to financial topics and their respective behavior (Beer et al., 2006a, Fessler et al., 2007). The OeNB’s Household Survey on Housing Wealth 2008 (HSHW) concentrates on different forms of housing wealth of Austrian housholds, their mortgage financing and covers a number of questions concerning inheritances (Fessler et al., 2009). 12.3.1 Descriptive statistics on housing wealth 49.9 per cent of the Austrian households own their main residence. Furthermore, 18 per cent of the tenants hold real estate (another flat/house, realities, hotel, premises, etc.). The highest ownership rate (65 per cent) is reached among the group of 60- to 69-year-olds (see Table 12.1).
243 Table 12.1
Ownership rates by socio-economic characeristics
Ownership rates
Ownership rate of main residence
Ownership rate of tenants (further real estate)
Ownership rate of main or further residence
Overall ownership
0.50
0.18
0.59
Ownership rate by age 18–29 30–39 40–49 50–59 60–69 70+
0.16 0.40 0.60 0.59 0.65 0.51
0.06 0.23 0.29 0.25 0.18 0.09
0.21 0.54 0.72 0.69 0.71 0.55
0.45 0.53
0.09 0.20
0.50 0.62
0.41
0.17
0.51
0.56
0.35
0.71
Ownership rate by occupation Free-professions, entrepreneur Salaried employee Civil servant Farmer Blue collar worker Other
0.65 0.44 0.61 0.96 0.41 0.48
0.33 0.23 0.54 1.00 0.12 0.26
0.76 0.57 0.82 1.00 0.48 0.61
Ownership rate by houshold net income* Up to a795 a796–1432 a1433–2388 a2389–3185 a3186–4999 a5000 or more
0.33 0.38 0.51 0.56 0.63 0.61
0.07 0.10 0.18 0.27 0.35 0.47
0.38 0.44 0.60 0.68 0.76 0.80
Ownership rate by marital status Unmarried Married** Divorced/separated Widowed
0.32 0.62 0.36 0.53
0.10 0.27 0.21 0.10
0.39 0.73 0.50 0.58
Ownership rate by number of persons 1 person 2 persons 3 persons
0.35 0.53 0.57
0.13 0.24 0.21
0.44 0.65 0.66
Ownership rate by education Compulsory schooling at most Apprenticeship, vocational or medium-level technical school Academic secondary school or higher-level school Technical college/university
(continued)
244 Housing Market Challenges in Europe and the United States Table 12.1
(continued)
Ownership rates
Ownership rate of main residence
Ownership rate of tenants (further real estate)
Ownership rate of main or further residence
4 persons 5 persons and more
0.63 0.66
0.20 0.26
0.71 0.75
Ownership rate by number of adult persons 1 adult 2 adults 3 adults 4 adults and more
0.35 0.54 0.71 0.84
0.13 0.25 0.14 0.35
0.43 0.65 0.75 0.89
Ownership rate by province Vorarlberg Tyrol Salzburg Upper Austria Carinthia Styria Burgenland Lower Austria Vienna
0.59 0.60 0.57 0.56 0.39 0.62 0.82 0.64 0.19
0.11 0.17 0.10 0.21 0.07 0.12 0.21 0.09 0.27
0.64 0.67 0.61 0.66 0.43 0.66 0.86 0.67 0.41
Notes: * These data were not imputed. Therefore, they are slightly different from those in Fessler et al. (2009). ** Data on partnerships are available but not displayed in the table. Source: OeNB Household Survey on Housing Wealth 2008 (HSHW).
By education, those with university and technical colleague (Fachhochschule) show the highest ownership rate of 56 per cent. This result is followed by the ownership rate of apprenticeship or medium-level technical school graduates (52.5 per cent) who start earning money comparably early. Tenants who own real estate are primarily university graduates (35 per cent). The tenure status varies remarkably with the occupational status of private households. 65 per cent of the entrepreneurs and free-professions own their home. As expected with 96 per cent most farmers are homeowners. Among tenants, 54 per cent of civil servants hold estate. Household housing wealth rises with higher net income. 63 per cent of those in a range of a3186 to 4999 own their main residence. The increase by income is proportionally much higher when looking at ownership rates of tenants with real estate. Above a level of a3186 household net income ownership rates of tenants reach a level higher than 30 per cent.
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 245
Those tenants with a household net income of higher than a5000 show an ownership rate of real estate of ‘just’ 47 per cent (compared to 61 per cent for owner-occupiers). Concerning martial status it is not surprising that married households depict an above average rate of 62 per cent of owner-occupiers, while only around a third of single households (unmarried or divorced) own their home. Married households are primarily those who own further estate (27 per cent). Ownership goes up with the number of persons living in a household together. This factor is not important regarding tenants with real estate holdings. Finally, the geographical place inside Austria is crucial for ownership – while just 19 per cent of Viennese households are owner-occupiers, on average 59 per cent of the households outside Vienna, in the rest of Austria, are homeowners. From tenants holding estate, 67 per cent own other flats or a further house, 33 per cent own a realty and 9.5 per cent have agricultural used land or forests. Only 1.2 per cent own premises or a business office and around 2 per cent have other real estates.2 47 per cent of all households holding further real estate inherited at least one of their further properties. 46 per cent bought it and 15 per cent got is as a gift. A closer look shows us that 50 per cent of the owner-occupier households bought their further flat or house, 37 per cent inherited it and 19 per cent got it as a gift. Looking at the tenants holding a house or flat, 50 per cent have inherited their property but it is interesting that even 40 per cent have bought the flat or house. Concerning the mortgage financing issue 18 per cent of all households holding further real estate have taken up a loan for their further estate, while 57 per cent of the owner-occupier households took a credit for their main home (regardless of whether there are outstanding amounts or not). Further details on households’ mortgage financing can be found in Albacete and Wagner (2009). 12.3.2 Descriptive statistics on financial wealth Austrian private households hold gross financial assets to average a55 0003 (median: a24 000). Austrian households’ gross financial assets are offset by consumer loans averaging a2900, so that average net assets come to a52 000 (median: a22 000). The median values are far lower than the averages, indicating that both gross and net financial assets are highly unevenly distributed. Considered by socioeconomic criteria, the level of financial assets is shown to depend markedly on household net income. Households
246 Housing Market Challenges in Europe and the United States
with a monthly net income of less than a750, for example, have net financial assets of a6621 (median: a3583); the net financial assets of households with incomes in excess of a3000 average a117 779 (median: a53 039). The education level of the household head also accounts for substantial differences in wealth positions. Households headed by a person with compulsory education only dispose of average net financial assets of a19148 (median: a7835). The amount of financial assets rises with the education level of household heads: the households of university and Fachhochschule (technical/medium-level vocational college) graduates own financial assets averaging a93 586 (median: a41 381). Broken down by the occupational status of the household head, households managed by entrepreneurs top the list by a wide margin, with average financial assets coming to a189 778 (median: a38 372). The large gap between the average and the median in this category is noteworthy. Household financial wealth rises from age group to age group; the share of households with negative net financial assets is higher than average among 30- to 39-year-old household heads, as especially many households in this category have taken out consumer loans. A presentation of financial assets across age groups produces a hump-shaped curve, which corresponds to the theoretical expectations about individuals’ asset developments according to the life-cycle model.4 Overall, more than 40 per cent of all Austrian households have taken out loans, 30 per cent of which are for consumption purposes, nearly 60 per cent for housing purposes and over 10 per cent for both purposes. As in the case of financial assets, there is a positive correlation between borrowing and household net income. Depending on the purpose of the loan, however, the relative share of consumer loans is higher among low-income households. Broken down by marital status, the share of household debt is highest among (married) couples. Households whose main residence is owneroccupied (homeowners) have an above-average number of loans and an especially high share of housing loans. The average Austrian household has borrowed some a20 000, with home loans accounting for 86 per cent of the loan volume. The average debt of households that have taken out home loans is roughly a41 562 (median: a18 000). Homeowners have higher average net financial assets than households with rental housing, and even after inclusion of home loans, their average net financial assets are only marginally lower than those of households with rental housing – their median financial assets are in fact considerably higher.
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 247
12.4 Who holds housing wealth and what for? The Household Survey on Housing Wealth 2008 (HSHW) shows that about 50 per cent of Austrian households own their home (57 per cent, according to Statistics Austria). The higher the education level is the higher the probability of owning the main residence (see Table 12.2). Furthermore, martial status and age is highly significant; and the number of persons in the household increases the probability of owning a house. Interestingly, controlling for a set of other characteristics, the data shows no significant influence of the net income level on homeownership. Also, there is no interaction effect between income and age. Homeownership in the western/countryside part of Austria is more widespread than in Vienna; this is clearly underlined by the logit results (being a household from Vorarlberg increases the odds of being homeowner by a factor of around 5.5 relative to the Viennese households). Finally, important to mention is the fact that those households that have already inherited real estate are more likely to own a house by a factor of 2.2. Furthermore, we conducted a cluster analysis on homeownership. Besides the variables on owning the main residence and on holding further real estate, the information on whether the household was ever before a homeowner was included in the dependent variables. Households are grouped into clusters that are statistically the most homogeneous groups in terms of their holdings of real estate. The cluster analysis on the household wealth data leads to four clusters. Cluster 1 households (22.2 per cent of households) are those with relatively high income and holding further housing wealth. Subcluster 1a (12.3 per cent of households) includes those households having further housing wealth besides holding their main residence as owner-occupiers, whereas subcluster 1b (8.8 per cent of households) are tenants holding a further residence but living in a rented flat/house. Cluster 1b comprises many Viennese households. Cluster 2 (12.6 per cent of households) comprises almost all the rest of tenants (besides cluster 4). These households have no further housing wealth. Cluster 3 (30.3 per cent of households) covers owner-occupiers. This cluster comprises academic households, entrepreneurs, married households with several persons living together. Half of the households holding housing wealth (either main residence or other residence) are captured in cluster 3. This cluster has above average shares in nearly all income classes.
Table 12.2
Logit-regression: determinants of homeownership
Logit-regression on homeownership
Education (reference variable: University) No compulsory school leaving certificate (1) Compulsory school (2) Apprenticeship (3) Medium-level technical and vocational school (4) Academic secondary school or higher-level vocational school (5) Marital status (reference variable: widowed) Unmarried (1) Married (2) Divorced/separated (3) Bundesland / province (reference variable: Vienna) Province (Vorarlberg) Province (Tyrol) Province (Salzburg) Province (Upper Austria) Province (Carinthia) Province (Styria) Province (Burgenland) Province (Lower Austria) Interviewer’s assessment of residential area (reference variable: rural area) Urban area, predominantly old buildings (1) Suburban, green area (2) Office center (offices, banks, etc.), few residential flats (3) Industrial area (4) Small city (5) Age Number of persons Monthly net income out of dependent work Age by monthly net income out of dependent work Monthly net income of total household Already inherited Constant Nagelkerkes R2 Cox & Snell R2
Coefficent
Standard error
Odds ratio
–2.42** –1.875*** –0.998*** –0.302
0.947 0.232 0.197 0.227
0.089 0.153 0.369 0.740
–0.507**
0.219
0.602
0.198 0.562*** 0.054
0.249 0.213 0.238
1.218 1.754 1.055
1.704*** 1.561*** 1.209*** 1.323*** 0.426 1.477*** 1.932*** 1.358***
0.316 0.245 0.260 0.215 0.260 0.211 0.380 0.209
5.493 4.765 3.352 3.753 1.532 4.378 6.906 3.887
–1.500***
0.193
0.223
–0.990*** –1.293*
0.135 0.705
0.372 0.274
–0.768 –0.605* 0.038*** 0.233*** 0.000
0.969 0.223 0.006 0.055 0.000
0.464 0.546 1.038 1.262 1.000
0.000
0.000
1.000
0.000
0.000
1.000
0.143 0.509
2.204 0.080
0.790*** –2.522*** 0.393 0.294
Note: *,**,*** denotes significance on the 10%, 5% and 1% level respectively; given standard errors refer to coefficients. Source: OeNB Household Survey on Housing Wealth 2008 (HSHW), authors’ calculations.
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 249
Cluster 4 (34 per cent of households) subsumes young households (77 per cent of households aged between 18 and 29 years are in this cluster), with quite low education. These households have a low income and are primarily unmarried single person households. The households are mainly tenants (71 per cent of the tenants are in this cluster 4). The share of workers is above average.
12.5 Households’ financial investments and risk-taking behavior The socio-economic characteristics of households play a key role in their choice of investment products. The question of which features have the biggest impact on households’ investment strategy can be analyzed using cluster analyses performed directly on the basis of investment strategies or choice models that estimate the probability of holding a certain investment product as a function of specific household characteristics.5 12.5.1 Cluster analysis groups households on their financial holdings The aim of these calculations is to draw conclusions about demographic characteristics on the basis of the financial products which households have chosen to invest in and to identify possible determinants of the investment decision. The cluster analysis produces four primary clusters. Cluster 1 covers ‘traditional’ investors (52 per cent of households). The financial wealth of households in cluster 1a (39 per cent of households) is limited to deposits, building loan contracts and life insurance policies. The net financial assets of the households in cluster 1a average a32 000. The shares of graduates of a medium-level technical and vocational school or an apprenticeship and of workers are very high in this cluster. Households in cluster 1b (13 per cent of households) invest above all in savings products with a higher return (e.g., a capital savings account, premiumaided savings). The net financial assets of the households in cluster 1b are twice as high at a68 141. The households subsumed in cluster 2 are capital market oriented (15 per cent of households). The average share of long-term securities in these households’ financial assets is around 30 per cent. The households in cluster 3 (21 per cent of households) may be defined as those with a minimum of investment products, as all investment products are only represented to a small degree. The level of net financial wealth is below the median in nearly 80 per cent of the households.
250 Housing Market Challenges in Europe and the United States
10 per cent of the households have negative net financial wealth. The high share of retired people in this cluster is striking – many of the households in the cluster are headed by older people and widows/widowers. Moreover, the share of self-employed persons is very high in this cluster. The households in cluster 4 (12 per cent of households) have a low level of assets, but endeavor to diversify their investment. Therefore, in relative terms, their investment in long-term securities is high. 12.5.2 Logit estimates on the decision of holding risky assets In a further step we calculated a logit model to estimate the probability of holding shares and/or mutual fund shares. Analyzing the model results helps to identify the determinants for holding shares and/or mutual fund shares and assessing their relative importance. In addition, the model controls whether or not variations in capital market participation associated with factors such as income, education level and financial wealth are correlated with each other and/or other control variables. The probability of a household owning shares and/or mutual fund shares rises significantly with its gross financial wealth, income and education level. The risk aversion dummy – capturing those respondents who disagree with the statement ‘When I invest, high profit is more important to me than high security’ – is significantly negative in the model, as expected. Gross financial assets have by far the highest explanatory power, though. A joint assessment of all three age variables (age, squared age and cubed age) highlights the influence of age on the probability of holding shares and/or mutual fund shares. We can show that this probability declines until the age of around 40, remains relatively constant after that and drops again sharply at retirement age. This means that, ceteris paribus, the probability of holding stocks is higher for relatively low-aged households. Our results are consistent with those reported by Bertaut and Haliassos (1995), who estimate a similar logit model for the USA and also find that financial wealth, income and education level have a positive influence on capital market participation. Their data set is superior in that it allows the authors to directly include unemployment risk in the model whereas, in our case, it was only possible to use the income and education dummy variables as imperfect proxies. In Austria, people’s awareness of stocks has increased over the last few years, which was probably attributable to the privatization of state-owned enterprises and the advertising activities of various public and private agents. Around 47 per cent of respondents agreed to the statement ‘It is easier to build financial wealth with stocks than
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 251
with traditional savings books/types of savings,’ which illustrates that fact. In particular, wealthy investors who in general have more possibilities (based on their income, job security, health status, etc.) make comparably higher investments in stocks. International comparison shows that Austria has a low capital market participation rate and that in Austria risky assets account for just a small share in the total financial assets of those who hold risky assets. While the percentage of those holding shares and/or mutual fund shares in Austria is around 22 per cent, it is approximately 32 per cent in the UK, around 34 per cent in the Netherlands and as high as around 49 per cent in the USA.6 The participation rates in Italy and Germany (around 19 per cent in both countries) are comparable to those in Austria. Differences across countries are primarily attributable to the different pension systems. The lowest wealth deciles in most countries are characterized by similarly low participation rates; significant differences can be observed only in the upper deciles, owing in part to institutional reasons and different financial systems (market-based versus bank-based). On average, households that invest in risky assets invest slightly below one-third of their financial wealth in such assets. While the total amount of households’ financial assets has a strong impact on the participation rate, it hardly changes the share of risky assets in total financial assets. The share of total financial assets invested in risky assets7 is no higher among wealthy households than less wealthy households (see Figure 12.1), whereas the absolute amounts invested in risky assets increase markedly as households’ financial assets go up. The phenomenon that the share of total financial assets invested in risky assets does not increase – or increases only marginally – from one wealth decile to the next can also be observed in most of the other analyzed countries (Guiso and Paiella, 2001). The share of risky assets in total financial assets is not a suitable indicator of households’ risk aversion. Less wealthy households, which – in terms of financial wealth, income, job security, health, etc. – typically face higher risks and dangers than wealthier households, mainly hold safe, low-return assets so as to minimize risk at least in their investment portfolios. This does not imply that they are generally more risk-averse than wealthy households. While wealthier households also hold safe assets, their willingness to complement their portfolios with risky assets increases in line with their financial assets. Assuming the same level of risk aversion and assuming that the risk of unemployment or illness decreases as financial wealth increases, wealthier households should opt for riskier assets than less wealthy households.
252 Housing Market Challenges in Europe and the United States Share in %
Share in %
80
60
70
50
60 40
50 40
30
30
20
20 10
10
0
0 Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10 Share of households owning stocks and/or mutual fund shares and related 95% confidence interval based on bootstrap with 2000 replications (left-hand scale) Average share of stocks and/or mutual fund shares in aggregate gross financial wealth of households and related 95% confidence interval based on bootstrap with 2000 replications; only those households with stocks and mutual fund shares (right-hand scale)
Figure 12.1 Stock and/or mutual fund share holdings by gross financial wealth decile Source: OeNB Survey on Financial Household Wealth 2004 (SFHW).
However, the collected data do not confirm this assumption, indicating that the share of risky assets tends to be disproportionately low in the portfolios of wealthy households. This hypothesis is confirmed by the fact that even households in the top decile of gross financial wealth invest in building loan or life insurance contracts.
12.6 Do private households save adequately? Taking a closer look at households’ ability to accumulate income and to put savings aside is crucial in assessing indicators for households’ vulnerability. A breakdown by socio-economic characteristics gives additional insight into which household groups are more or less affected by economic shocks – important information when discussing financial stability issues. More than half of the respondents report that they save regularly or make deposits under a savings plan; 44 per cent save at irregular intervals or put aside whatever income is left at the end of the month. 5 per cent of households are unable to save. The higher households’ income and financial wealth are, the more they save on a regular basis. 20 per cent of households with monthly net incomes of below a750 state that they are unable to save; 12 per cent have no savings.
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 253
Households’ saving capacity varies strongly among clusters. As is to be expected, the affluent households have the greatest saving capacity; more than 16 per cent of these households state that they are able to save more than a10 000 a year. Only 3–6 per cent of the households in the other clusters fall into this category. Households in cluster 3 have the lowest saving capacity. The savings behavior of cluster 3 households also reflects their low saving capacity. Some 70 per cent of households save at irregular intervals or cannot put aside any funds. In the other clusters, the share of regular savers comes to over 50 per cent. Around 43 per cent of Austrians put savings aside on a regular basis, a share which increases with both income and age and furthermore with education. Older respondents save on a more regular basis than their younger counterparts. Significant shares of households that are unable to set aside funds are only found in the low income group (net equivalent income under a749) and in the group of under 30-year-olds (12 per cent). The share of households stating that they were putting money aside ‘for a rainy day’ turned out to be roughly 85 per cent. The tendency to save is more frequent among members of groups with higher age and income, and is found to be weakly correlated with education.
12.7 Household debt – a financial stability risk? One aspect of economic policy concern is that highly indebted households may ultimately be unable to service their loans. In terms of economic policy, there are three different levels of concern: • the macro level: macroeconomic risks resulting from a slump in
consumer demand; • the financial sector level: the risk of financial instability resulting from
households being unable to service their debt; and • the individual level: risk of household over indebtedness.
According to the OeNB Survey on Households’ Financial Wealth, 39 per cent of Austrian households had taken out a loan. A growth in household indebtedness is not a problem per se, but it does become an issue when the borrowers’ housing or consumer loan payment obligations exceed their financial means. Examined by the purpose of the loan, housing loans predominate and account for some 85 per cent of the total volume of loans, and about 70 per cent of the households with debt have taken out a housing loan. The highest level of household debt was found among households
254 Housing Market Challenges in Europe and the United States
headed by persons aged 30 to 39, the reason for indebtedness being the purchase of consumer durables and investment in owner-occupied housing. Consequently, most of the households with negative net wealth belong to this category. An international comparison of debt frequency (Brandolini et al., 2006) shows that, compared to Austria, the proportion of households with loan commitments is lower only in Germany (30 per cent) and Italy (22 per cent). The figure for Germany may be underestimated, however, as the national data used for the Luxembourg Wealth Study (LWS) only reflect loans to the extent that they exceed a threshold of a2500. At 80 per cent, the loan frequency is highest in Norway, followed by the US (75 per cent) and Sweden (70 per cent). Possible reasons for Austria’s relatively low ranking are the tax treatment of loan interest rates, the low loan-to-value ratio for housing loans and a specific aversion to taking out loans. Loans are more prevalent among high-income households, and this difference in distribution is even more pronounced for housing loans than for consumer loans. The differences in the levels of debt broken down by income group are repeated on an international scale and are evident in all of the countries examined by Girouard et al. (2006). There are also considerable differences across income quartiles in foreign currency loans because this type of borrowing is mostly used to finance new homes, i.e., for investments that high-income households are more likely to be able to afford. Consumer loans are most commonly taken out by households with a low level of financial assets, whereas the frequency of housing loans initially increases in line with financial wealth and only declines in the most affluent households. In Austria the ratio of debt to gross financial assets is at 36 per cent, somewhat higher than the euro area average (33 per cent), and the accumulation of financial assets is comparatively low. Debt-to-wealth ratios are a lot more uniform within the euro area than the ratios of debt to disposable income, which may reflect, among other reasons, the fact that in countries where bullet loans are widely common (e.g. the Netherlands), part of people’s savings are earmarked for loan repayment at the end of the term, whereas in Austria, bullet maturities are only customary for foreign currency loans. In other countries, for example in Italy, the steady accumulation of financial assets can partly be ascribed to the fact that banks applied very restrictive lending policies over long periods of time. The average size of housing loans far exceeds that of consumer loans. Both consumer and housing loans evidently increase in size as the levels
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 255
of income and gross financial assets rise. One problem in interpreting this spread arises because neither the original loan amount nor the time it was taken out is known. This problem is especially relevant if the research is focused on establishing how the size of household debt is influenced by socio-economic characteristics, whereas the size of debt at the time of the survey is relevant in addressing the implications of household debt for financial stability. The relatively high share of housing loans in the total debt volume is an indicator of a low financial stability risk, as housing loans are generally secured by the underlying property. One housing loan risk factor is the high proportion of foreign currency loans, which presents a quite considerable exchange rate risk to the liabilities of households.
12.8 Is financial literacy a precondition of financial stability? Against the background of the current financial and economic crisis the importance of private households’ financial knowledge and skills to deal with everyday financial matters and to make the right choices for their needs was underlined. Micro data show that people’s attitudes about money often depend on their education, income and age. A higher level of education, income and age tends to go hand in hand with a more positive financial attitude. 12.8.1 Cluster analysis on people’s money management capability The cluster analysis of households’ money management capability relies on responses to the question how well respondents keep track of their finances, whether they prefer low-risk saving instruments and whether they tend to overdraw their checking accounts. The related responses yielded four clusters: two groups of respondents with generally positive bank account balances, one cluster with basically balanced accounts and a fourth cluster – containing 15 per cent of all respondents – with those people whose accounts are usually overdrawn. The households in this last cluster have the lowest financial wealth, the least diversified investment portfolios (primarily passbook savings accounts; 61 per cent hold life insurance policies) and the largest overall volume of consumer loans. The cluster analysis confirms that the capability to ensure not to overdraw one’s accounts and to keep track of one’s finances is correlated with age and not with education or income. The indicators on which this analysis was based were answers to the questions
256 Housing Market Challenges in Europe and the United States
whether respondents put aside money for emergencies, whether they have made individual retirement provisions and whether they would like to be better informed about financial issues. The analysis yielded four clusters. The first two clusters are characterized by the fact that about 93 per cent of respondents have made private old-age provisions, have a high income and strongly diversified portfolios, which makes the detailed analysis of these two portfolios particularly interesting. Moreover, the respondents in these two clusters on average have a relatively high level of education. The second cluster also includes respondents which have made considerable equity investments. The third cluster (35 per cent of households) encompasses households without private pension schemes (about one-quarter of respondents). The respondents in this cluster have invested lower amounts in all investment categories than respondents in the first two clusters. The third cluster contains households with high income and wealth groups together with respondents who preferably take advice from their family and seek out information provided by various advisors. Hence, cluster analysis confirmed once more that respondents with a higher level of education and income tend to shop around, not least because wealthier respondents are in a stronger negotiation position. In order to classify households based on how they make financial decisions, we used questions on the ownership of different investment instruments (passbook savings account, bonds, stocks, etc.) and the question of whether respondents compare offers provided by different banks. Households in the high income cluster have clearly more diversified portfolios. An indicator for responsible financial behavior is whether respondents rely on more than one information source in financial matters. More than half of the households answered that they primary rely on their home banks’ advice (see Figure 12.2). In the second cluster, almost onethird of respondents rely on advice by their own bank. Of all clusters in this area, this cluster has the highest share of over-70-year-olds. Cluster analysis thus confirms that loyalty concerning one’s bank is particularly strong among older people.
12.9 Conclusions Housing wealth and financial wealth are the major parts in private households’ wealth holdings; and housing wealth is most often used as
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 257 I talk to my financial advisor at my bank
59% 27%
I talk to my family I gather information at different banks
22% 20%
I talk to friends I rely on information from newspapers I browse the internet for information I get brochures at my bank I talk to independent financial advisors I talk to my colleagues at work I consult with the staff council at my workplace
19% 18% 18% 11% 8% 3% 0%
10%
20%
30%
40%
50%
60%
70%
Figure 12.2 Which sources do you rely on when you seek information on financial issues?* Note: * Multiple answers were possible. Source: OeNB Survey on Financial Household Wealth 2004 (SFHW).
collateral. The chapter tries to illustrate how micro data can help when analyzing several topics including financial stability issues. Logit estimates of both datasets (2008 Household Survey on Housing Wealth (HSHW) and the 2004 Survey on Financial Household Wealth (SFHW)) show that education, age (and, for financial wealth, also net income) are decisive factors for holding these wealth categories. It can be shown that housing wealth in Austria is mainly for living and not for transaction purposes. Given that only few Austrian households actually hold stocks and that these holdings are concentrated among wealthy households, however, the related risks to financial stability are comparably low. The participation in capital market increases in line with financial wealth, net income and the level of education. Risk-taking is strongly correlated with wealth. The participation rate is highest in the highest wealth decile, in the group with the highest net income, the group with the highest level of education, and in the group of entrepreneurs. Our analysis shows that high-income households tend to have more debt than low-income households. The risks associated with private debt that could threaten financial stability in Austria are therefore rather small. Because of the low levels of debt compared with other countries and owing to the fact that borrowing is concentrated in affluent
258 Housing Market Challenges in Europe and the United States
households, the level of household indebtedness in Austria does not pose a threat to financial stability. Future efforts directed toward improving financial literacy should pay special attention to the relationship between financial attitudes and actual financial behavior. Personal attitudes toward financial issues (e.g., risk orientation, propensity to invest in complex financial products and to shop around, self-confidence in financial matters) are often linked with the level of education, income or age. A striking result of the survey is that half of the respondents seldom compare financial services offers. As shopping around allows consumers to get better deals, encouraging consumers to do so would improve their basis for sound financial decision-making. We have shown a very heterogeneous picture of factors determining financial stability. We conclude by underlining that household wealth data is of crucial importance for analyzing financial stability.
Annex Gross financial assets = Current account holdings8 + savings deposits including deposits made under building loan contracts + value of bonds + value of stocks quoted on the stock exchange + value of mutual fund shares (equity funds, bond funds, mixed funds, real estate funds, hedge funds, money market funds) + value of equity investments + accumulated payment of life insurance premiums. Net financial assets are defined as gross financial assets excluding consumer loans. In the OeNB Survey on Financial Household Wealth 2004 the interview partner was the household head or the household member with the most accurate knowledge about the respective household’s finances, whereas in the OeNB Household Survey on Housing Wealth 2008 questions were put to the owner of the property or the person named in the tenant contract. Cluster analysis To cross-check clustering results, we used Ward’s method to establish a hierarchy among respondents and the K-means procedure to
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 259
partition respondents. First, the number of clusters was determined with Ward’s hierarchical clustering method; then this number was confirmed by applying the K-means algorithm.
Notes 1. Financial stability in our understanding is broadly defined in terms of its ability to facilitate economic processes, manage risks, and absorb shocks. 2. These percentage shares do not add up to 100% as multiple answers are possible (e.g., part of the home was inherited and part of it was bought). 3. A comparison of these data with financial accounts data shows that gross financial assets as recorded in the survey represent about 50 per cent of the financial assets recorded in the financial accounts. The degree of consistency differs among investment categories. 4. The cross-sectional data from a (static) age distribution must not be interpreted as dynamic across the life cycle at a specific survey date. 5. The methods used for the cluster analysis are described in the Annex. 6. For reasons of comparability, bonds are not included in this calculation (data are from the 1990s). 7. The precision of the estimated shares weighted for Austria increases in the higher deciles owing to higher participation rates; in the lower deciles, it is very low. Overall, the participation rate increases across financial wealth deciles; there is no significant difference in the share of risky assets in total financial assets for those households that hold risky assets. 8. Only data sets that could be evaluated fully were used in this analysis. A value of zero is used where no assets were reported. However, only holdings were counted, as respondents were not asked to provide exact amounts of overdrafts on current accounts. The survey did not cover cash holdings. After all, whether to include cash in assets is a matter of debate (transaction balances, loss of value and the like).
References Albacete, N. and K. Wagner (2009) How Do Private Households Finance Their Home Equity? In: Monetary Policy & the Economy Q3/09, OeNB, forthcoming. Banca d’Italia (2008) The Bibliography of the SHIW, available at: www.bancaditalia.it/statistiche/indcamp/bilfait/biblio/biblio_ing_2.pdf Banco de España (2005) The Wealth of Spanish Households: A Microeconomic Comparison, Economic Bulletin, July, available at: www.bde.es/informes/be/ boleco/2005/be0507e.pdf Banks, J. and S. Tanner (2002) Household Portfolios in the United Kingdom, in: Guiso, L., M. Haliassos and T. Jappelli (eds), Household Portfolios, MIT Press, pp. 219–51. Barnes, S. and G. Young (2003) The Rise in US Household Debt: Assessing its Causes and Sustainability, Bank of England Working Paper 206, available at: www.bankofengland.co.uk/publications/workingpapers/wp206.pdf
260 Housing Market Challenges in Europe and the United States Bartzsch, N. and E. Stöß (2007) Measuring German Household Debt: Financial Accounts Data and Disaggregated Survey Data as Complementary Statistics, IFC Bulletin 26, Irving Fisher Committee on Central Bank Statistics. Barwell, R., O. May and S. Pezzini (2006) The Distribution of Assets, Income and Liabilities across UK Households: Results from the 2005 NMG research survey, Bank of England Quarterly Bulletin, Spring. Beer, C., P. Mooslechner, M. Schürz and K. Wagner (2006a) Austrian Households’ Financial Wealth: An Analysis Based on Microeconomic Data, in: Monetary Policy & the Economy Q2/06, OeNB, pp. 94–110. Beer, C., P. Mooslechner, M. Schürz and K. Wagner (2006b) Financial Wealth Holdings of Austrian Private Households – Data Sources and Institutional Features, paper presented at the LWS Conference held on 14–15 December, 2006. Beer, C. and M. Schürz (2007) Characteristics of Household Debt in Austria, Does Household Debt Pose a Threat to Financial Stability? In: Monetary Policy & the Economy Q2/07, OeNB, pp. 58–79. Benartzi, S. and R. Thaler (2001) Naive Diversification Strategies in Defined Contribution Saving Plans, American Economic Review 91, 79–98. Bertaut C. and M. Haliassos (1995) Why Do so Few Hold Stocks?, Economic Journal 105(432), September, 1110–29. Bertaut, C. and M. Starr-McCluer (2002) Household Portfolios in the United States, in: Guiso, L., M. Haliassos and T. Jappelli (eds), Household Portfolios, MIT Press, pp. 181–219. Bover, O. (2004) The Spanish Survey of Household Finances (EFF): Description and Methods of the 2002 Wave, Banco de España, Occasional Paper 0409. Brandolini, A., L. Cannari, G. D’Alessio and I. Faiella (2004) Household Wealth Distribution in Italy in the 1990s, Bank of Italy, Economic Research Department, Economic Working Papers 530. Brandolini, A., E. Sierminska and T.M. Smeeding (2006) Comparing Wealth Distribution Across Rich Countries: First Results from the Luxembourg Wealth Study. Working Paper 1, Luxembourg Wealth Study. Brown, S. and K. Taylor (2005) Household Debt and Financial Assets: Evidence from Great Britain, Germany and the United States, Working Paper 05/2005, Universität Leicester. Bucks, B., A. Kennickell and K. Moore. (2006) Recent Changes in US Family Finances: Evidence from the 2001 and 2004 Survey of Consumer Finances, Federal Reserve Bulletin 92 (February), A1–A38. Campbell, J.Y. (2006) Household Finance, Journal of Finance LXI(4), 1553–1604. Case, K.E., J.M. Quigley and R.J. Shiller (2005) Comparing Wealth Effects: The Stock Market versus the Housing Market, Advances in Macroeconomics, Volume 5, Issue 1. Coricelli, F., F. Mucci and D. Revoltella (2006) Household Credit in the New Europe: Lending Boom or Sustainable Growth, Centre for Economic Policy Research, Discussion Paper 5520, available at: www.cepr.org/pubs/newdps/dplist.asp?dpno=5520&action.x=8& action.y=6 Craig, A. (2006) Managing Debt is Important as Managing Assets, TD Economics Special Report, TD Bank Financial Group, May, available at: www.td.com/economics/special/ca0515_debt.pdf Cronqvist, H. and R.H. Thaler (2004) Design Choices in Privatized Social-Security Systems: Learning from the Swedish Experience, American Economic Review 94(2), May, 424–28.
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 261 Crook, J. (2006) The Demand and Supply of Household Debt: A Cross Country Comparison, in: G. Bertola, R. Disney and C. Grant (eds), The Economics of Consumer Credit: European Experience and Lessons from the US, MIT Press. Crook, J. and S. Hochguertel (2006) Household Debt and Credit Constraints: Comparative Micro Evidence from Four OECD Countries, Working Paper 05. University of Edinburgh. available at: http://www.iue.it/FinConsEU/Research Activities/DevelopingCreditJune2006/JCrook_2006.pdf Cucurucu, S., J. Heaton, D. Lucas and D. Moore (2004) Heterogeneity and Portfolio Choice: Theory and Evidence. Prepared for the Handbook of Financial Econometrics, North-Holland. Debelle, G. (2004) Macroeconomic Implications of Rising Household Debt, BIS Working Paper 153, available at: www.bis.org/publ/work153.htm Del-Río, A. and G. Young (2005) The Determinant of Unsecured Borrowing: Evidence from the British Household Panel Survey, Bank of England Working Paper 263, available at: www.bankofengland.co.uk/publications/workingpapers/ wp263.pdf Doms, M. and M. Motika (2006) Property Debt Burdens. FRBSF Economic Letter 2006-18, available at: www.frbsf.org/publications/economics/letter/2006/ el2006-18.html Duygan, B. and C. Grant (2006) Household Debt and Arrears: What Role Do Institutions Play? Third Annual DG ECFIN Research Conference. 7–8 September, available at: www.iue.it/FinConsEU/ResearchActivities/DevelopingCredit June2006/Duygan&Grant_Jun%2706.pdf Dynan, K., K. Johnson and K. Pence (2003) Recent Changes to a Measure of US Household Debt Service, Federal Reserve Bulletin, October, available at: www.federalreserve.gov/pubs/bulletin/2003/1003lead.pdf ECB (2005) Financial Stability Review, December, available at: http://www.ecb.int/ pub/pdf/other/financialstabilityreview200512en.pdf Fano, D. (2005) Assessing Risk in Household Portfolios: Progress and Open Issues at the Macro and Micro/Survey Level. Luxembourg Wealth Study, S.A.DI.BA. Perugia, January. Faruqui, U. (2007) Are There Significant Disparities in Debt Burdens Across Canadian Households? IFC Bulletin 26, Irving Fisher Committee on Central Bank Statistics. Fessler, P. and M. Schürz (2008) Stockholdings in Austria, in: Monetary Policy & the Economy, Q2/08, OeNB, pp. 83–100. Fessler, P., M. Schürz, K. Wagner and B. Weber (2007) Financial Capability of Austrian Households, in: Monetary Policy & the Economy Q3/07, OeNB, pp. 50–67. Fessler, P., P. Mooslechner, M. Schürz and K. Wagner (2009) Housing Wealth of Austrian Households, in: Monetary Policy & the Economy Q2/09, OeNB. Girouard, N., M. Kennedy and C. André (2006) Has the Rise in Debt Made Households More Vulnerable? OECD Economics Department Working Paper 535, available at: http://www.olis.oecd.org/olis/2006doc.nsf/linkto/ecowkp(2006)63 Gollier, C. (2002) What Does Theory Have to Say About Household Portfolios?, in: L. Guiso, M. Haliassos and T. Jappelli (eds), 2001, Household Portfolios, MIT Press, pp. 27–55. Guiso, L. and M. Paiella (2001) Risk Aversion, Wealth and Background Risk, CEPR Discussion Paper 2728.
262 Housing Market Challenges in Europe and the United States Guiso, L. and T. Jappelli (2002) Household Portfolios in Italy, in: L. Guiso, M. Haliassos and T. Jappelli (eds), Household Portfolios, MIT Press, pp. 251–91. Guiso, L., M. Haliassos and T. Jappelli (2002) Household Portfolios, MIT Press. Guiso, L., M. Haliassos and T. Jappelli (2003) Household Stockholding in Europe: Where Do We Stand, and Where Do We Go? CEPR Discussion Paper 3694. Guiso, L., L. Sapienza and L. Zingales (2005) Trusting the Stock Market. CRSP Working Paper 602. Haliassos, M. (2006) Household Portfolios, in: Palgrave Dictionary, 2nd edition. Herrala, R. (2006) Household Indebtedness, Bank of Finland Bulletin 2006-1, available at: http://www.suomenpankki.fi/en/julkaisut/bulletin/2006/b_1_2006.htm Herrala, R. and K. Kauka (2007) Household Loan Loss Risk in Finland – Estimations and Simulations with Micro Data, Bank of Finland Discussion Paper 5. Honohan, P. (2006) Household Financial Assets in the Process of Development, in: World Bank Policy Research Working Paper No. 3965. Johansson, M.W. and M. Persson (2006) Swedish Households’ Indebtedness and Ability to Pay – A Household Level Study, Economic Review 3, 24–41. King, Mervyn A. and Jonathan I. Leape (1987) Asset Accumulation, Information, and the Life Cycle, in: NBER Working Paper No. 2392, September. Klyuev, V. and P. Mills (2006) Is Housing Wealth an ‘ATM’? The Relationship Between Household Wealth, Home Equity Withdrawal, and Saving Rates, IMF Working Paper 162, available at: www.imf.org/external/pubs/cat/longres. cfm?sk=19396 Kohler, M., E. Connelly and K. Smith (2004) The Composition and Distribution of Household Assets and Liabilities: Evidence from the 2002 HILDA Survey. Reserve Bank of Australia Bulletin. April, available at: www.rba.gov.au/PublicationsAnd Research/Bulletin/bu_apr04/bu_0404_1.pdf Magri, S. (2002) Italian Households’ Debt: Determinants of Demand and Supply, Temi di discussione del Servizio Studi 454, Banca d’Italia. May, O., M. Tudela and G. Young (2004) British Household Indebtedness and Financial Stress: A Household-Level Picture, Bank of England Quarterly Bulletin, available at: www.bankofengland.co.uk/publications/quarterlybulletin/ qb040401.pdf May, O. and M. Tudela (2005) When Is Mortgage Indebtedness a Financial Burden to British Households? A Dynamic Probit Approach, Bank of England Working Paper 277, available at: www.bankofengland.co.uk/publications/workingpapers/ wp277.pdf Rinaldi, L. and A. Sanchis-Arellano (2006) Household Debt Sustainability. What Explains Household Non-Performing Loans? An Empirical Analysis. ECB Working Paper 570, available at: www.ecb.int/pub/pdf/scpwps/ecbwp570.pdf Riiser, M.D. and B.H. Vatne (2006) Developments in Household Debt. An Analysis of Microdata for the Period 1986–2003, in: Norges Bank Economic Bulletin 2/06. Tudela, M. and G. Young (2005) The Determinants of Household Debt and Balance Sheets in the United Kingdom. Bank of England Working Paper 266, available at: www.bankofengland.co.uk/publications/workingpapers/wp266.pdf Van Els, P., W. van den End and M. van Rooij (2003) Financial Behaviour of Dutch Households: Analysis of the DNB Household Survey 2003, DNB Research Report 744, available at: www.dnb.nl/dnb/home/file/744_tcm13-38874.pdf
Pirmin Fessler, Peter Mooslechner, Martin Schürz, Karin Wagner 263 Van Rooij, M. (2002) Financial Vulnerability of Mortgagors and Mortgagees. Research Memorandum WO 706.E, De Nederlandsche Bank. Vatne, B.H. (2006) Developments in Household Debt, An Analysis of Microdata for the Period 1986–2003, IFC Bulletin 2/2006, Irving Fisher Committee on Central Bank Statistics. ˙ Zochowski, D., Zaja˛czkowski, S. (2006) The Distribution and Dispersion of Debt Burden Ratios Among Households in Poland and its Implications for Financial Stability, MPRA Paper 692, University Library of Munich, Germany.
Index capital gains anti-speculation 167 house prices 45, 135 imputed rent 113 taxation 72, 117, 118, 163, 166, 167, 173 central banks asset prices 3, 15, 31–2, 34, 36 interest rates 3 securitization 54 subprime mortgage market 47–8 Collateralized Debt Obligations (CDOs), subprime mortgage market 2–3, 41, 46–7 commodities, bubble 3, 40–1, 48–9, 53 consumption equity withdrawal (HEW) 6, 105, 109 housing wealth 6, 19–20, 22–3 interest rates 113 marginal propensity to consume (MPC) 19, 22, 116 CPI-inflation 3, 40, 48, 53 Credit Default Swaps (CDSs), risk 3, 41 credit markets credit rating institutions 47 credit scoring 30 crisis see financial crisis liberalization 19–20
André, Christophe xi, 6–7, 109–30 Arestis, Philip xi, 1–14, 40–59 Asian–Russian crisis (1997–8) 3, 40, 51 asset backed securities financial crisis 49 financial engineering 57 forms 3, 41 residential mortgage-backed securities (RMBS) 23, 29–30, 49, 66–8, 117, 120, 125 securities and derivatives 41 see also securitization asset prices borrowing 31 bubbles 6, 30, 31, 32, 33, 48, 120–4 central banks 3, 15, 31–2, 34, 36 correction 31 cycles 31 inflation 3 interest rates 113–14 monetary policy 6, 31, 32, 34–5, 120–4 shock 32, 113 stability 32 Austria data 242–6 descriptive statistics 242–6 financial investments 249–52 homeownership rates 243–4, 248 housing wealth 13–14, 239–63 savings 252–3
data house prices see house price statistics housing investment 21 housing markets 10–11, 203–20 rental housing 91 stability issues 239–63
Bank of Japan, liquidity easing 41 Bear Stearns 48, 125 Bernanke, Ben 3, 40 business cycles house price volatility 20–2 housing investment 20–2 housing markets 6–7, 109–30
Eastern Europe housing regimes 178–80 see also Hungary 264
Index 265 Eiglsperger, Martin xi, 221–38 equity withdrawal (HEW) consumption 6, 105, 109 credit standards 30 mortgage markets 22, 23, 158, 159 OECD countries 115–17 European Union (EU) fiscal policy 8–9, 158–77 house prices 12–13, 62, 221–38 housing system 7, 131–57 rental housing 133–6 Fannie Mae 48, 49, 56, 60, 67, 68, 125 Fessler, Pirmin xi, 13–14, 239–63 finance, housing see financial crisis financial crisis contagion 47, 49 current economic policies 50–2 housing bubbles 40, 61–2 lessons 53–6 liberalization 2, 42–4 monetary policy 3, 50–1 preceding observations 61–2 quantitative effects 52–3 subprime mortgage market 2–3, 40–59 United States 49–50 financial innovation assets see asset backed securities excessive lending 6 financial structured products 2–3 future income 13 liberalization 2, 40, 41, 44–50 monetary policy 24, 35, 40 portfolio behaviour 14 subprime mortgage market 2, 6 financial intermediaries indirect holdings 14 securitization 29–30, 66 financial liberalization see liberalization financial stability asset prices 32 automatic stabilizers 167 financial literacy 255–6 household debt 253–5
prices 50 shock 158 survey data 239–63 fiscal policy financial crisis 50 housing markets 8–9, 158–77 structural fiscal measures 9 see also taxation Freddy Mac 48, 49, 60, 68, 125, 307 Girouard, Nathalie xi, 6–7, 109–30 Greenspan, Alan 3, 33, 40, 51 Gstach, Dieter xii, 5–6, 85–108 Hegedüs, Jószef xii, 9–10, 178–202 homeownership rates academic literature 63–6 Austria 243–4, 248 determinants 3–4, 60–84, 247–8 developments 16–18, 63–6 house prices 64 housing bubbles 4 housing finance 60–84 interest rates 63–4 sociological/demographic factors 64 tax incentives 4, 16, 26, 64, 65, 69, 72–8, 118 variation 16–18 house price statistics mix adjusted indices 206–10 mix adjusted/disaggregated indices 204–6 assessment of data 232–3 available data 223 compilation of indicators 223–6, 232–4 coverage 226–7 data set 223–33 data sources 223–6 European Union (EU) 12–13, 221–38 frequency 226 market data 10–11, 203–20 modeling 210–16 prices 228, 236–7 samples 228–32 timeliness 227
266 Index house price volatility business cycles 20–2 economic implications 1 structural reduction 9, 167–8 taxation 119 variation 16–18 wealth effects 18–20 house prices capital gains 45, 135 European Central Bank (ECB) analyses 12, 221–38 European Union (EU) 12–13, 62, 221–38 homeownership rates 64 mortgage markets 22–4 rents compared 112 role 221–3 shock 61, 86, 115, 119, 126, 213 Spain 144–5 housing bubbles financial crisis 40, 61–2 homeownership rates 4 monetary policy 6–7 housing expenses, taxation 2, 6, 8–9, 72–3 housing finance competition 2, 17, 28 contract saving systems 66 globalization 1 homeownership rates 60–84 household wealth 2 housing markets 26–30 international dimension 29–30 liberalization 1, 4, 28 market-based systems 2 modes 66–8 mortgage see mortgages recommendations 64–5 role of state 66–72 securitized see securitization subprime see subprime mortgage market vulnerabilities 61–2 housing investment business cycles 20–2 data 21 housing markets business cycles 6–7, 109–30 challenges 1–14
data 10–11, 203–20 economic policies 6–7, 109–30 fine-tuning 168–70 fiscal instruments 8–9, 159–74 historical perspective 110–13 housing finance 26–30 imputed rent 25, 113 indicators 17 international dimension 29–30 liquidity extraction 6 macroeconomics 16–24 monetary policy 2, 30–5, 113–16 OECD countries 109–30 regulation 25–6 relevant issues 15–39 role of state 68–72 subsidies 169 taxation 166–70 transaction costs 118 housing sector diversity 132–6 economic development 1 housing wealth academic literature 240–2 Austria 13–14, 239–63 cluster analysis 249–50 collateral value 20 financial investments 249–52 indebtedness 26–9, 253–4 money management 255–6 private consumption 6, 19–20, 22–3 structural factors 117–18 taxation 118–20 withdrawal see equity withdrawal Hungary housing affordability 195–6 housing policy 184–6 housing program 186–91 new housing system 9–10, 178–202 post-socialism 196–8 recession 180–4 rental housing 192–5 social sector 10, 192–5 imputed rent capital gains 113 CPI 91, 104 housing markets 25, 113
Index 267 indexation 90–1 taxation 8, 24–5, 72, 78, 91, 160–2, 171 indebtedness housing wealth 26–9, 253–4 private households 62 interest, mortgage relief/tax deduction 2, 9, 24–6, 152, 160, 162–3, 167, 172, 187 interest rates asset prices 113–14 central banks 3 consumption 113 homeownership rates 63–4 LIBOR 46 short-term nominal rates 28 United States 48 Karakitsos, Elias xii, 40–59 liberalization credit markets 19–20 debt crises 43 development policy 43 efficient markets 42 financial crisis 2, 42–4 financial innovation 2, 40, 41, 44–50 housing finance 1, 4, 28 rental housing 7–8, 131–57 sequencing 43 theoretical debate 137–41 life cycle model 19, 114, 246 liquidity easing 41 macroeconomics housing markets 16–24 New Consensus 42 rental housing 85–108 marginal propensity to consume (MCP) 19 microeconomic data 239 monetary policy asset prices 6, 31, 32, 34–5, 120–4 changing financing structures 33–5 conflicting approaches 32–3 efficiency 6–7
financial crisis 3, 50–1 housing markets 2, 30–5, 113–16 shock 2, 22 transmission 113–16 Mooslechner, Peter xii, 1–14, 15–39, 239–63 mortgage interest, tax deduction 2, 9, 24–6, 152, 160, 162–3, 172, 187 mortgage markets characteristics 22–3 house prices 22–4 variety 23 mortgages down payments 17 foreign currency 29 indebtedness 26–9 liberalization 4 loan to income (LTI) ratios 11, 205, 217 loan to value (LTV) ratios 6, 11, 17, 23, 26, 27, 30, 73, 117, 119, 205, 211, 215, 217 long-term fixed rate mortgages 2, 28 product variety 23 subprime see subprime mortgage market variable rates 28 Murphy, Anthony xii, 10–11, 203–20 ownership, rates see homeownership rates Pareja-Eastaway, Montserrat xii, 7–8, 131–57 policies development 43 fiscal see fiscal policy Hungary 184–6 monetary see monetary policy prices assets see asset prices commodities 3, 40–1, 48–9, 53 housing see house prices stability maintained 50 volatility see house price volatility provisioning 7, 124, 125, 127
268 Index regulation banking 124 housing markets 25–6 rental housing 7–8, 131–57 securitization 126 theoretical debate 137–41 rental housing academic literature 86–91 contracts 8, 140, 144–5, 148–9, 152 data issues 91 European Union (EU) 133–6 Hungary 192–5 macroeconomics 85–108 market share 96–7 markets 5–6, 18, 20, 85–108 OECD countries 118 price-to-rental ratio 95, 121 regulation/liberalization 7–8, 131–57 rent control 137–41, 147–50 rent-to-income ratio 95–6, 102–4, 121 rental rates 92–6 renters/owners 98–102 social sector 7, 18, 136, 192–5 Spain 7–8, 141–54 stylized facts 98–104 vacancy rates 97–8 see also tenure rents house prices compared 112 imputed see imputed rent risk Credit Default Swaps (CDSs) 3, 41 financial investments 249–52 financial system supervision 124–6 logit analysis 250–2 subprime mortgage market 23 Sánchez-Martínez, María Teresa xiii, 7–8, 131–57 Schürz, Martin xiii, 13–14, 239–63 securitization central banks 54 credit standards 30 financial crisis 35–6
financial engineering 57 financial intermediaries 29–30, 66 international dimension 2, 29–30 regulation 126 residential mortgage-backed securities (RMBS) 23, 29–30, 49, 66–8, 117, 120, 125 subprime mortgage market 23, 30, 45 usage 66, 67 see also asset backed securities shock absorbers 150, 168 asset prices 32, 113 financial stability 158 house prices 61, 86, 115, 119, 126, 213 impact 13, 240, 252 monetary policy 2, 22 supply 191 social sector Hungary 10, 192–5 rental housing 7, 18, 136, 192–5 Spain balanced system 150–2 contracts 8, 144–5, 148–9, 152 economic growth 141 house prices 144–5 housing policies 143–7, 150–2 regulated/unregulated dwellings 146–7 rent control 147–50 rental housing 7–8, 141–54 subsidies 145–6, 151 supply/demand 145–6, 150 tenure 143–7, 150–2 transition 145–6 special investment vehicles (SIVs) 44–7, 54 Springler, Elisabeth xiii, 3–4, 60–84 stock markets indirect participation 14 wealth 19 subprime mortgage market central banks 47–8 Collateralized Debt Obligations (CDOs) 2–3, 41, 46–7 default risk 23
Index 269 financial crisis 2–3, 40–59 financial innovation 2, 6 refinancing 45 securitization 23, 30, 45 subsidies housing decisions 8 housing markets 169 interest rates 9, 10 national differences 2, 24 second homes 5–6 Spain 145–6, 151 see also taxation taxation automatic stabilizers 167 capital gains 72, 117, 118, 163, 166, 167, 173 categories 161 deduction 2, 9, 24–6, 152, 160, 162–3, 172, 187 housing expenses 2, 6, 8–9, 72–3 housing markets 166–70 housing wealth 118–20 imputed rent 8, 24–5, 72, 78, 91, 160–2, 171 indirect 174 regulatory change 25–6 stamp duties 167, 169 tax incentives 4, 16, 26, 64, 65, 69, 72–8, 118
value added tax 163–6 see also fiscal policy tenancies, structures 16 tenure choice 17, 85, 91, 95 duration 92 European housing systems 131–6 residual income 100 Spain 143–7, 150–2
7,
United Kingdom, mix adjusted house price indices 208–10 United States bail outs 48 financial crisis 49–50 interest rates 48 loan portfolios 44 long-term fixed rate mortgages 2 value added tax
163–6
Wagner, Karin xiii, 1–14, 15–39, 60–84, 239–63 Washington Consensus 42 wealth house price volatility 18–20 houses see housing wealth stock markets 19 Wolswijk, Guido xiii, 8–9, 158–77