The State at Work: Public Sector Employment in Ten Western Countries

The State at Work, Volume 1

The State at Work, Volume 1 Public Sector Employment in Ten Western Countries

Edited by

Hans-Ulrich Derlien University of Bamberg, Germany

B. Guy Peters University of Pittsburgh, USA

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Hans-Ulrich Derlien and B. Guy Peters 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library

Library of Congress Control Number: 2008929287 European universities in transition : issues, models and cases / edited by Carmelo Mazza, Paolo Quattrone, Angelo Riccaboni. p. cm. Includes bibliographical references and index. 1. Education, Higher—Europe. 2. Educational change—Europe. I. Mazza, Carmelo, 1967– II. Quattrone, Paolo, 1968– III. Riccaboni, Angelo, 1959–. LA628.E974 378.4—dc22

2008 2007050570

ISBN 978 1 84376 516 5 SET ISBN 978 1 84720 774 6 Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents vii x

Contributors Acknowledgements 1. Introduction: the state at work Hans-Ulrich Derlien and B. Guy Peters 2. Public employment in Britain: from working in to working for the public sector? Brian W. Hogwood 3. Breaking sharply with the past: government employment in New Zealand Robert Gregory 4. Public employment in Australia: in competition with the market Helen Nelson 5. Public employment in Canada: downsizing in a multi-layered state James Iain Gow and Sharon L. Sutherland 6. Public employment in the United States: building the state from the bottom up B. Guy Peters 7. The German public service: between tradition and transformation Hans-Ulrich Derlien 8. Working for the government in Spain: from authoritarian centralism to democratic political devolution Carlos R. Alba and Carmen Navarro 9. The French paradox: a huge but fragmented public service Luc Rouban 10. The political allocation of incessant growth in the Danish public service Lotte Bøgh Andersen, Jørgen Grønnegaard Christensen and Thomas Pallesen 11. The welfare state is female: trends in public sector employment in Sweden Jon Pierre v

1

19

40

76

106

140

170

196 222

249

268

vi

The state at work, 1

12. Conclusion Hans-Ulrich Derlien

283

Index

293

Contributors Carlos R. Alba is Professor in the Department of Political and Administrative Sciences at the Universidad Autonoma de Madrid since 1986. He is an expert on the Spanish civil service and administrative elites and has published numerous articles in national and international journals. His book publications include: Politique et Administration en Espagne: Continuité Historique et Perspectives; El sistema politico local: un Nuevo escenario de gobierno; and Administrative Reforms in Spain: Old Inertias and New Developments. Lotte Bøgh Andersen is Assistant Professor in the Department of Political Science, University of Aarhus, Denmark. There she received her PhD in public administration. Her dissertation dealt with the political strategies of public employees. Jørgen Grønnegaard Christensen is Professor of Public Administration, Department of Political Science, University of Aarhus, Denmark. His Danish and English publications deal with many aspects of public sector governance, especially central government organization. Hans-Ulrich Derlien is Professor of Public Administration in the Social Science and Economics Faculty of the University of Bamberg, Germany. His German and English publications deal with the political aspects of national and local government in general, the public service and the administrative elite in particular. James Iain Gow is Emeritus Professor of Political Science at l’Université de Montréal, Canada. Educated at Queen’s (MA) and Laval universities, after a period as a federal public servant he spent most of his academic career teaching and researching Canadian and Quebec administrative history, the Quebec public service, administrative innovations, politics and administration and administrative culture. Robert Gregory is Professor of Political Science in the School of Government, Victoria University of Wellington, New Zealand. His main fields of interest are public administration and state sector reform, policy analysis and public policymaking theory. He has published in journals such as Governance, Public Administration Review, Public Administration and the Australian Journal of Public Administration. vii

viii

The state at work, 1

Brian W. Hogwood is Professor at the Faculty of Law, Arts and Social Sciences at the University of Strathclyde, Glasgow. His main interests are: public policy, particularly public employment, and the relationship between policy change and organizational change; structures and processes in executive decision-making. He is author or co-author of several books and published articles in a variety of journals. Carmen Navarro is Associate Professor at the Departament of Political Science, Universidad Autónoma de Madrid. Her research and lecturing focus is on public policy, with emphasis on environmental policy and local government policies. Helen Nelson is an Honorary Research Associate in the discipline of government and international relations at the University of Sydney, Australia. She is a consultant in the fields of public administration and comparative federalism. Thomas Pallesen is Professor in Public Administration, Department of Political Science, University of Aarhus, Denmark. His publications deal with many aspects of public sector governance, especially the governance of hospitals and the organization of local and regional government. B. Guy Peters has been Maurice Falk Professor of American Government at the University of Pittsburgh, USA since 1984 and is Honorary Professor, City University of Hong Kong. His main interests include: comparative public policy and administration, and American public administration. He has published numerous books and countless articles in many international journals. He also served as one of the co-editors at the founding stage of the journal Governance: An International Journal of Policy and Administration. Jon Pierre is Professor of Political Science at the University of Gothenburg in Sweden. He is also adjunct professor at the University of Pittsburgh and previously held a chair in politics at the University of Strathclyde. He is a previous editor of Governance: An International Journal of Policy and Administration. One of his recent publications, co-edited with B. Guy Peters, is Handbook of Public Administration (Sage 2003). Luc Rouban is Research Director at the Centre National de la Recherche Scientifique (CNRS), at Sciences Po, Centre d’Etude de la Vie Politique Française (CEVIPOF), Paris. His research interests are public administration, civil service and political elites sociology. He is currently working on higher managers in the public and private sectors as well as on the sociopolitical evolution of the civil service. He is author of many books and articles in international journals dealing with the civil service, state reform and administrative elites.

Contributors

ix

Sharon L. Sutherland is Professor at the School of Policy Studies, Queen’s University, Canada. She studied at the University of Alberta in Edmonton, and at the University of Essex in Britain. Her research interests centre on methodology, accountability issues in responsible government, including the accountability of public servants, and political ethics.

Acknowledgements Internationally comparative projects take a long time to accomplish before results are finally published. For various reasons including multiple commitments of researchers, the initiative to projects like this often dates back ten years and more. In our case, work started with a fund provided for the period 1994–98 by the Humboldt Foundation in their Transco-op Program. This fund was co-founded by Pittsburgh University and the University of Bamberg, whom we thank for their support. Out of this initial two-country cooperation grew a preliminary comparative study of public services in the USA and Germany (Derlien and Peters 1998). This study set the framework for the second phase of the larger comparative project, for the country reports included in Volume 1 here. The initial funding enabled us to hold a conference in Bamberg in 1998, which brought together most of the teams represented in the two-volume work. The Bamberg conference was followed by similar conferences in Sandbjerg, Denmark (organized by Jørgen Grønnegaard Christensen and financed by the University of Aarhus Research Foundation) in 1999 and at the University of Strathclyde in Glasgow (organized by Brian W. Hogwood and financed by the Future Governance Program: Lessons from Comparative Public Policy) in 2000. Without these conferences during the take-off period of the project, much of the enthusiasm needed for a long shot like this endeavour would not have been generated. Our special debt, furthermore, goes to James Iain Gow, who not only contributed to these volumes, but also helped polish chapters written by authors whose native tongue is not the English language. The same job was done by Helen Nelson in several cases. To both of them we are deeply grateful. Also, we are indebted to those colleagues who were particularly engaged in this project, for their unrelenting emotional support and encouragement to complete this project. In a similar way, various people at Edward Elgar Publishing’s headquarters, who expressed continuing support for the project, are to be thanked. Finally, the tremendous job of bringing up the manuscripts to the publisher’s style requirements was done by Gisela Baumgärtner (Bamberg) with admirable precision, and Stefan Frank (Bamberg) assisted in cross-checking data and rearranging tabulations. We owe them thanks too. The remaining x

Acknowledgements

xi

weaknesses and flaws of this publication, as usual, are the responsibility of the editors. Hans-Ulrich Derlien B. Guy Peters Bamberg/Pittsburgh, August 2007

REFERENCE Derlien, Hans-Ulrich and B. Guy Peters (1998), Who Works for Government and What Do They Do? Common Trends and Structural Differences of Public Employment in the United States and Germany, Verwaltungswissenschaftliche Beiträge, 32, Bamberg: University of Bamberg.

1.

Introduction: the state at work Hans-Ulrich Derlien and B. Guy Peters

Since the 1990s it has become fashionable to consider the ‘waning of the state’, the ‘transformation of the state’ or ‘the hollowing out of the crown’ (for the latter, see Rhodes 1994; Weller et al. 1997). At the extreme, some scholars have discussed the possibility of ‘governance without government’ (Rhodes 1996). These discussions also reflect waning public confidence in the capacity of governments to govern effectively and fairly (Putnam and Pharr 2000). Underlying these observations about the decline of the state are phenomena such as the loss of traditional national sovereignty to supranational polities like the European Union and non-governmental organizations such as the World Trade Organization, the easy permeability of borders in the European Union and the lack of centrality of national governments in functional governance networks where the state allegedly figures only as one among many actors (Sorensen and Torfing 2006). In addition, privatizations have affected the shape of the state and its capacity to steer society. Similarly, retrenchment of the welfare state has resulted in increased private provision of social services in some countries, although at the same time, transfer payments for social subsidies and health care costs have increased dramatically (Castles 2004). The regulatory activities of the state have also changed. While it may be true that regulatory activities as one part of policy-making have, to a certain extent, moved upwards in multi-level governance systems, rule enforcement and monitoring as corresponding administrative activities have generally remained national. The European Union, for example, does not command implementation agencies but relies on national bureaucracies. Furthermore, despite supra-nationalization at the one hand, core state activities like determining and levying taxes, the judiciary and law enforcement as well as defence remain basically national despite trends towards international cooperation in order to cope with international problems. This holds true for NATO in the defence area as well as for growing security and police communication at the European and international levels. Finally, with increased corporatization, privatization and the use of notfor-profit organizations, the boundaries of the state have become blurred. 1

2

The state at work, 1

While the private law construction of enterprises such as telecommunications, energy and transportation in continental Europe excluded by definition these services from the public sector (as defined by public law), the corporations remain to various degrees financially and ultimately politically controlled by national governments. In other words: during the last 30 years, the picture of national public services has become more difficult to draw. Radical changes have taken place, but their substantive impact remains open to question. This presents some difficulties for international comparative research, but at the same time provides new challenges for focusing on the trees of public activities behind the forest of state images that frame sweeping theoretical analyses. Since the 1980s there have been several important efforts to come to grips with the changing nature of the public sector, in particular the restoration of the concept of the state, the introduction of the concept of ‘governance’ and the emergence of the New Public Management (NPM) reform movement. Our endeavour to map public services in ten OECD countries, trace their growth and recent shrinking and inspect their anatomy, albeit motivated by the challenge of comparative research, is related to these concepts. Certainly our findings reflect the reality of the changes identified through these concepts.

THE CONCEPT OF THE STATE With the renaissance of the continental European concept of the state in international research implying the notion of so-called ‘stateless societies’ of the Anglo-American type (Dyson 1980), a heated theoretical debate developed, not the least because of the normative underpinning of the state concept. In public law countries, the concept was linked to the 19th century political struggle about what the state (the crown and its bureaucracy) should do and must not do to civil society and the economy. As an analytical concept, though, the ‘state’ may be also a useful concept to describe the sum total of what national governments are doing: regulating and producing services. There is also consensus on using the term with respect to the state’s territorial dimension and thus its usefulness in international law and international relations (Nettl 1968). The title of this book The State at Work claims that we want to detail ‘who works for government and what do they do?’ (Derlien and Peters 1998). It alludes further to the definition given to public administration by the 19th century German theorist Lorenz von Stein: public administration is the state at work.1 Stein had set out to distinguish in a functional analysis, sectors of activity in modern societies that are either performed by society itself (in

Introduction

3

those days, for instance, education or care for the elderly) and, increasingly, those that have been taken over by the state through its agent, public administration. This notion comes close to Max Weber’s verdict that ‘political domination in everyday life means public administration’. This notion may appear problematic for various reasons. First, to those adverse to the use of the term ‘state’ in academic discourse, it conveys too much of the state–society dichotomy of 19th century monarchist Europe where the ‘state apparatus’ was, in Marxist perspective, a means of suppression. In this view, bureaucracies staffed with professionally trained, rule-bound and hierarchically appointed civil servants were an instrument of political domination, as reflected in Max Weber’s sociology of political domination. However, Weber’s ‘bureaucratic domination’ said nothing about the constitutional framework nor the public tasks performed by the state. Today there is a broad academic consensus that democracy contributed more than anything else to the emergence of the modern welfare state and that civic society is tied to public services delivered by bureaucracies (Marshall 1964). On the other hand, to a Weberian, the concept of the state can hardly be defined by referring to specific public tasks taken on and executed at a specific point in time, for ‘there is no single task that was not public in some case and from the protection of personal security to the administration of justice, there is none which all have recognized (Weber 1978, p. 55) – with one exception, however, namely, successfully claiming the monopoly of legitimately applying physical force to citizens. As our concern is less about the state than about public administration and the public service, in the post-war period and in a clearly defined number of OECD countries, we may as well consider the Weberian position a research question: to what extent are public tasks reflected in the public services and shape their composition (Rose 1975)? Third, the state certainly does not work only by providing public services and goods: infrastructure, personal social services, education, safety at home and peace abroad. Rather, it also legitimately extracts money from private pockets, finances the services mentioned and – most important to the argument – distributes and redistributes revenues to citizens and enterprises (as well as within the public sector, say to sub-governments). Even a ‘skeleton state’ (Ridley 1995) would have to employ revenue officers. In fact, beside the military apparatus (the standing army), fiscal administration (the sitting army) is a nucleus of the modern state apparatus (Rosenberg 1958). When enquiring about the mix of spending and service provision we are not only back to the historically variable array of public tasks and the modern privatization debate, our subject also links up to investigations about modern welfare regimes as undertaken by Esping-Andersen (1990). According to this author, three such regimes can be distinguished: the

4

The state at work, 1

Nordic welfare states producing a very broad assembly of public tasks through public services, with these services financed through high taxes and thus a large GNP share of the public sector; second the minimalist US ‘state’ with a low taxation level, only providing welfare to narrowly defined social groups at a minimal level while leaving the procurement of services like education also to the market. The third arrangement would be found in continental Europe, a mixture of relatively high public transfer payments (and thus a relatively high taxation level) on the one hand, and mixed public and private service procurement on the other hand. We might assume that the type of welfare regime is reflected in public employment.

GOVERNANCE Whereas the concept of the state focuses on ‘stable’ elements like public tasks, albeit acknowledging that these are subject to historical change and constitutionally defined elements of government, the concept of governance that has dominated theorizing since the 1990s, emphasizes the processes associated with the state apparatus and the branches of government (Pierre and Peters 2000, 2004). Furthermore, since the inception of the journal Governance: An International Journal of Policy, Administration, and Institutions’ in 1986 (first volume 1988), the linkage between state and society/economy has become a central concern. This is not only a reflection of earlier NPM-inspired deregulation and privatization. Since the historical landslide of 1989 and the decline of state socialism, but also in view of supra-national state building in Europe and more generally of economic and communication globalization, theorists have tried to come to grips with increasingly problematic national policy-making, in particular with the gap between the steering pessimism and (socialist) steering euphoria of previous decades (Kooiman 1993; Mayntz and Scharpf 1995). In its more radical and present-day formulations governance has become the core concept of a theoretical school claiming that informal processes within networks of political and economic, administrative and civic actors have replaced the public formal organization and hierarchy as channels of problem-solving. This, among other things, is a development ascribed to the influence of deregulation and privatization that under the philosophy of NPM has blurred the formerly clear demarcation of the private–public border; the proponents of governance theory also link it to such diverse phenomena as the growth of NGOs and the development of multilevel policy-making with supra-national institutions such as the EU. Though so suggestive and bold, the empirical foundation for this radical version of governance theory is notoriously quite weak, not the least due to a

Introduction

5

preference for case studies in various policy areas. It remains an open question whether the state is merely one among several actors engaged in solving some problem or if the state, bolstered by the classical authoritative policy areas and instruments, as well as the public service, remains a dominant actor even manipulating the networks of interaction. Regardless of whether the state is perceived as helplessly responding to the new challenges from NPM, in particular deregulation, privatization and decentralization or de-concentration, or whether NPM is a sophisticated intended way of response, of applying a new less authoritarian policy style than previously (Pierre and Peters 2000), our research can have a bearing on the discussion about governance. First, the identification of large numbers of public (state) servants might underline that there is no governance solely by the market, however incomplete state-led governance attempts may be. Second, by inspecting public service size and composition in various countries our awareness of differentiated arguments may be sharpened, thus bringing theories down to earth again. Empirical research might be especially helpful in sorting out the impact of NPM that, as a philosophy of public affairs, gave rise to huge reforms and subsequent theorizing of ‘governance without government’. Although we neither go into public finance and the welfare regime of transfers without services nor into rule production and the deregulation-reregulation paradox, it is our hope that The State at Work might provide at least important background information.

NEW PUBLIC MANAGEMENT The international reform movement stimulated by neo-conservative governments in the UK (since 1979) and the USA (since 1980) was about redrawing the boundaries of the state – towards ‘hollowing out the state’ – basically for fiscal reasons. The accompanying toolkit of reform measures became summarized under the heading ‘New Public Management’ (Pollitt 1991; Hood 1991). Whatever the impact of the ensuing reforms in the individual country – there is quite a bit of national variation (Olsen and Peters 1996; Peters and Savoie 1998; Pollit and Bouckaert 2000) – they moved public sectors in three directions: ●

Privatization of previously public tasks and publicly provided services, in particular in countries with a large array of public enterprises, according to Osborne and Gaebler (1993) led governments to reconsider whether they wanted to concentrate on steering or also to engage in rowing. First, many countries privatized nationalized

6

The state at work, 1

●

industries that did not produce public goods but, for instance, motor cars or coal. Second, the reorganization of communication systems in a broader sense, including the public monopolies of railways, postal services and telecommunications, has taken different forms in different countries and in different sectors. Privatization’s linkage to the internal market of the European Union is constituted by the EU’s claim that national markets for public service utilities should be opened for competition from the outside. Privatization did not, however, always lead to complete selling of government assets to non-governmental investors. In organizational terms its main implications have been an insistence on the separation of regulatory tasks from operative tasks. Furthermore, in political perspective the drive behind these reforms in the EU has also been fiscally motivated. The notorious 1992 Maastricht criteria fixing thresholds for national debts for entry to the European Monetary Union forced member states to cut back public debts by selling the ‘family silver’ allegedly hidden in public enterprises or more realistically, by discharging debts accumulated by the often unruly offspring of the state in the economic sector. For public services privatization, be it mere corporatization or genuine selling of assets, meant at least a formal decrease of the size of the public labour force (in budgets and national statistics). Decentralization became the second imperative, in both unitary and federal systems. In line with NPM reasoning the centre sought to hive off political and fiscal responsibility for many public services. In unitary states this fiscally motivated reform impetus was however complemented by a longstanding politically motivated drive to grant more autonomy to sub-governments or even to federalize the system. In the European Union, this trend was fostered by developments under the Single European Act 1987. The establishment of a committee of the regions induced unitary states to define politicoadministrative units to match the states in federal systems such as Germany and (later) Austria. It saw the federalization of Belgium and Spain (also prompted in both countries by powerful political forces to grant ethnic and cultural autonomy to subsystems), decentralization in France and devolution in the UK. Even within federal systems such as the USA or Germany, national policy attempted to redefine centre–periphery relations and to turn around a trend that had emerged since 1945 towards ever greater centralization of rule-making and spending power. Thus we would expect also a decentralization of public employment and a relative shrinking of central government as an employer. De-concentration, the transfer of

Introduction

●

7

ministerial personnel to agencies concerned with the implementation of policies (agencification in the UK) though would not affect the importance of central government as an employer. Although NPM was about the confines of the state, its more dramatic influence was on the state’s internal operations in the public services. Not only would the size of these be affected by privatization but their structures were also to be at the disposition of policy-makers and reformers. The historical particularity of the public services and their alleged privileges compared to employment in the private sector were to be reduced, both on normative grounds and because of presumed efficiency gains from using private sector personnel management principles. In particular, public bureaucracies with their closed career systems, with tenured positions and status-related pay schemes were the target of reformers. To what extent management practices, performance pay and contractualism affected the public service remained to be seen; previous research indicated differences in the rate of reform and compatibility with national administrative cultures, with the Anglo-American countries as spearheads.

It can be reasonably assumed that, whatever the strength of these reforms in the individual national polity, these NPM trends would have an impact on ‘the state’. However, reformers at the time of launching the reforms, did not have sufficient empirical evidence other than crude fiscal data, nor do we today have an international empirical account of the outcomes of the attempts to reshape the state. Furthermore, there are aspects of national public service systems not taken into consideration by reformers such as the ethnic composition of the service and long-term trends such as increasing part-time employment that nevertheless deserve closer inspection. It is the aim of this book to take stock of the historical development of public services, starting long before NPM, to shed light on the most important public task areas or policy programmes, to illuminate the distribution of public employment between national government and sub-governments, and to trace the impact of NPM on the public service as a social system that, in most countries, comprises at least 15 per cent of the labour force.

THE RESEARCH CHALLENGE Unlike many other areas of political science, the analysis of public administration has not yet progressed very far in the objective measurement of the basic phenomena about which we do research. Most work in the field of administrative studies involves impressionistic analyses of the structures

8

The state at work, 1

of government, rather than more objective measurement of the relevant phenomena. In large part because of the measurement problem, there has been little or no development of basic data necessary for the analysis of administrative executives. These deficiencies of measurement and data collection are even more evident when comparative analysis is considered (Derlien 1992). Although, in particular, the UK Economic and Social Research Council’s ‘Whitehall Programme’ gave rise to a number of important structural–functional comparative studies, the comparative study of public services has remained less attractive. Certainly, administrative elites have attracted periodic attention since the 1970s. Since Armstrong’s (1973) important historical book, some interesting and important comparative studies of the attitudes of civil servants (Aberbach et al. 1981; Aberbach 2003) and the structure of the ‘mandarins’ of western Europe (Page and Wright 1999) were carried out. Hood and Peters published findings of one important aspect of executive systems, rewards in high public office (1994; see also Brans and Peters, forthcoming). Also Gow and Simard (1999) dealt systematically with unorthodox, non-career public service employment in five countries. But there remains limited comprehensive data on the basic elements of public service systems. There has however been one major study of public employment in western countries (Rose et al. 1985) in which the scale and the growth of public employment in six countries (Britain, France, Germany, Italy, Sweden, USA) were investigated for the period 1951–81. That study is now largely out of date. Further, data were confined to large aggregates (generally defined in functional terms) so that little comparison could be made of the structural evolution of governments. Recently Bekke and van der Meer (2000), as part of the ambitious Leiden project (Raadschelders and Perry 1994), published Civil Service Systems in Western Europe, with a strong emphasis on the history of the core group of state personnel but neglecting the public service at large and its modern welfare functions. Analogous volumes were published for other parts of the world. Also, Auer et al. (1996) and Bossaert et al. (2001) of the European Institute of Public Administration provide a formal, legal analysis of the public service systems in the 15 European Union core states, dealing basically with aspects of personnel management including recruitment, performance measurement, pay system, pension system, work hours and disciplinary law and the ethical codex. An introductory section (Bossaert et al. 2001, pp. 45–58), cites statistics for the size of the public services and their distribution among levels of government, but the developmental perspective as well as the distribution of public service personnel among policy areas, is not dealt with. Nevertheless, the work shows the growing interest in comparative public service analysis in general and for specific European Union

Introduction

9

harmonization purposes. Finally, there have also been some post-NPM surveys (Pollitt and Bouckaert 2000; OECD 1997; 1999), whose scope is though, limited to the implementation of reform measures. Thus, while there have been some useful quantitative studies of administrative systems, there is still a relative dearth of information when evidence about the executive branch is compared with that available for other areas of political science inquiry. This has led to other efforts, last but not least by the OECD, at improving the conditions for systematic comparative analysis, but up to now they are at the incipient stage. This most recent endeavour stated: ‘Very early data are suggesting that the new measures of public employment totals are leading to significantly revised understanding of employment in the public domain’ (OECD 2007, p. 9).We anticipate that the research reported here will advance the comparative endeavour. Delineating the Public Service In this present study, when discussing the state at work we refer to millions of people who are members of the public service in each of the ten countries under investigation. Members of the public service are employed by national, regional and local governments. Their relationship to the state is defined by contracts or legal status. This relationship implies or presupposes that they are paid out of taxes and that (in principle) salaries and number of positions are contained in state budgets. Thus, we do not regard those who work in a purely functional relationship for governments as members of the public service, that is, people working in areas that are contracted out to private enterprises or who are in privatized public enterprises, even if financial responsibility is partly or totally retained by governments. Further, subsidized industries and their employees are not in the public domain. Unless these merely functionally related areas are excluded from our definition, we would end up with inflated employment figures for the public service, a weakness that partly invalidated the earlier analysis by Richard Rose and his contributors (Rose et al. 1985). This definition carries its own complexities. For instance, railways, airlines and post and telecommunications, in some states, remain government enterprises, while in other states they have been privatized. Lufthansa has always been a private law company (with national and state governments holding sizeable proportions of the shares), whereas Air France was privatized only in 1999 with the state keeping 19 per cent of the shares. Railways in the United States never went through a period of nationalization as they did in Europe; European railway systems were privatized in the 1990s. The US Postal Service however is still in the public domain, while the postal service in Germany, for instance, has been privatized. In countries with a

10

The state at work, 1

public law tradition, the delineation between public and private law companies is much easier than in the Anglo-American world. In continental Europe it is necessary to distinguish between full-scale privatization where governments sell former state-owned enterprises (SOEs) to private investors and corporatization where a SOE is transferred to private law status although the enterprise remains heavily controlled by governments. In any case, employees in these companies no longer have contracts with governments but with the relevant company. These legal differences have implications for public service statistics quite like the difference between the British National Health Service (public employment) and other health systems, where providers, medical doctors and general practitioners are private entrepreneurs although their services are paid from compulsory contributions by patients. Further, there are countries with overwhelmingly publicly run schools and universities while in other countries there is a mixture of public and private schools and universities. The delineation between public and private, however, remains valid, even if private schools and universities are heavily subsidized from public coffers. This delineation has a bearing on the estimate of the public service size and its share of total public employment. Public and Civil Service Another definitional clarification of some importance in many countries is the distinction between the public service and the civil service. In this study, public service is the broader concept including employees of all kinds: labourers, employees and a group with a special legal status commonly called civil servants. The civil service, in other words, is just a subgroup of the public service – a distinction which is not made in all of the countries. When so identified, civil servants have an exceptional position, owing to the terms of recruitment, employment, function and retirement. In most countries where the distinction is made, civil servants are found in authoritative areas of public activity where they make binding decisions interfering with the basic rights of citizens and, in most countries, they are working in government departments giving political advice and/or controlling policy implementation. If the public/civil service distinction is not observed, we arrive either at underestimating the size of the public service, a problem from which the study by Bekke and van der Meer (2000) partly suffered, or shading the qualitative difference between the larger public service including employees and labourers and the civil service proper with its particular status and function. Therefore, one methodological aim of this project is to compare the quantitative importance of the various subgroups of the public service (see Derlien and Rouban in Volume II).

Introduction

11

The above discussion elucidates the complexity of the public sector ‘employment contract’ when a comparative perspective is applied. This complexity becomes even more pronounced if the varying scope and pace of civil service reform is considered. NPM reform in general has questioned the continued reliance on the entire civil service concept, preferring contractual relations as in the rest of the labour force, and radical reformers such as New Zealand have taken action accordingly. But in other countries not normally counted among zealous reformers, the civil service has also gradually lost its importance as the basic form of government employment. In Denmark, for example, this development had already started in the 1960s and today most public employees outside the judiciary and the police are hired on the basis of contracts backed up by collective agreements (Gregory and Christensen 2004). Coming to Grips with Policy Areas At a first glance, the definition of what a policy area includes in terms of activities and personnel employed is apparently an easy one. At least from a national perspective the operational definition of a policy area might be clear-cut in national statistics. However, even at the national level, slight redefinitions often take place, for instance how to deal with personnel employed in public sector university hospitals: because of their multifunctionality, these personnel can be subsumed under the heading of either health care or research and teaching. In internationally comparative perspective, the classifications adopted by national statistical offices vary considerably and pose ultimately insurmountable problems. To mention just one of these, there are countries (for instance, Germany) where personnel working in national ministries are not counted among the staff of the policy area the ministry represents, but in a category of ‘general administration’, a category that at the same time encompasses similar personnel at subgovernment levels. However, as the extent of employment in the government departments themselves is normally relatively small, entailing not more than 20 or 30 000 people, the bias in the policy analysis data is negligible. On the other hand, in unitary states such as France, employment figures of national government departments may be deceptive, as for instance teachers in France are included in departmental statistics, whereas in federal countries, education is regularly a matter of sub-governments (state and local) with relevant regulatory activities and personnel policy often being reserved by higher levels of government. In their recent survey, the OECD foreshadows working with ‘definitions of the public domain which encompass activities that represent a significant contingent fiscal liability to government’ and which are based on the system of national

12

The state at work, 1

accounts (OECD 2007, p. 9). Nor could we solve the problem. Knowing that there are national definitional differences, our analysis focuses on the relative size of the major policy areas to end up with a reliable picture for the importance of at least the three to four major policy areas in terms of personnel employed. We do not, however, maintain that employment in minor policy areas, for instance agriculture (regulatory activities and state domains), lend themselves to comparative analysis, in particular in time series data. We therefore had to abandon the plan to devote a specific chapter to public sector employment by policy area in Volume II. In order to analyse employment by policy area in detail, one could have gone into theoretical classifications like those stipulated by Theodore Lowi (1972), distinguishing distributive, regulatory, redistributive and polity decisions, or take the sticks and carrots approach (see for instance Bemelmans-Videc 1998; Vedung 1997). In terms of decreasing intervention capacity into citizens’autonomy, such a classification would include: ●

● ● ● ● ● ●

●

Regulatory activities of the law and order type including regulation of competition, financial services and environmental and consumer protection; Taxes, monetary incentives and disincentives; Transfer payments such as pensions and social subsidies; Infrastructure projects offering usage (motorways or public toilets); Personal social services such as kindergartens or hospitals; Contracts with private service providers; Pedagogical programmes with highly normative information like health- or traffic-education, for example, teaching children how to cross a street; Provision of ‘neutral’ information like weather broadcasts or traffic information.

Such a typology of instruments (see also Salamon 2002) would be useful for explaining employment differences between particular policy areas. For example, where personal social services provided by governments through the public service are characteristic, the manpower employed is large, whereas transfer payments (for instance pensions), although amounting to huge expenditures, can be operated by few people with the help of computers.

PROJECT HISTORY AND DESIGN The research work on which this book is based started in 1994, initially funded by the Transcoop Program of the Humboldt Foundation and a

Introduction

13

matching fund of the University of Pittsburgh. It resulted in the pilot study by Derlien and Peters (1998) concerning public service development in the USA and Germany. Subsequently, conferences in Bamberg (1998), Sandbjerg (1999) and Glasgow (2000) were organized to bring together the international research team. The dimensions of analysis derive from the core aspects of the state concept, specifically public tasks and state structure or to put it differently, policy areas and the structure of the polity: ● ● ● ● ● ● ● ●

Historical development (and growth) of public services since the 1960s when data became available in most countries; Development of part-time employment; Development of female employment; Distribution of employment among levels of government; Changes in the composition by legally defined groups of public servants, among these the core civil service; Composition of the services and change of this composition by policy area; Composition by ethnicity as far as this is of concern; Influence of NPM-inspired reforms since the 1980s.

The brief national accounts contained in Volume I try to deal with all of these aspects and provide brief historical and qualitative background information. The analytical chapters contained in Volume II summarize and discuss the national findings along the theoretical dimensions set out for the national reports. Country Selection From the beginning, we have tried to avoid the ‘coins, flags and anthems’ approach of comparative research or comparative anthologies. In selecting the countries represented here, we started with two basic considerations. First, we found it necessary from the beginning to compare federal and unitary states; our reasoning was, among other more obvious typological interests, that the degree of centralization might induce different management challenges and thus reform inclination in national governments. Federal states represented here are the USA, Germany, Spain, Canada and Australia. Second, we found it reasonable to compare public services in countries with different political cultures, in particular countries that went along the path of stateless, minimalist welfare traditions, the USA and Commonwealth countries, to Nordic and Continental European countries based on public law traditions and the notion of rather strong states,

14

The state at work, 1

Table 1.1

Design: countries/contributors Polity Federal

Anglo countries Family Non-Anglo countries

Unitary

USA

Peters

UK

Hogwood

AUS

Nelson

NZ

Gregory

CAN

Gow/Sutherland

G

Derlien

FRA

Rouban

ESP

Alba/Navarro

DK

Christensen/ Pallesen

SW

Pierre

reflecting welfare regimes other than the Anglo-American family. Thus the design depicted in Table 1.1 emerged (with contributors indicated). Further ‘variables’ we wanted to catch with this selection were: ● ● ●

Countries involved in the process of devolution: UK, Spain (and France, to a certain extent); Countries with a presidential form of government (USA and France, to some extent) compared to parliamentary governments; Countries with recent government changes between liberal and conservative parties where supposed policy changes might produce redirections of public service employment policy.

Time Horizon It was agreed to begin the time series of data as early as possible after the Second World War. As is pointed out in the chapter on long-term trends (Christensen and Pallesen in Volume II), detailed data have been available in most countries only since the beginning of the 1960s, while the 1950s can be covered only with rough aggregates. Due to the time it takes to bring so many researchers together in a single project, the latest data we were aiming at were for 1998, and then as the project drew along, for 2000. It seemed reasonable to stop the update at this year, for the major reform factors supposed to impinge on public sector development had by then had time enough to work out following the start of reforms in the early 1980s. In fact, major privatizations of public tasks had been carried through, and – as far as the German case is concerned – temporary post-unification effects on the public service were levelled by 2000. Ultimately, in a secular study, it hardly

Introduction

15

matters if data depict the state of affairs in the year of publication. The important thing has not been to put together an up-to-date collection of comparative statistics but rather to make a first and serious step towards paving the field for systematic comparative analysis applying a public employment perspective. Furthermore, given the amount of privatization and the number of years spent on structural public sector reform in the countries under investigation, more recent changes in government, whether from conservative to social-democrat or in the opposite direction, did not imply major changes in public sector policy. International competition for capital investment by providing attractive private sector tax levels, prevent even leftist governments from expanding the public sectors in the way that occurred in the 1960s and 1970s. Database Any project attempting to provide comprehensive comparative information on a subject such as ours encounters substantial difficulties. To a great extent we were all at the (not so tender) mercy of the official statistics of our respective countries. We all attempted to be as diligent as possible both in identifying the sources of our data and the definitions of particular entries, especially when they differed from the agreed-upon definitions of the project. The problems we encountered were in part a function of the nature of the statistics that the official sources chose to collect and report. This problem was particularly apparent when attempting to establish a historical baseline for the systems prior to 1960. Furthermore, not only did the statistics formally change, with some entries disappearing and others appearing, but in some cases the countries themselves changed. For example, the significant changes in the Spanish polity from the 1970s onwards and the unification of Germany required careful attention to the details of what was being measured and in relationship to what. We can only say that we have made every effort to make the data as comparable as possible and to expose the cases in which the data are insufficient or inconsistent.

CONCLUSION This project has produced the most extensive and directly comparable research on public employment now available. This does not mean that the research is without its problems and potential deficiencies. Still, the coverage of countries and the quality of the individual country studies provides

16

The state at work, 1

an important expansion of our knowledge of ‘the state at work’. Further, the time series data used demonstrate the extent of change that has occurred in the public sector in most industrialized democracies and to some extent foreshadows future change. Like all studies in the social sciences this one may raise as many questions as it solves.

NOTE 1. To be precise, von Stein speaks of administration as ‘the organic activity of the state’ (p. 10). Stein’s concept has already been referred to by Derlien (1985).

REFERENCES Aberbach, Joel D. (ed.) (2003), Bureaucrats in Turbulent Times, special issue of Governance, 16, no 3. Aberbach, Joel D., Robert D. Putnam and Bert A. Rockman (1981), Bureaucrats and Politicians in Western Democracies, Cambridge, MA: Harvard University Press. Armstrong, John A. (1973), The European Administrative Elite, Princeton, NJ: Princeton University Press. Auer, Astrid, Christoph Demmke and Robert Polet (1996), Civil Services in the Europe of Fifteen: Current Situation and Prospects, Maastricht: European Institute of Public Administration. Bekke, Hans A.G.M. and Frits M. van der Meer (eds) (2000), Civil Service Systems in Western Europe, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Bemelmans-Videc, Maria L. (1998), Carrots, Sticks & Sermons. Policy Instruments and their Evaluation, New Brunswick, NJ: Transaction Publishers. Bossaert, Danielle, Christoph Demmke, Koen Nomden and Robert Polet (2001), Der öffentliche Dienst im Europa der Fünfzehn. Trends und neue Entwicklungen, Maastricht: European Institute of Public Administration. Brans, Marleene and B. Guy Peters (forthcoming), Rewards of High Public Office. Castles, Francis G. (2004), The Future of the Welfare State. Crisis Myths and Crisis Realities, Oxford: Oxford University Press. Derlien, Hans-Ulrich (1985), State and Bureaucracy in Prussia and Germany, Verwaltungswissenschaftliche Beiträge, 20, Bamberg: University of Bamberg. Derlien, Hans-Ulrich (1992), ‘Observations on the state of comparative administration research in Europe – More comparable than comparative’, Governance, 5, 279–311. Derlien, Hans-Ulrich and B. Guy Peters (1998), Who Works for Government and What Do They Do? Common Trends and Structural Differences of Public Employment in the United States and Germany, Verwaltungswissenschaftliche Beiträge, 32, Bamberg: University of Bamberg. Dyson, Kenneth H.F. (1980), The State Tradition in Western Europe. A Study of an Idea and Institution, Oxford: Robertson.

Introduction

17

Esping-Andersen, Gösta (1990), The Three Worlds of Welfare Capitalism, Cambridge: Polity Press. Gow, James Iain and François Simard (eds) (1999), ‘Symposium on the non-career public service’, special issue of the International Review of Administrative Sciences 65 (1). Gregory, Robert and Jørgen Grønnegaard Christensen (2004), ‘Similar ends, differing means: contractualism and civil service reform in Denmark and New Zealand’, Governance, 17, 59–82. Hood, Christopher (1991), ‘A public management for all seasons?’, Public Administration, 69, 3–19. Hood, Christopher and B. Guy Peters (1994), Rewards at the Top: A Comparative Study of High Public Office, London: Sage, II8, II9 Kooiman, Jan (1993), Modern Governance, New Government-Society Interactions, London: Sage. Lowi, Theodore J. (1972), ‘Four systems of policy, politics and choice’, Public Administration Review, 32, 298–310. Marshall, Thomas H. (1964), Class, Citizenship and Social Development, Garden City, NY: Doubleday. Mayntz, Renate and Fritz W. Scharpf (eds) (1995), Gesellschaftliche Selbstregulierung und politische Steuerung, Schriften des Max-Planck-Instituts für Gesellschaftsforschung Köln, vol. 23, Frankfurt a.M.: Campus. Nettl, J.P. (1968), ‘The state as a conceptual variable’, World Politics, 20, 559–92. OECD (Organization for Economic Cooperation and Development) (1997), Managing the Senior Public Service: A Survey of OECD Countries, Paris: OECD. OECD (Organization for Economic Cooperation and Development) (1999), The State of the Higher Civil Service After Reform: Britain, Canada and the United States, PUMA/HRM(99)1, Paris: OECD. OECD (Organization for Economic Cooperation and Development) (2007), ‘Towards better measurement of government’, OECD Working Papers on Public Governance, Paris: OECD. Olsen, Johan and B. Guy Peters (eds) (1996), Lessons from Experience: Experiential Learning from Administrative Reforms in Eight Democracies, Oslo: Scandinavian University Press. Osborne, David and Ted Gaebler (1993), Re-Inventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector from School House to State House, City Hall to the Pentagon, Reading, MA: Addison-Wesley. Page, Edward C. and Vincent Wright (eds) (1999), Bureaucratic Elites in Western European States. A Comparative Analysis of Top Officials, Oxford: Oxford University Press. Peters, B. Guy and Donald J. Savoie (eds) (1998), Taking Stock: Two Decades of Administrative Reform, Montreal: McGill/Queens University Press. Pierre, Jon and B. Guy Peters (2000), Governing, Politics and the State, Basingstoke: Macmillan. Pierre, Jon and B. Guy Peters (2004), Governing Complex Societies: Trajectories and Scenarios, London: Routledge. Pollitt, Christopher (1991), Managerialism and the Public Services: The AngloAmerican Experience, Oxford: Blackwell. Pollitt, Christopher and Geert Bouckaert (2000), Public Management Reform. A Comparative Analysis, Oxford: Oxford University Press.

18

The state at work, 1

Putnam, Robert and Susan Pharr (2000), Disaffected Democracies: What’s Troubling the Trilateral Countries, Princeton, NJ: Princeton University Press. Raadschelders, Jos C.N. and James L. Perry (1994), Protocol for Comparative Studies of National Civil Service Systems, Comparative Civil Service Research Consortium, School for Public and Environmental Affairs, Indiana University, Bloomington, USA and Department of Public Administration, University of Leiden, the Netherlands. Rhodes, Rod A.W. (1994), ‘The hollowing out of the state: the changing nature of the public service in Britain’, Political Quarterly, 65, 138–51. Rhodes, Rod A.W. (1996), ‘The new governance: governing without government’, Political Studies, 44, 652–67. Ridley, Frederick F. (1995), ‘Towards a skeleton state? Changes to public sector management’, in John Wilson (ed.), Managing Public Services: Dealing with Dogma, Eastham, UK: Tudor, pp. 19–36. Rose, Richard (1975), ‘On the priorities of government: a developmental analysis of public policies’, European Journal of Political Research, 4, 247–89. Rose, Richard et al. (1985), Public Employment in Western Nations, Cambridge: Cambridge University Press. Rosenberg, Hans (1958), Bureaucracy, Aristocracy and Autocracy. The Prussian Experience 1660–1815, Cambridge, MA: Harvard University Press. Salamon, Lester M. (2002), The Tools of Government: A Guide to the New Governance, Oxford: Oxford University Press. Sorensen, Eva and Jacob Torfing (2006), Theories of Democratic Network Governance, New York: Palgrave Macmillan. Stein, Lorenz von (1866), Die Verwaltungslehre, Part II, Werkliche Innere Verwaltung [German Internal Administration], Aalen: Scientia Verlag. Vedung, Evert (1997), Public Policy and Program Evaluation, New Brunswick, NJ: Transaction Publishers. Weber, Max (1978), Economy and Society, vol. 1, edited by G. Roth and C. Wittich, Berkeley, Los Angeles/London: University of California Press (German original: Tübingen, 1921). Weller, Patrick, M., Herman Bakvis and Rod A.W. Rhodes (1997), The Hollow Crown. Countervailing Trends in Core Executives, Basingstoke: Macmillan.

2.

Public employment in Britain: from working in to working for the public sector? Brian W. Hogwood

DISTINCTIVE BRITISH FEATURES There are special features in Britain which cause problems of analysis of trends in Britain across time within Britain as well as comparison of apparently similar public employment status in other countries. An important feature of British public administration is its lack of public law basis. There is no legal definition of what constitutes public employment. Even the concept of a civil servant lacks such a basis (see Drewry and Butcher 1991, pp. 9–30) – and in fact the United Kingdom has more than one civil service! A second feature of Britain is that the concept of ‘civil servant’ is both narrower and wider than obtains in many other European countries, being confined to direct employees of central government departments (leaving aside the problem of defining what those are). Employees of bodies in receipt of state funds in Britain, such as universities even when these have a state charter, are not only not civil servants, but are not considered to be public employees at all. On the other hand, with minor exceptions, employees of central government departments, even the declining number of manual workers, are nearly all classified as civil servants. This leads on to a broader point of great relevance to analysing post1970s trends in Britain: the government classifies employment as either public or private. Within the ‘private’ sector it does not distinguish for employment purposes, public function bodies in receipt of government funds (and often dependent on those funds for the bulk of income) from private interest and for-profit organizations. As this chapter will show, it is precisely in this grey area that many of the most interesting developments have occurred.

19

20

The state at work, 1

TOTAL PUBLIC EMPLOYMENT In contrast to the statement of Derlien and Peters (1998, p. 6) concerning the United States and Germany that ‘Both countries entered the time period in question [1950] after the Second World War with extremely small public sectors’, the United Kingdom public sector had swollen between 1938 and 1950 as a result of the war and post-war policies. As Rose (1985) points out, Britain was an early growth country in public employment terms. Parry (1980, p. 3) suggested that there were three major changes in Britain involving structural expansions and the assumption of new tasks by government: 1. 2. 3.

The fivefold increase in local government between the 1890s and the 1920s. The expansion of central government during and after the Second World War. The industrial nationalizations and welfare state policies of 1945–51.

Parry argued that, ‘Each of these changes altered the composition and orientation of the public sector workforce as well as its aggregate size, and each marked an important extension of public sector activity which could not easily be reversed.’ In fact, one of the major expansions – the post-war industrial nationalizations – has subsequently been reversed, and it could in turn be argued that it would be difficult to reverse the privatizations entirely in the short term. Like the earlier three major changes, privatization has altered the balance of the public sector workforce, in this case away from industrial workers to administrators and those delivering social policies. Public employment has also shown a long-run growth as a percentage of total employment in the economy. In 1851 public employment was only 2.4 per cent of employment (Parry 1980, p. 4), and by 1890 it was still less than 4 per cent. There was a sharp rise to 6 per cent by 1901. The First World War led to a temporary sharp rise in public employment, before falling back in 1921 to around 10 per cent of all employees, around which it remained until the beginning of the Second World War. Although the number of employees fell after the end of the Second World War, it remained substantially higher than the pre-war figure, as a result of Parry’s second and third major changes listed above. If we turn our attention to the post-war period (Table 2.1), we can see that numbers in public employment were more or less the same in 1960 as in 1950, emphasizing that the growth of government on this indicator was not inevitable even before the advent of the Thatcher government. From 1960 to 1978 there was

21

Public employment in Britain

Table 2.1

Total public employment in the United Kingdom, 1950–2000

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1975 1970 1965 1960 1955 1950 Percentage change 1990–2000 1980–90 1970–80 1960–70 1950–60 1950–2000

Employees (000s)

Percentage of population

Percentage of labour force

5111 5094 4944 4954 5069 5211 5292 5477 5783 5848 6052 6569 7387 7249 6515 5995 5758 6062 5757

8.5 8.6 8.3 8.4 8.6 8.9 9.1 9.4 10.0 10.1 10.5 11.6 13.1 12.9 11.7 11.0 11.0 11.9 11.4

17.4 17.7 17.4 17.6 18.3 19.0 19.5 20.3 20.3 20.8 20.9 25.0 29.0 28.9 26.3 23.8 23.2 24.9 24.7

15.6 18.1 13.4 13.1 0.0 11.2

Sources: Hogwood (1998a); data collected for author’s project ‘Public Employment as a lens on Goss-National Learning and its Impact’ in the ESRC Future Governance Programme.

consistent growth in public employment with particularly sharp increases one or two years into a Labour government. Parry (1980, p. 8) notes that the substantial increase in public employment which occurred between 1966 and 1976 was, in contrast to the 1940s, ‘internally generated rather than an expansion of the boundaries of the public sector’ (original emphasis). Public expenditure cuts from 1976 were not initially reflected in public employment reductions. The trend in public employment since 1980 provides dramatic evidence of a Thatcher effect, one continued under Major,

22

The state at work, 1

and subsequently partly reversed under Blair. However, we will have to be careful in assuming that this is true of all aspects of government policy, and be on our watch for artificial impacts of reclassifications. These changes are also reflected in public employment’s ratio to population and share of the labour force. Britain’s population grew from 50.6 million in 1950 to 55.4 million in 1970. Thereafter growth has been modest to 58.8 million in 2000, with dips in some years. Thus post-war population growth merely dampens slightly the relative significance of the growth in public sector employment to 1970 and has minimal implications thereafter. The labour force expanded from 23.6 million in 1950 to 28.7 million in 2000. The column in Table 2.1 showing percentage of labour force mirrors to a remarkable extent the profile of the total number in the public sector. Following the rundown after the Korean War, public employees as a percentage of the labour force dipped to 23.2 per cent in 1960. Public employment then rose as a percentage, to its post-war peak of 29.0 per cent in 1980. Between 1980 and 1998 public employment as a share of the total labour force fell from 29.0 per cent to 17.4 per cent.

EMPLOYMENT BY LEVEL OF GOVERNMENT In contrast to Germany and the United States, Britain is a unitary rather than a federal system. It also has had the same boundaries since 1922, longer than either Germany or the United States. This might be thought to make the task of tracking public employment on a consistent basis easier for the United Kingdom. Alas, no! A fuller characterization of the United Kingdom than describing it as unitary would be to describe it (at least until July 1999) as a highly centralized multinational state with a strong element of territorial administration and a lack of systematic constitutional or public law governing public administration. Such a description may appear to be confusing and contradictory, and indeed it is, but that aptly sums up the structure of government and the state in the United Kingdom. First, the lack of a constitutional or systematic public law basis to governmental organizations or public employment means that we lack basic operational definitions of what constitute central government departments, civil servants, local government, public undertakings and so on. Different and incompatible characterizations have been used at different times since 1950. In the absence of a clear public law basis to classification, aggregate public employment data in Britain now depend on (now international) classification of public sector financial accounts. Second, the territorial dimension to public administration means that much initial collection of data is in terms of the component nations of the

Public employment in Britain

23

United Kingdom, in the past not always on a compatible basis. Some data are available on a Great Britain basis (that is, England plus Wales plus Scotland, but excluding Northern Ireland), some on an England and Wales basis, and some for each country separately. For example, prior to 1957, data on public employment were produced by government only on a Great Britain basis. Government data provide a (reasonably compatible) basis for the United Kingdom only since 1961. Between 1922 and 1972 Northern Ireland had its own government and parliament, though the United Kingdom parliament had sovereignty, to the extent that it could and did legislate to abolish the Northern Ireland parliament and government. Northern Ireland had its own civil service, and has continued to do so under direct rule. The division between central government and local government in Northern Ireland is different from the rest of the UK since a range of policies with potentially sectarian implications are not under local government: police, housing, education, social services. Scotland, though an integral part of the United Kingdom, has its own legal system and (central government) department, the Scottish Office, and many public employment statistics are collected on a Scottish basis. For some categories, those interested in public employment trends have to aggregate the data themselves. Recent constitutional developments have important implications for the concept of ‘level of government’: ●

●

●

●

From July 1999, Scotland has had a directly elected parliament with legislative powers in designated policy areas. The United Kingdom parliament will retain the right to overrule the Scottish parliament or executive on any matter, and in theory abolish them. Civil servants working for the Scottish executive will continue as UK civil servants. The Scottish executive will be overwhelmingly dependent on the UK government for a block grant, its only direct tax raising power being to vary the basic rate of income tax by 3p in the pound. From July 1999 Wales has had an elected Assembly and Executive with no legislative or tax-raising powers, though with the power to pass or not pass secondary legislation within the framework of Westminister statutes. Although subject to periods of suspension and relapse to direct rule, Northern Ireland has an elected assembly with law-making powers. It will not initially be given responsibility for policing, but unlike Scotland, has nominal authority over social security. It will be overwhelmingly dependent on the UK government for financing. London has an elected assembly and directly elected mayor. These are arguably about local government rather than an intermediate

24

The state at work, 1

category. The government has proposed indirectly nominated regional chambers for the remaining ‘regions’ of England, with no effective executive powers, which might after a referendum be transformed into elected assemblies. The government is pursuing this proposal with a distinct lack of urgency. The medium-term implications of these constitutional changes for total public employment are explored in Hogwood (1998b), and can be summarized as involving little change in trends, in part arising from the fact that the devolved administrations are largely or wholly dependent on central government funding. Central and Devolved Employment In comparing Germany and the United States, Derlien and Peters (1998, p. 13) identify the significance of state railways and the post office (including post office banking and telecommunications) as an important explanation of the apparently greater weighting of federal employment compared to the state in Germany. In Britain the railways and post office are also important factors, but there are others which have to be discussed: ●

●

Confining ourselves to civilian employment, we find additional British features. One of the most significant is the National Health Service (NHS), whose employees are considered to be state employees. Second, the range of state enterprises before privatization covered not only railways and the post office, but also major employing industries such as coal and other aspects of transport.

Thus it is worth separating out the NHS and total public corporations from the central government total, as in Table 2.2. In terms of what most people would think of as central government the most important element is the civil service. The British concept of the civil service is both narrower and broader than the German. First, it is narrower, since it only applies to central government (and the devolved level in Scotland and Wales), and categories such as university lecturers are not even regarded as public employees, let alone civil servants. Second, it is wider in that there is not the same differentiation between civil servants and employees as in Germany; with some exceptions all employees in government departments are civil servants. Contrary to widespread assumptions, the civil service did not show inexorable growth before 1979. Closer inspection reveals a substantial fall in the size of the civil service between 1946 and 1960, and a rise between 1961

25

Public employment in Britain

Table 2.2

Public employment by level of government, 1950–2000 (000s) Total Civilian central government civilian Total Of which: public central employees NHS nationalized industries

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1975 1970 1965 1960 1955 1950

4904 4886 4735 4744 4848 4981 5042 5206 5493 5551 5749 6312 7058 6917 6143 5572 5233 5258 5053

2222 2183 2156 2149 2221 2330 2397 2527 2594 2604 2782 3354 4102 4000 3584 3418 3496 3753 3624

1251 1222 1201 1199 1186 1182 1177 1205 1231 1222 1221 1223 1174 1042 741 650 552 485 397

245 247 248 242 287 345 382 437 457 497 675 1131 1816 1816 1879 1894 2176 2318 2335

Percentage change 1990–2000 14.7 1980–90 18.5 1970–80 14.9 1960–70 17.4 1950–60 3.6 1950–2000 2.9

20.1 32.2 14.5 2.5 3.5 38.7

2.5 4.0 58.4 34.2 39.0 215.1

63.7 62.8 3.4 13.6 6.8 89.5

Local government As % of total

N

As % of total

45.3 44.7 45.5 45.3 45.8 46.8 47.5 48.5 47.2 46.9 48.4 53.1 58.1 57.8 58.3 61.3 66.8 71.4 71.7

2682 2703 2579 2595 2627 2651 2645 2679 2898 2947 2967 2958 2956 2917 2559 2154 1737 1505 1429

54.7 55.3 54.5 54.7 54.2 53.2 52.5 51.5 52.8 53.1 51.6 46.9 41.9 42.2 41.7 38.7 33.2 28.6 28.3

9.6 0.4 15.5 47.3 21.6 87.7

Source: See Table 2.1.

and 1979 from about 670 000 to 740 000. Since then the civil service has fallen to 498 000 in 2000 (see Table 2.3). This reduction in the size of the civil service was a clearly stated Conservative objective, though the downward trend was reversed three years in to the Labour government elected in 1997. When we disaggregate the civil service into industrial and non-industrial civil servants we get a rather different picture. As Table 2.3 shows, since the

26

Table 2.3

The state at work, 1

Civil servants: industrial and non-industrial, 1950–2000 Total

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1975 1970 1965 1960 1955 1950

Non-industrial

Industrial

000s

As % of public sector

000s

As % of civil service

000s

As % of civil service

498 481 484 493 512 532 553 579 592 576 579 608 715 717 722 668 643 720 746

9.7 9.4 9.8 10.0 10.1 10.2 10.4 10.6 10.2 9.8 9.6 9.3 9.7 9.9 11.1 11.1 11.2 11.9 13.0

446 429 431 439 459 475 494 509 509 495 499 500 549 526 495 420 380 386 433

89.6 89.2 89.0 89.0 89.6 89.3 89.3 87.9 86.0 85.9 86.2 82.2 76.8 73.4 68.6 62.9 59.1 53.6 58.0

29 31 33 36 36 42 46 52 61 64 67 102 159 178 209 235 263 334 313

5.8 6.4 6.8 7.3 7.0 7.9 8.3 9.0 10.3 11.1 11.6 16.8 22.2 24.8 28.9 35.2 40.9 46.4 42.0

Percentage change 1990–2000 14.0 1980–90 19.0 1970–80 1.0 1960–70 12.3 1950–60 13.8 1950–2000 33.2

10.6 9.1 10.9 30.3 12.2 3.0

56.7 57.9 23.9 20.5 16.0 90.7

Source: See Table 2.1

end of the Korean War the number of industrial civil servants has declined consistently until the effect of expansion under the Blair government manifested itself in 2000. The decline from 1979 accelerated sharply, to the extent that the category has almost disappeared to only 5.8 per cent of the civil service, but it was an acceleration of an existing trend. The bulk of ‘civil service’ cuts have not come from the traditional British image of administrative or clerical civil servants but from industrial workers. Their number declined from 168 000 in 1979 to 29 000 in 2000.

Public employment in Britain

27

The pattern for non-industrial civil servants is rather different. Their number did decline persistently but gently in the 1950s, and thereafter showed a consistent rise to the mid-1970s, with a sharp increase in 1976. The Conservative government did cut the number of non-industrial civil servants between 1979 and 1985, but their number levelled off thereafter until 1993. The overall decline in non-industrial civil servants between 1979 and 2000 was from 568 000 to 446 000, a much less dramatic percentage decline at 21 per cent than for industrial civil servants (83 per cent). In interpreting changes to the numbers in the civil service we have to take into account the way in which some activities previously carried out by civil servants have been reclassified, in some cases as a prelude to privatization, but in other cases still within the public sector. Major reclassifications alone accounted for around 53 000 of the reduction in civil service numbers, and exclude cases of direct full privatization, which in some cases have involved the services of the newly privatized organization being bought back by the government (for a list see Safford and MacGregor 1998, pp. 28–9). One important development in the civil service has been the establishment from 1988 onwards of ‘Next Steps’ agencies, so named after the report Improving Management in Government: The Next Steps (Efficiency Unit, 1988). As of July 1997, these accounted for 77 per cent of the UK civil service. Staff remain civil servants (though to confuse matters, agencies also include Northern Ireland civil servants, Forestry Commission civil servants, and military staff). The exceptions are former agencies, such as Her Majesty’s Stationery Office (HMSO), now The Stationery Office, which have been privatized on the basis of the government contracting to buy back the services which were previously provided within government. Their employees are now, however, ‘off the books’. Agencies range in scale from the Social Security Benefits Agency with a staff of overall 71 000 to a singlesite conference centre with under 30 staff. Until July 1999 comparison of central and devolved government applied only up to 1972 for a small portion of the United Kingdom, Northern Ireland (amounting to 2.8 per cent of the population in 1996). Systematic records of the relative share of central (United Kingdom), devolved and local employment were never kept. We do have figures for all aspects of public employment in Northern Ireland for 1996 (Northern Ireland Digest of Statistics). These confirm the minimal employment of the central UK government, even post the 1960s troubles, but relate to the period when many former functions of local government had been taken over by appointed boards. It is not possible to give a full analysis of the position in Scotland, Wales and Northern Ireland of central compared to devolved government, for the simple reason that government does not publish statistics of total central

28

The state at work, 1

government employment by country. We do have data about civil servants, the health service, and appointed bodies reporting to the Scottish, Welsh and Northern Ireland Offices, but the missing link is for British-wide appointed bodies and public corporations. We can say that the ratio of devolved public employment is by far the greatest in Northern Ireland, followed by Scotland, then Wales. Local Government Employment Superficially, local government structure in Britain looks relatively simple and has a distinctly low number of local authorities for its population compared to other countries. In terms of general purpose authorities, Scotland, Wales, Northern Ireland and the bulk of the population (though not the area) of England live in 195 single-tier (‘unitary’), authorities. The remainder of the population of England has a two-tier structure: 34 counties, which deal with the bulk of local services, including education and social services, and districts. There are also special purpose authorities (which sometimes coincide with the areas of counties, but are almost always larger than the areas of unitary authorities), dealing with police, fire, and in major conurbations, public transport. Authorities may also be grouped for land-use planning purposes. In contrast to the United States, there is no overall census of authorities in Britain. Summary published data do not distinguish between types of local authority, whether general or special purpose, though police can easily be separated out. The picture of relative simplicity becomes more complex when we note that there are variations in the powers of unitary authorities, with those in Northern Ireland having few responsibilities (education, housing, and policing all being administered through non-local government bodies), while those elsewhere cover most local government functions directly, other than police, fire and, in major conurbations, public transport. The real complexity in Britain results from the major changes in the structure of local government, not just boundary changes, which render it not meaningful to try to track trends in public employment in particular categories of authority. To underline the scale of the upheavals, there is not a single authority in Britain which has not undergone at least one major upheaval since the 1960s. Broadly speaking, the position until the 1960s was that the differing systems in Scotland and in England and Wales were as complex as those in the United States and Germany. Many major cities were single-tier county boroughs (counties of cities in Scotland), covering all or nearly all local government functions. Inner London had a two-tier system of county and boroughs. The rest of the country had a system of

Public employment in Britain

29

counties, below which were various types of boroughs (some of which exercised functions which elsewhere in the same county were carried out by the county level), and urban or rural districts. Since then there have been major reorganizations in 1974–5, 1985, and 1995–98. Thus, even if data were readily available for public employees in each category of local authority in any given year, they would not be particularly helpful for tracing public employment in local government over time. The number and scale of reorganizations of local government in Britain emphasize the extent to which its structure is the creature of central government, rather than having an entrenched constitutional status. Another indicator of this is the loss of local authority functions to central government appointed bodies. These considerations having been set out, we can now turn to the trends in local authority employment. Local authority employment rose steadily under both parties from 1961 to 1973. The dip in 1974 is the reciprocal effect of the increase in NHS employment arising from the transfer of local authority health functions. Allowing for this, there was a levelling off of employment in the mid-1970s; there was not a drop because current expenditure was maintained, while capital expenditure was cut. Employment levels from 1979 to 1988 provide support for neither ‘draconian cuts’ nor ‘local extravagance’ claims. Underlying local authority employment levels were effectively static from 1975 to 1988. The apparent gradual increase from 1982 reflects the numbers on the job creation Community Programme who were included as local authority employees. The apparent sharp drop between 1988 and 1989 reflects the ending of the Community Programme, and the transfer of polytechnics from nominal local authority control. However, local authority employment is an aggregate which covers a number of programmes, of which education is the largest. Both education and police saw substantial increases over the decades, but education employment fell from 1989 with the loss of functions. Local authority personal social services have shown a very high percentage growth rate – from 170 000 in 1961 to 400 000 in 1997, a total increase of 135 per cent. By contrast, local authority construction, which includes some direct manual labour operations, fell from 103 000 to 65 000 over the same period, a decline of 37 per cent. Public Employment by Geographical Area Derlien and Peters provide a breakdown of state and local employment by state/Länder. Since the United Kingdom is a unitary state, such a breakdown is neither practicable nor meaningful. What is possible is to provide employment estimates for the devolved activities in Scotland, Wales and

30

The state at work, 1

Table 2.4 ‘Devolved’ and local government employment: population ratios (per 1000 inhabitants) by country, 1997

England Wales N. Ireland Scotland

Devolved

Local

Total

Not applicable 19.9 91.6a 29.7b

43.5 53.6 5.6 56.9

43.5 73.5 97.2 86.6

Notes: a. Calculated on the assumption that public order and security will not be devolved. b. Does not include public corporations coming under the Scottish Parliament. Sources: Hogwood (1998c); Civil Service Statistics, Public Bodies 1997; Monthly Digest of Statistics, Britain 2000; Northern Ireland Digest of Statistics.

Northern Ireland, and local public employment in these and in England (see Table 2.4). This shows that Northern Ireland has the lowest per capita employment in local government, reflecting its limited functions. Conversely, partly as a result of this but also of the wide range of functions other than law and order to be devolved, devolved share of public employment is higher in Northern Ireland. Scotland has a higher ratio of devolved employment than Wales, reflecting the way in which many functions are still administered on an England and Wales basis. Scotland and Wales both have noticeably higher ratios for local government.

PUBLIC EMPLOYMENT BY CENTRAL GOVERNMENT Since 1996 it is possible in Britain to distinguish between the Senior Civil Service (all civil servants at the old Grade 5, or Assistant Secretary, and above) and other personnel, but this is not the same as the distinction between core staff in London and field staff working elsewhere since: 1. 2.

Even leaving aside the special cases of Scotland, Wales and Northern Ireland, many headquarters staff are now based outside London; Most civil servants working in London are field staff delivering services to the London population.

A further complication in making comparisons is that the concept of the civil service is more narrowly drawn in Britain, certainly than in Germany.

Public employment in Britain

31

Central Government Structure Derlien and Peters state of the USA and Germany: ‘The structure of federal government in both countries is relatively simple.’ In Britain the structure is relatively complicated to the extent that we notoriously do not know what a government department is (Hood et al. 1978). This reflects the lack of a constitutional law basis to public administration. However, a grouping within this diversity can be produced to attempt a comparison with other countries. We can consider the current structure of central government to be grouped into ‘Whitehall families’ (Hogwood 1995), each of which is focused on a ministerial department, the head of which is a member of the Cabinet. The 2004 list of 17 consists of two ‘core departments’, 14 functional departments, and one territorial department: Cabinet Office Chancellor of the Exchequer’s departments Home Office Law Officers (England, Wales, Northern Ireland) Department of Constitutional Affairs Foreign and Commonwealth Office Department for International Development Ministry of Defence Department for Work and Pensions Department of Health Office of the Deputy Prime Minister Department for Transport Department for Education and Skills Department of Trade and Industry Department for Environment, Food and Rural Affairs Department for Culture, Media and Sport Northern Ireland Office For the last two decades, all heads of ministerial departments have been members of the cabinet (though the Attorney General, responsible for the Law Officers department is not), but prior to then it was normal for some ministers heading departments not to be members of the Cabinet. (Some ministerial departments have two ministers in the Cabinet.) A distinctive feature of the organization of British government is the frequency of major changes, reflecting the ability of the Prime Minister to make such changes effectively by announcement. This has major consequences for the ability to track public employment by department or policy area over time. Within each family at any given time there are to a considerably varying extent:

32

The state at work, 1 ● ●

● ● ●

Since 1988, agencies within departments carrying out largely executive tasks; Non-ministerial departments with varying degrees of policy and operational autonomy, some of which are agencies, and some of which have agencies within them; Non-departmental public bodies not normally staffed by civil servants and relying to varying degrees on government finance; Public corporations; Bodies not recognized by government as public bodies but set up as a result of government action or with government approval, and carrying out public policy delivery functions, including regulation.

In principle, all the above report in some way via a head of a ministerial department, though the minister’s ability to intervene or control varies both legally and in practice. An important exception, not listed above, is the National Audit Office, which reports directly to Parliament. Personnel Including Field Staff It should be clear that the problems outlined above cause considerable difficulties in analysing change over time. Ironically, it is sometimes easier to make comparisons between the 1950s and the 1990s than between the 1970s and the 1990s because it is easier to aggregate the smaller ministries of the 1950s to 1990s departments than compare the 1990s with the ‘giant’ departments of the late 1960s and early 1970s. The analysis presented here covers only civil servants in the narrow British sense (which amongst other things excludes the Northern Ireland and Forestry Commission civil services) and excludes the Post Office prior to 1969. As in Germany and the United States, civil service employment has in the past been dominated by civilian defence employment, at 17.3 per cent of the total in 1997. However, this relative dominance has been dramatically reduced since 1950, or even since 1980. With the transfer of responsibility for the collection of social security contributions from the Department of Social Security to the Inland Revenue, the Chancellor of the Exchequer’s departments form a grouping comparable in scale to Defence civil service employment. Even with the loss of the Contributions Agency, the Department of Work and Pensions (which replaced the Department of Social Security) is a major civil service employer. The picture would be substantially different if non-civil-service civilian employees of bodies under the aegis of central departments were included (that is excluding local government). The Department of Health would emerge as the clear winner with 708 700 in 1997. The Department of Trade

Public employment in Britain

33

and Industry would be a major player, since it is the Department responsible for the Post Office. This emphasizes the extent to which different ‘Whitehall families’ have different structures of civilian employees. Core Departmental Personnel The concept of core personnel is ambiguous in the British context, having at least three meanings. The first is the staff of core departments, currently considered to cover the staff of the Prime Minister’s Office (very small), the Cabinet Office, the Privy Council Office (33 staff), and the Treasury (but not the Chancellor’s other departments: 893 staff). These departments handle the key coordination, personnel management and resource allocation functions. They account for only a very small proportion of the civil service (0.4 per cent). The second meaning of core is the part of departmental activities not carried out by executive agencies or non-ministerial departments. Executive agencies and bodies working on similar lines account for 77 per cent of the civil service, and other non-ministerial departments for a further proportion. However, as I have pointed out elsewhere (Hogwood 1995), even at the stage that the process of establishing agencies was nearly complete, there was considerable variation in both the proportion of activities retained and the extent to which these actually contained policy or executive functions. The final possible meaning of core refers to the top civil servants. Identifying such a group has now become easier, since the formation of the Senior Civil Service, including all civil servants from Grade 2 down to Grade 5 and all agency chief executives. This together with the topmost officials, accounted for 0.8 per cent of the civil service in 1997. Interestingly, as the size of the civil service as a whole has declined, the topmost level has grown slightly.

PUBLIC EMPLOYMENT BY POLICY AREA This section compares the public employment devoted to different policy areas in all the component parts of the UK, and all levels of government. Looking at public employment by policy area gives an interestingly different perspective from that offered by public expenditure. Social Security is by far the largest public expenditure programme, but it is dwarfed in public employment terms by defence, health, and education, since social security staff dispense cash to others, whereas much of the expenditure on defence, education and health goes on employing those

34

The state at work, 1

who are delivering the service. In this section the key features and overall pattern are examined. The last four decades (1960s to 1990s) have seen important changes in the relative make-up of the public sector. In 1961, the largest single category was nationalized industries and other public corporations at 38 per cent; by 1975 they had fallen to 28 per cent of total public employment. By 1997, as a result of reductions in the number of employees in nationalized industries and the post-1979 privatizations, they accounted (excluding NHS trusts) for only 7.9 per cent. The armed forces, already down from 14 per cent in 1954 to 8 per cent in 1961 and 4.6 per cent in 1975, fell further to 4.2 per cent in 1997. The one clear area of growth has been in social policies. The National Health Service share of total public employment grew from 8 per cent in 1954, to 10 per cent in 1961, to 14 per cent in 1975 and 24 per cent in 1997. Local authority education’s share grew from 13 per cent in 1961 and 21 per cent in 1975 to 24 per cent in 1988 before declining slightly to 23 per cent in 1997 following successive removals of education responsibilities, but also in the context of reductions elsewhere in the formal public sector. Local authority social services grew from 2.9 per cent in 1961, to 4.3 per cent in 1975 and 7.9 per cent in 1997. Thus social services (health, local education and social services) now account for 55 per cent of public sector employment, even without including civil servants working on these policy areas and the penumbra of nominally private sector but publicly funded education and social services. Over the last 100 years public employment has developed from being dominated by defence and policing, through the growth of social service functions, a rapid growth of both social service and economic functions after the Second World War, to a situation now where the armed forces and police account for only a small proportion of the official total (8.2 per cent combined), as do total public corporations (excluding health) at 7.9 per cent. Public employment is now dominated by the provision of social policies, a domination which is greater if the government-funded and regulated quasi-public sector is included.

ASPECTS OF SOCIAL DIFFERENTIATION OF THE PUBLIC SERVICE The gender and ethnic basis of public employment are both issues of public concern, though only the gender basis is collated by government for public employment as a whole. The issue of the growth of female employment is strongly related to the growth of part-time employment.

Public employment in Britain

35

Public Employment by Legal Status There is no legal basis to the status of any category of public employment in Britain, because there is no systematic basis of constitutional or public law in Britain. Even the concept of a civil servant lacks such a basis (see Drewry and Butcher 1991, pp. 9–30). That said, we can make some meaningful distinctions, as well as drawing attention to problems with these distinctions. The concept of the Civil Service is very much narrower in Britain than in Germany. The Civil Service category currently only covers (some) central government employees. In addition to ‘the Civil Service’ (which can be divided into the Home and Diplomatic Service) there are the Forestry Commission and Northern Ireland Civil Services, both relatively modest in scale. The devolved Scottish and Welsh Executives are staffed by persons in ‘the Civil Service’, while that for Northern Ireland is staffed by the separate Northern Ireland Civil Service. Staff in central government departments (for which there is no definition) are invariably civil servants, but some civil servants work for other bodies, with the Health and Safety Executive and the Advisory, Conciliation and Arbitration Service being the main bodies. Prior to its abolition in 1988 the largest non-departmental public body, the Manpower Services Commission, was staffed by civil servants. In the annual consolidated summary of public employment, civil servants working in what are considered for national accounts purposes to be public corporations, are classified under that heading rather than central government, even though they are working for central government departments. Within the civil service there is a distinction between industrial and nonindustrial civil servants. Industrial civil servants certainly do not correspond to the German concept of labourer and probably not even to the US concept of blue collar, since they include senior supervisory and technical staff. The size of the industrial civil service has undergone a dramatic decline over the decades as a result of rundown of some functions (for example defence support), privatization, and reallocation to other parts of the public sector. With the near disappearance of the industrial civil service, the typical civil servant now conforms to the conventional image of a bureaucrat working in an office. The decline in industrial civil servants is more dramatic than either labourers in Germany (which include Länder and local employees) or blue collar federal employees in the United States. There used to be a distinction between Administrative, Executive and other classes in the civil service, but following the Fulton Report of 1968 a unified grading structure was developed. More recently there have been two important developments:

36

The state at work, 1

1.

the formation of a Senior Civil Service covering Grade 2 down to Grade 5; the decline in significance of grades as such, with an increased linkage to the responsibilities and salary to be attached to particular responsibilities. For statistical purposes, the civil service has devised ‘responsibility level’ bands, but consistent data on this basis only goes back to 1984.

2.

Public Employment by Gender and Ethnicity The changing gender basis of public employment in Britain is strongly related to the relative growth of part-time public sector employment. Conventionally, the gender basis of public employment is discussed in terms of the changing female composition, but before doing so it is worth remarking on the changes to male employment in the public sector, since these have important social implications for assumptions about employment and income in relation to family structure. There has been a dramatic decline in male full-time public sector employment and, of possibly even greater significance, a decline in job openings for young, unskilled males. This arises in a number of ways: ● ● ●

The decline in size of the armed forces, and the increased requirement for skills in recruits; The decline in employment in industries such as coal, steel, railways, even prior to their privatization; The contracting out and down-sizing of many unskilled services such as gardening, garbage disposal, property maintenance and so on.

Parry (1980) found that the public sector workforce was approximately 24 per cent female in 1951 and 44 per cent in 1976 and from lagging behind the private sector in the proportion of female employees had overtaken it. Since then the proportion of female employees in the public sector workforce has risen to 59 per cent in 1996 (calculated from data in Economic Trends). However, females accounted for only 43 per cent of full-time employees, while they accounted for 91 per cent of part-time employees. This trend reflects the decline and privatization of the full-time maledominated nationalized industries and the growth in education and health, which employ a large number of women on a part-time basis. However, the total number of female employees in the public sector has declined since the mid-1980s, with the number of full-time employees falling, while part-time female employees continued to rise. As interesting as the overall proportion of male and female in different parts of the public sector is the relative status of genders. We have already

Public employment in Britain

37

noted that female employment is heavily biased towards part-time rather than full-time. Figures for grading by gender are not available for the public sector as a whole but are available for staff by responsibility level in the nonindustrial civil service. At first sight, the figure of 51.1 per cent of nonindustrial civil servants being female in 1997 appears to indicate gender equality. However, within that total there is a clear preponderance of female staff in the lower responsibility levels and much lower proportions of staff in the higher. The gradient is a smooth one across all levels: the higher the level, the lower the proportion of women. When we turn to trends across time, the future is much more interesting than the average increase in the proportion of women from 47.8 per cent in 1984 to 51.1 per cent in 1997 would suggest. The proportion of women has declined substantially in the lowest level and declined slightly in the second lowest level, but has risen substantially in all levels upwards from Executive Officer – defined by the Civil Service as the first management level. Thus, the gender gradient curve, while still marked, is much less steep than in 1984. The gap in 1984 between top and bottom was 73.3 per cent (79.2  5.9), while in 1997 it had been reduced to 51.3 per cent (66.3  15.0). In other words, the gender gap has been reduced not simply by an increased female proportion in the higher levels but by an increased male proportion (and absolute numbers) in the lower levels. There is no constitutional bar in Britain, unlike in Germany, to being ethnic-aware in collating public employment data. Data are not published for the ethnic composition of public employment as a whole, but are published for the civil service. This information should be treated with considerable caution because it is based on voluntary self-reporting, and there is a high number of persons classifying themselves as ‘other’. In total, ethnic minority representation across the Civil Service has increased from 4.2 per cent in 1989 to 5.7 per cent in 1997, compared to 5.3 per cent of the economically active population in 1997 (Cabinet Office 1998, p. 71). However, ethnic minority representation is highest at the lower responsibility levels and lowest at the highest, though the gradient is not quite as smooth as we found for gender. There are huge geographical variations in ethnic minority representation among regions, largely but not completely reflecting the ethnic minority composition of the local population (see Cabinet Office 1998, p. 79).

WHAT DOES WORKING FOR GOVERNMENT MEAN? There is some evidence of moves towards a different structure of public employment within the totals, some representing continuing trends, some

38

The state at work, 1

more recent developments. There are continuing moves towards a greater proportion of public employment being part-time, in part reflecting the decline or privatization of full-time areas such as the armed forces and the nationalized industries. There is a very strong relationship between this continuing development of part-time employment and the continuing growth in the female proportion in the public sector workforce. Considering the broader public sector as a whole, including the quasipublic sector, there have certainly been moves, particularly since 1987, from ‘hierarchy to contract’ (Harrison 1993), notably in the areas of training, some education, health, the social services, and compulsory competitive tendering in local government and market testing in central government. In some cases the contract is with bodies whose employees are still in the public sector, in others, such as training, in bodies considered to be in the private sector. In other cases, such as education, the arrangement is not formally a contract, but funding is awarded on the basis of conditions and with the availability of sanctions from the government body allocating funds. One consequence of the trend to contracting as a means of providing inputs to the provision of public services is an increasing ambiguity about what employment in relation to the public sector now entails: working in, or working for? There is a substantial and increasing discrepancy, amounting to over 1 million, in the (lower) number of public employees as reported through the employer statistics primarily employed in this book and the reporting of employees about their sector status through the government’s Labour Force Survey. One possible partial explanation which has been put forward is that individuals’ views of their job may differ from the government’s definition of their status (see Bassett 1996). These developments have coincided with restructuring and privatization of nationalized industries, leading to a dramatic decline of those dependent on government funding in that sector. Since 1979 there has been a clear reduction in direct industrial employment in all parts of the public sector. However, this has not led to a simple retreat to a traditional core of defence, law and order and tax collection. Defence itself has seen substantial reductions. Proportionately, social policies dominate public employment to an even greater extent than at any previous time, even if we exclude those who are officially ‘off the books’ in public employment terms, though not in public expenditure ones.

REFERENCES Abramovitz, Moses and Vera Eliasberg (1957), The Growth of Public Employment in Great Britain, Princeton, NJ: Princeton University Press.

Public employment in Britain

39

Bassett, Philip (1996), ‘Public sector puzzle casts shadow over jobs figures’, The Times, 16 July 1996. Butler, David and Gareth Butler (1994), British Political Facts 1900–1994, Basingstoke: Macmillan. Cabinet Office (1998), Equal Opportunities in the Civil Service 1995–97, London: Cabinet Office. Derlien, Hans-Ulrich and B. Guy Peters (1998), Who Works for Government and What Do They Do?, Verwaltungswissenschaftliche Beiträge 32, Bamberg: University of Bamberg. Drewry, Gavin and Tony Butcher (1991), The Civil Service Today, 2nd edn, Oxford: Blackwell. Economic Information Branch, Office for National Statistics (1997), ‘Employment in the public and private sectors’, Economic Trends, March, (552), 15–25. Efficiency Unit (1988), Improving Managment in Government: The Next Steps, London: HMSO. Harrison, Anthony (ed.) (1993), From Hierarchy to Contract, Hermitage, Berkshire: Policy Journals. Hogwood, Brian W. (1995), ‘Whitehall families: core departments and agency forms in Britain’, International Review of Administrative Sciences, 61, 511–30. Hogwood, Brian W. (1998a), ‘Towards a new structure of public employment in Britain’, Policy and Politics, 26 (3), 321–41. Hogwood, Brian W. (1998b), ‘Reinventing public employment: the restructuring of public sector employment in Britain’, paper delivered to the Annual Conference of the American Political Science Association, Boston, September 1998. Hogwood, Brian W. (1998c), ‘The administration of Scotland: from Scottish Office to Scottish government’, in P. Norton (ed.), The Consequences of Devolution, King Hall Paper no. 6, London: Hansard Society. Hogwood, Brian W. and B. Guy Peters (1983), Policy Dynamics, Brighton: Wheatsheaf. Hood, Christopher, Andrew Dunsire and Suky Thomson (1978), ‘So you think you know what a government department is . . .?’, Public Administration Bulletin, 27, 20–32. Parry, Richard (1980), ‘United Kingdom public employment: patterns of change 1951–76’, Studies in Public Policy, no. 62, Glasgow: Centre for the Study of Public Policy, University of Strathclyde. Rose, Richard (1985), ‘The significance of public employment’, in R.Rose et al. (eds), Public Employment in Western Nations, Cambridge: Cambridge University Press. Safford, J. and Duncan MacGregor (1998), ‘Employment in the public and private sectors’, Economic Trends, March (532), 22–33.

3.

Breaking sharply with the past: government employment in New Zealand* Robert Gregory

INTRODUCTION New Zealand is a country of 3.8 million people (at 1999). Its indigenous people are the Maori, probably of Pacific origins some 1000 years ago. From the early decades of the 19th century the country was colonized by Britain and thus became established as a constitutional monarchy with the British sovereign as Head of State (although sovereignty powers are exercised by a Governor-General), and with a Westminster model of representative government. New Zealand has a unitary system of government, with a unicameral Parliament. Unitary government has been constituted since 1876, when provincial governments were abolished. In the latter part of the nineteenth century the basis of a welfare state was created, administered almost exclusively by central government, which remains by far the dominant source of administrative power. New Zealand’s welfare state was consolidated and expanded greatly by the first Labour government, which held office between 1935 and 1949. During this period the Parliamentary process became dominated by two main political parties, Labour and National. Apart from two Parliamentary terms (1957–60 and 1972–75) when Labour held power after 1949, the National Party controlled the treasury benches until, in 1984, the fourth Labour government was elected to office under Prime Minister David Lange. Traditionally, New Zealand’s political culture has been marked by pragmatic expediency rather than ideological or doctrinaire commitment, and its political economy has commonly been depicted in the post-war years as something of a model of the modern ‘mixed economy’. Abstract ideological interpretations were commonly dismissed as being irrelevant in a frontier, pioneering, society, where progress was seen to depend more upon the ability of the ordinary person to use their own resourcefulness than upon 40

Government employment in New Zealand

41

ideas and theories imported from foreign parts. Particularly during the three decades after the Second World War, New Zealand politics was based on a bipartisan consensus that was committed to the welfare state, full employment, financial stability, economic growth and a favourable balance of trade. From the mid-1930s until the early 1980s the development and management of New Zealand’s welfare state was governed largely by Keynesian and neo-Keynesian economic ideas, on the back of assured markets for primary produce. Although New Zealand’s is a capitalist political economy, the Marxist view of the state as an instrument of the ruling class has never taken firm root (Roper 1997). Certainly New Zealand inherited from Britain a strong trade union disposition to depict politics in terms of the class struggle, and to that extent political cleavages could be drawn more clearly along these lines than has been the case in, say, America. But from the mid-1930s the development and maintenance of a liberal welfare state, in the British tradition, rather than a social democratic welfare state in the Scandinavian mode (Rudd 1997), was the focus of a political middle-ground which the two main political parties sought to stake out as their own. In its earlier years the Labour Party (established in 1916) built its power base on the aspirations of a growing urban working class, but it soon found it electorally necessary to cast aside its avowedly socialist principles, in favour of policies which were seen to be much less threatening to the interests of property owners, both rural and urban. In more recent years, the party has depended less on its trade union base and much more on its appeal to educated urban liberals. The National Party, on the other hand, successfully coalesced around the sometimes conflicting interests of farmers, urban employers including manufacturers, and other white-collar workers. Up to and including the general election of 1993, New Zealand operated a first-past-the-post electoral system. However, in 1996 a system of proportional representation was introduced (based on the German model of mixed member proportionality), following a popular vote for change. The egalitarian spirit that characterized the relationship between the state and society in New Zealand was also manifest, unsurprisingly, in the traditions of the New Zealand Public Service. In the latter years of the 19th century and until the introduction of a merit-based, unified career service in 1912, the public service was a major source of political patronage (Henderson 1990; Lipson 1948). Entry to the permanent service could be gained only by an examination, administered on a non-competitive basis, but governments of the day could, and did, swell the ranks of public servants by making ‘temporary’ appointments – some of whom stayed on for more than 30 years – as a means of distributing political favour.

42

The state at work, 1

From the establishment of a unified career service administered by a central agency through until the 1980s, the size of the New Zealand public service grew steadily, as gauged not only by the numbers of employees but also by the range of functions that government sought to administer through its own offices. Demographically, during this period New Zealand public servants were overwhelmingly male and non-Maori of European descent; but the service provided secure careers for steady performers who lived their lives in a style almost totally indistinguishable from the mores of society at large. From 1912, the career service was administered by a central personnel agency – originally a Public Service Commissioner, later the Public Service Commission, and later still the State Services Commission (SSC). This tradition has continued ever since then, and unlike some European countries and America, changes of government in New Zealand are not marked by the appointment of executive officials identified with the political colour or programmatic commitment of incoming governments, even though the current State Sector Act 1988 gives ministers more say than they had previously in the appointment of departmental chief executives. However, since the mid-1980s there has been a major increase in the numbers of political appointees made within ministerial offices, for executive, secretarial, policy advice and media liaison purposes. Such appointees are not part of the public service. In New Zealand, patterns of public sector employment, as distinct from numbers, were relatively stable during the post-war decades, until the advent of the fourth Labour government in 1984. This government embarked on a major programme of neo-liberal economic policy reform, which provided the rationale for the radical transformation of state sector structures and processes. As a result, not only the patterns but also the numbers for state sector employment were drastically changed, especially during the 1990s. State Sector Reform, 1986–98 The whole configuration of state sector employment was drastically altered by the reforms introduced from the mid-1980s by the fourth Labour government, and continued under subsequent regimes. Here it is sufficient to record that these reforms involved, inter alia, a massive programme of privatization of publicly-owned utilities (Duncan and Bollard 1992), the corporatization of government trading organizations in the form of state-owned enterprises (SOEs), the creation of a plethora of other single-purpose state-owned organizations (crown entities) existing in an ‘arm’s-length’ relationship with the government of the day, and huge reductions in the numbers of employees officially designated as public servants.

Government employment in New Zealand

43

The attempts to improve allocative and productive efficiency within the New Zealand economy as a whole had to target the public sector, which in 1984 accounted for about 22 per cent of GDP, while government expenditure (including transfers) was around 39 per cent of GDP (Scott et al. 1990). The main intellectual tools used to achieve this end were public choice theory and agency theory, and these were applied through the statutory instruments of the State Owned Enterprises Act 1986, the State Sector Act 1988 and the Public Finance Act 1989. The institutional driving force behind these reforms was the Treasury, which applied the theoretical frameworks with what Schick (1998, p. 124) has called ‘full fidelity to their internal logic’. The main features of the restructuring have been well documented and discussed elsewhere (see Scott 1995; Boston et al. 1996; Schick 1996). They involved the corporatization of government trading activities, placing them on a market-competitive basis; the privatization of a wide range of state assets (including the telecommunications services, railways, the national airline and the Bank of New Zealand); managerialist (based on New Public Management) and commercializing innovations within the Public Service; the formalizing of contractual relationships within the executive arm of government; the ‘functional’ separation of policy advice from departmental operations; the monitoring and review of executive performance, and the ‘outsourcing’ of many public goods and services. The Public Service and the State Services New Zealand, like Britain, has no firm basis in public law for its system of public administration and management. There is no constitutional definition of what ‘public employment’, as a generic category, actually comprises. However, the different types of organization that comprise the state sector are categorized in statute. The main categories are those of (1) central government, and (2) local government. The bulk of this report focuses on central government data. Within this category the public service is a distinct subset of the state services in general. Organizations which comprise the New Zealand public service have existed under the jurisdiction of the Public Service Act 1912, later the State Services Act 1962, and the State Sector Act 1988, and are currently listed in a schedule to the latter Act. They are the organizations – ministries and departments – that are subject to full ministerial direction. Traditionally in New Zealand, the state services comprised the public service, together with the air force, army, naval and police personnel, and staff employed in the postal/telecommunications agencies, the railways, the legislative department, public education services, and public hospital services (see Report of the Royal Commission

44

The state at work, 1

of Inquiry 1962). However, there have always been a number of other central government agencies, such as statutory boards and state corporations, whose staff are paid directly or indirectly from state funds. It has only been possible to use estimates of the total staff numbers in these agencies prior to 1990 (see next section). Some Methodological Issues This project provides data on employment in the New Zealand state sector during the period 1950 to 1998. Except where otherwise indicated, the figures have been generated by independent research for the current project, using official sources – Statistics New Zealand, the SSC, the appendices to the journals of the House of Representatives, Annual Government Estimates of Expenditure, and the New Zealand Official Yearbook (published annually by Statistics New Zealand). A relatively small component of the data is necessarily based on estimates rather than precisely documented information. In Tables 3.1 and 3.2, for example, the figures for the total number of state sector employees have been compiled from a range of different sources some of which provide conflicting data. The figures from 1951 through to 1981, in particular, are best regarded as estimates only, in which the numbers of state sector employees outside the state services (see previous section) have been calculated by a 5 per cent rule-of-thumb addition to the total numbers of state servants. Further, the state sector reforms begun in the 1980s have made it much more difficult to collect full data on employment numbers. This is particularly the case when it comes to measuring the decrease in the numbers of public servants that has occurred in New Zealand during the 1990s. While data on the changing numbers of employees of the New Zealand public service (public servants) are accurately recorded by the SSC, the Commission does not keep full figures on the numbers of staff employed by all the crown entities that have come into existence during this period. Because many of the functions of former government departments were transferred to single-purpose crown entities during the restructurings and therefore officially removed from the public service category, the rapidly declining numbers of public servants can give a misleading picture. The actual decline, while still radical, is not quite as dramatic as it appears to be, if measured only by the numbers of employees officially designated as public servants. The SSC collects accurate figures not only on public servants, but also on ‘selected state sector organizations’, including SOEs and crown entities established to carry out work previously done within the public service. However, because this selection is by definition not a complete category of

Government employment in New Zealand

45

non-public service agencies, this research project has attempted to provide a full picture of actual public service change between the years just prior to the reforms and 1998. To do this public service restructuring was traced to identify, as far as possible, all those state sector agencies that derive their existence from former government departments, and to determine their actual staff numbers as at 1998. The SSC provides full-time equivalent numbers (FTEs), not headcounts. However, the latter were provided for the purposes of this project by the individual agencies approached. Moreover, figures provided for the years prior to the reforms are, so far as is known, headcounts rather than FTEs. These differences inevitably result in some distortions in the total comparative figures, but it is unlikely that these would exceed 5 per cent. There is no single central agency keeping track of employment levels of all SOEs and crown entities, although the Crown Company Monitoring Advisory Unit (CCMAU), a subsidiary of the Treasury, keeps figures on many of them, including all SOEs. It was variously suggested in the course of our research that particular organizations ought to keep full records of state sector employment data. Suggested agencies included Statistics New Zealand, the Controller and Auditor General, the Inland Revenue Department, and the SSC. In the absence of any such centralized data source, the project contacted various crown entities and SOEs, and also examined their annual reports to Parliament. It found no standardized style of reporting to Parliament. All reports contain the financial information that is legally required, but reportage of staffing levels varies considerably. Government departments and ministries, and some selected crown entities are required to report staffing levels and to gather data on ethnicity and gender. SOEs and many of the other crown entities are not required to gather this type of data or to report staffing levels. Smaller crown entities often include photos and lists of staff in each annual report. Some of the larger ones provide comprehensive data on staff numbers including the gender mix of staff. Others provide no data at all. Many crown entities and SOEs do not gather data on the ethnic composition of their staff, although most (not all) compiled data on the gender mix. A number of smaller crown entities commented on the difficulties they had in recruiting staff of either Maori or Pacific Island ethnicity. The research demonstrated that the state sector reforms had seriously disrupted information gathering and institutional memory. Some organizations had been disestablished and their functions merged with other organizations. Others had been privatized in their entirety or incrementally. Data had disappeared and no one knew where they were held (if held at all). Frequently there would be no employees or at the most one employee

46

The state at work, 1

who had been with the organization throughout the period of the reforms, indicating high levels of staff turnover, with corresponding destruction of institutional memory. Generally speaking, organizations were willing to be as helpful as possible in responding to requests for information on staffing. Many were very concerned about the shortcomings in the information they were able to provide. The least helpful organizations tended to be those with commercial functions. Finally, it may be observed that one of the espoused purposes of New Zealand’s state sector reform was to increase the transparency and accountability of governmental agencies. The difficulty experienced in collecting information on such a fundamental matter as staffing numbers suggests that this aspiration may not have been entirely fulfilled.

AN OVERVIEW OF TOTAL PUBLIC EMPLOYMENT, 1950–98 From 1912, when the Public Service Act introduced a merit-based, unified, public service career system in New Zealand there was steady growth in the numbers of permanent public servants, from about 4700 to about 30 000 in 1950. By that year the total figure for all public servants – permanent, temporary and casual – was 52 000. The total number of state sector employees (including public servants) in New Zealand by this time was nearly three times that figure, and by 1961 had risen to more than 179 000 (of whom about 62 000 were public servants). New Zealand’s welfare state was consolidated and expanded greatly by the first Labour government, which held office between 1935 and 1949. From the late 1930s there was a sharp increase in public service temporary staff, until in 1947 about 75 per cent of the temporary staff were made permanent, creating the impression of a massive jump in public service employment. As shown in Tables 3.1 and 3.2, the numbers of public servants, and state sector employees generally, continued to grow steadily during the decades after the Second World War. Table 3.1 shows that while the numbers of state sector employees (excluding local government) grew steadily in the three decades from 1951, by 1998 they had fallen abruptly by around 34 per cent from 1981, to be about the same as they had been in 1961. The major factor in this decline was the post-1986 state sector reforms, particularly the public service restructuring and the massive privatization process of the late 1980s and early 1990s.

47

Government employment in New Zealand

Table 3.1

1998 1990 1981 1971 1961 1951

State sector employees (excluding local government), 1951–98 Employees

Percentage of population

Percentage of employed labour force

173 877 184 510 265 860 216 300 179 235 155 500

4.8 5.5 8.3 7.4 7.4 7.9

18.0 22.0 19.0 19.0 20.0 21.0

Table 3.2 Institutional composition of state sector employees, 1951–98 (excluding local government)

1998 1990 1981 1971 1961 1951

Total state sector employees

Non-public service employees

Percentage of non-public service employees

Percentage of permanent public servants

Percentage of total public servants

173 877 184 510 265 860 216 300 179 235 155 500

143 241 132 670 181 360 144 800 117 035 103 500

82.4 71.9 68.2 66.9 65.3 66.5

17.6 24.3 24.4 24.0 22.3 19.3

17.6 28.1 31.8 33.1 34.7 33.5

Table 3.2 gives a breakdown of the figures in Table 3.1 by way of the percentages of non-public service, permanent public service, and total public service employees in the period 1951 to 1998. It can be seen from Table 3.2 that the proportional make-up of total state employees has changed markedly in the period since the reforms of the 1980s. By 1998, public servants, all of whom were permanent employees, comprised only around 17 per cent of the total number of state employees, excluding local government, whereas since the early 1950s their proportion had been reasonably stable at around twice that figure up until the early 1980s. Table 3.2 also shows that while the number of non-public service employees increased by 75 per cent between 1951 and 1981, it dropped by 19 per cent between 1981 and 1998, largely as a result of the privatization process. The proportion of public servants among all state sector employees almost halved during the period 1981 to 1998. More on this in the next section.

48

Table 3.3

1998 1997 1996 1995 1994 1993* 1992 1991 1990 1989 1988 1987 1986 1985 1984 1980 1975 1970 1965 1960 1955 1950

The state at work, 1

Public servants, 1950–98 Permanent

Temporary and casual

Total

30 636 31 663 31 810 34 656 39 553 33 441 42 832 45 387 44 845 51 587 60 940 70 412 70 240 66 548 66 160 64 830 56 436 50 189 44 099 38 716 32 230 29 937

– – – – 600 7 513 4 791 4 418 6 995 7 406 7 660 2 005 18 267 18 875 19 578 19 533 20 359 20 441 22 871 22 795 20 661 21 988

30 636 31 663 31 810 34 656 40 153 40 954 47 623 49 805 51 840 58 993 68 600 72 417 88 507 85 423 85 738 84 363 76 795 70 630 66 970 61 511 52 891 51 925

Note: * In 1993 the SSC changed its method of data collection.

The Public Service Table 3.3 shows the numbers of permanent and temporary/casual employees in the New Zealand public service, 1950–98. There is one clear pattern discernible: a steady increase in total numbers from 1950 through to a peak in 1986. From 1987 the effects of the state sector reforms become apparent, as the numbers of public servants decline sharply and progressively through to 1998 – a reduction from 88 500 to 30 600 in the space of 12 years. This reduction resulted directly from the state sector reforms as a whole, and the contracting out of many governmental activities, departmental restructuring, and fiscal constraints. The figures also confirm that until the late 1980s, public service numbers increased steadily regardless of whether the National Party or the Labour Party held governmental office. A clearer picture of the recent decline in the

49

Government employment in New Zealand

Table 3.4 Permanent public servants by population and employed labour force, 1950–98

1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 1988 1987 1986 1985 1984 1980 1975 1970 1965 1960 1955 1950

Permanent public servants

Percentage of population

Percentage of employed labour force

30 636 31 663 31 810 34 656 39 553 33 441 42 832 45 387 44 845 51 587 60 940 70 412 70 240 66 548 66 160 64 830 56 436 50 189 44 099 38 716 32 230 29 937

0.80 0.87 0.86 0.97 1.12 0.96 1.24 1.33 1.33 1.52 1.82 2.10 2.12 2.01 2.01 2.05 1.82 1.76 1.70 1.61 1.49 1.55

1.75 1.86 1.93 2.20 2.62 2.27 2.94 3.07 3.05 3.45 3.93 4.55 4.62 4.33 4.83 4.95 4.59 4.60 4.45 4.42 4.02 4.07

size of the New Zealand public service can be seen in the proportionate figures given in Table 3.4 for permanent public servants, 1950–98. From 1988, the numbers of permanent public servants as a percentage of both the total population and of the total labour force within the country have decreased markedly and progressively, on both counts. In 1987, the year before the introduction of the State Sector Act, permanent public servants comprised 2.10 per cent of the population and 4.55 per cent of the labour force. Between 1988 and 1998 the proportions dropped progressively to 0.80 per cent and 1.75 per cent. The ‘Virtual’ Public Service Crown entities have existed as such since 1992 and were previously known as ‘crown agencies’. Generally they are the New Zealand equivalent of what

50

The state at work, 1

are popularly called ‘quangos’. They are state agencies which are neither SOEs nor full government departments subject to direct ministerial control. Existing in an ‘arm’s-length’ relationship with the political executive, crown entities usually comprise a board of directors appointed by the government, a chief executive appointed by the board and staff appointed under the authority of the chief executive. The number of such agencies grew from 164 in 1992 to 234 in 1998, though the numbers have reduced from 1995 when they ‘peaked’(at least for this period) at 243.1 As indicated above, the staff of crown entities are not members of the public service, but many of their functions were originally undertaken within the service. Thus, any accurate picture of the changing patterns of public service employment must incorporate data on staff numbers in crown entities, particularly those that were established as a direct result of the restructuring of former government departments. A truer picture of the number of ‘virtual’ public servants in New Zealand can be gained by adding to the number of official public servants, the numbers of staff employed by crown entities and SOEs established to carry out functions that were run by departments and ministries before the state sector reforms. Thus, research done for this project indicates that, at a conservative estimate, in 1998 there was a total of about 44 600 such ‘virtual’ public servants, a figure about 46 per cent higher than that for officially designated public servants.2 Notwithstanding this, the post-reform decline in the overall numbers of government employees has been dramatic. Outsourcing The figures in Table 3.3 show the large decline in the numbers of ‘in-house’ public servants, but they say nothing about the contractualized ‘outsourcing’ of work to virtual employees outside the public service. As Kelsey has observed, ‘Consultants had been used before by New Zealand government agencies, but never so many, for so much, so rapidly’ (1995, p. 143).3 Just how many external consultants are involved, how much money is spent on acquiring their services, how many of them are former ‘in-house’ public servants who left the service voluntarily or were made redundant, and what the net gains or losses have been between reducing numbers of ‘in-house’ officers and spending more on outsourcing, is unknown (and perhaps unknowable). However, some available information is indicative. The Treasury, the Department of the Prime Minister and Cabinet, and the Ministry of Maori Development are perceived to be among the biggest departmental spenders on consultants, most of it for policy advice. Whereas in 1987/88 the Treasury’s expenditure on consultants was 5.8 per cent of its total expenditure, by 1996/97 it was 21.5 per cent. In the

Government employment in New Zealand

51

intervening eight years, the proportions fluctuated markedly between a high of 70.7 per cent in 1990/91 and a low of 17.8 per cent in 1988/89. The average for the years 1989/90 to 1996/97 was 39.1 per cent. In 1998/99 the Treasury spent $NZ61.5 million on consultants, more than half of its total annual spending. (Most was spent on fees involving the sale of state assets.) In the case of the Department of the Prime Minister and Cabinet, which was established in October 1989, its expenditure on consultants averaged about 15.5 per cent of its total spending, between 1990/91 and 1992/93 (figures for later years have not yet been forthcoming). And the Ministry of Maori Development, in the years 1991/92 to 1996/97 has spent on consultants an average of 12.65 per cent of its total expenditure. A more or less random sample from other departments shows that the Ministry for the Environment spent on consultants an average 12.48 per cent of its spending between 1990/91 and 1996/97; the Ministry of Science, Research and Technology, during the same years, 7.71 per cent; but the Ministry of Transport, only 1.46 per cent. In 1999 the Child, Youth, and Family Services Agency – the government department involved in social work – spent 22 per cent of its $NZ295 million appropriation on contracting out social services to community providers. Public Sector Unionization4 Before 1988, New Zealand state sector unions were largely reliant on a tradition of ministerial recognition in securing their bargaining role, and enjoyed little legal basis in representing the interests of their members. The State Sector Act 1988 allowed what were called ‘public sector service organizations’ to register as trade unions for the first time. However, in 1991 the Employment Contracts Act (ECA), introduced by the National government, ended nearly a century of state sponsorship of trade unionism (Walsh 1997). It abolished the regulations that governed trade unions and deemed each registered union to be no different to any other nontrading body corporate. Collective bargaining was replaced by the fragmented negotiation of individual or collective employment contracts. An employment contract bound only those individuals who were party to its negotiation. Traditionally public sector unions in New Zealand enjoyed high membership levels, with membership typically including senior public servants. The state sector reforms of the late-1980s, together with the introduction of the ECA, radically changed the face of public sector unionism. The state sector reforms markedly reduced the number of state sector employees, and the pool of potential members for public sector unions. They also fragmented bargaining structures, and removed public managers from

52

Table 3.5

The state at work, 1

Public sector union membership

Unions/sectors

Approximate membership (31 March 1990)

Approximate membership (31 December 1997)

81 250 44 500 42 250 42 250

51 500 12 000 28 500 55 750

PSA Other public service Public health unions Education unions Source: Harbridge et al. 1998.

collective bargaining coverage. The ECA removed the secure institutionalized presence enjoyed by public sector unions for the brief period from 1988. The Public Service Association (PSA) remains the traditional representative of public service employees, and its coverage extends to workers in the health and education sectors, in current or former SOEs and in local government. The ECA ended official collection of data on union membership, and any such information is now gleaned through private surveys. A survey by Brosnan and Walsh (1997) has shown that over 90 per cent of public sector workplaces surveyed were unionized, compared with 13 per cent of privatesector workplaces and 19 per cent of the whole sample. And density of membership in unionized public sector workplaces was considerably higher than in unionized private workplaces. Table 3.5 provides a partial picture of trends in public sector union membership, based on unofficial surveys covering the years 1990–97. Contractual Appointments and Rewards for High Public Officials The introduction of a contractualized appointments regime in the public service from the late-1980s saw a steady rise in the numbers of public servants appointed on such contracts, from 17 per cent by 1994 to 34 per cent by 1998. For most of the twentieth century, pay rates for New Zealand’s top government officials were justified, though mainly implicitly, by a mix of what Peters and Hood (1994, p. 222) call the ‘representativeness’ and ‘alimentation’ criteria. The former, harking back to classical democracy, is based on the belief that high public officials should be generally representative in their lifestyle of ‘ordinary citizens’. The latter notion says they should be rewarded with a comfortable upper-middle-class living, in order to shield

Government employment in New Zealand

53

them from corrupting temptations and to help ensure that they are not distracted from the onerous burdens of statecraft. Probably the representativeness criterion was stronger, being consistent with the country’s dominant egalitarian norms. However, the state sector reforms have been based overwhelmingly on the third criterion presented by Peters and Hood, the ‘market’ criterion, which is reflected in the ‘developing ideology . . . that top civil servants should be paid at rates similar to those of top business executives’ (ibid.). The remuneration packages paid to chief executives of state-owned enterprises and many other crown entities are not subject to centralized control by the State Services Commissioner, and many of them are in line with the top rates paid in the private sector. The rationale for this is that state-owned enterprises must be able to compete effectively with large private corporations in the market place of executive talent. In 1997, the government abandoned its policy of trying to directly link the remuneration of departmental chief executives to that of private chief executives. The latter’s base salaries for positions broadly comparable with the average-sized public service position, rose by about 130 per cent in the nine years to 1997, compared with a rise of about 10 per cent in real terms for the salaries of departmental chief executives. Instead, the government sought relativities across the public sector as a whole. With fewer truly commercial functions left in the public service (most having been corporatized), few managers from the private sector were interested in taking up positions in the public service, especially in light of the growing gap between public and private remuneration rates in the years since 1988, and there have been hardly any departmental chief executives recruited from outside the public sector. By the late-1990s, the pay packages of departmental chief executives had not only increased very substantially in comparison with those formerly paid to permanent heads under the previous system, but had done so during a period when ordinary New Zealand employees were becoming worse off. (General public sector workers were having to be content with annual salary increases of up to only 2 per cent; and under industrial legislation, New Zealand workers in general had lost penal rates for work at nights and the weekends, access to employer-supported pension schemes, and special allowances.) In 1987 the average annual income earned by New Zealand salary and wage workers was about 25 per cent of the average salary package paid to public service permanent heads. By 1998 it was 17 per cent of this average, and continuing to trend downwards in comparison. Consistent with the trend in OECD countries (Hood and Peters 1994, p. 16) top New Zealand bureaucrats’ remuneration is no longer linked to that of politicians; all ministers are paid less than the chief executives of the

54

The state at work, 1

main government departments who report to them, and several heads of government agencies earn considerably more than the Prime Minister.5 (But it has been estimated that when private-sector rules are applied to the salary packages of New Zealand Members of Parliament, taking in all their allowances and entitlements, their total yearly remuneration increases by 142 per cent on base salary, in the case of a backbench MP; by 127 per cent for a Cabinet minister; and by 106 per cent in the case of the Prime Minister.) In New Zealand the state sector reforms introduced a pay-forperformance regime at the upper levels of governmental employment, including the public service, as the other side of the contractualist coin, in further emulation of the private sector. In 1997 the government moved to strengthen the linkages between departmental chief executives’ performance and their remuneration. It increased from 10 per cent to 15 per cent the proportion of their total remuneration package that is conditional upon performance. (Formerly the percentage was calculated only on the base salary.) At sub-chief executive levels pay for performance is devolved to individual agencies; there is no standardized process in place. Moreover, governments have been generally unwilling to provide the money needed to fund pay for performance. The upshot is that general state-service workers have not seen ‘productivity’ gains reflected in their own incomes to anything like the extent enjoyed by those at the top. The visibility of the remuneration packages paid to many top public sector executives has been diminishing. Under the former system the precise salaries of permanent heads, indeed of all public service staff, were published annually. Similar publications were available for other large components of the wider state services. These are no longer published; now only the remuneration paid to public service chief executives is published in a single source: the annual report of the State Services Commissioner. Precise figures are not provided, however, as total remuneration paid (that is, base salary and benefits in kind) is listed in $10 000 bands. Moreover, the remuneration paid to the large number of top executives of crown entities is not registered in any single publication, and the annual reports of these agencies present such information in differing, and often incomplete, form.

PUBLIC EMPLOYMENT BY POLICY AREA From the early 1970s, the high standard of living enjoyed by New Zealanders was threatened by uncertain markets for its primary produce. The main market was increasingly closed off by Britain’s entry into the European Community, and economic problems were exacerbated by the oil

Government employment in New Zealand

55

shocks, and subsequent increasing levels of inflation and unemployment. The markets for New Zealand’s primary products became more geographically diversified, and the export of manufactured goods, together with a growing tourism industry, are currently central components of the national economy. Between 1975 and 1982 there was no major growth in the New Zealand economy. Annual inflation averaged 15 per cent, and unemployment had risen to a peak of 5 per cent by 1983. A serious fiscal crisis emerged with the deficit running at 9 per cent of GDP in 1983–84. The 1984 election of the fourth Labour government heralded the advent of ‘Rogernomics’, an eponymous economic strategy guided by the Minister of Finance, Roger Douglas. It transformed the New Zealand economic landscape by deregulating financial markets, floating the New Zealand dollar, abolishing agricultural subsidies, and generally following policies based on neo-classical, or ‘monetarist’, economic principles concerned more with correcting ‘government failure’ than ‘market failure’ (Bollard and Buckle 1987; Rudd and Roper 1997). The Labour government was re-elected in 1987 (with an increased majority), but after the stock-market crash in October of that year it became increasingly riven by internal dissent over the pace and direction of its reforms. The new National government elected in 1990, led by Prime Minister Jim Bolger, was widely expected to slow up the pursuit of marketled reforms, in the face of a resounding political defeat for the Labour Party, but instead chose to deregulate the labour market, and to extend ‘economic rationalism’ into social policy through major reforms in health, tertiary education, and social welfare. The thrust of these changes was to reduce radically the extent of the state’s commitment to the promotion of individual welfare (Boston and Dalziel 1992; Kelsey 1995). The Bolger government’s Parliamentary majority was drastically reduced – to barely marginal proportions – at the 1993 election, further reflecting a broad popular desire for a slowing down of and reduction in the pace and scope of policy change (Vowles et al. 1995). While the narrowness of the governing party’s majority did slow the pace of change, during this period the reorganization of the country’s public health system along quasi-market/competitive lines was a major source of public controversy (Easton 1994; Gauld 1996). Changing Patterns of Employment across Public Service Agencies and their Equivalents, 1984–98 As discussed above, some former departments and ministries have been radically reformed through the creation of crown entities to carry out single

56

The state at work, 1

functions formerly undertaken within a larger departmental framework. Other departments have been reconstituted into more than one department, each with a separate mission.6 For example, the former Ministry of Transport, which had a staff of 4278 in 1984, by 1998 had only 63, involved mainly in policy formation. But several crown entities carried out its former functions, employing among them some 950 people, giving a combined total of only 24 per cent of the 1984 figure. Conversely the former Department of Justice in 1984 had about 4100 staff members, while in 1998 the three new departments created from the single entity (the Department of Courts, the Department of Corrections, and the Ministry of Justice) together had more than 5600. In the few cases where the departmental structure remained largely intact, staff numbers have mainly risen. The Department of Social Welfare in 1975 had nearly 3000 staff, in 1984 5358 and in 1998 about 5500. Also significant, given the fiscal concerns displayed in the post-1984 economic reforms, is the 29 per cent rise in the numbers of staff employed within the Inland Revenue Department, which has also retained its structural integrity. As for the main central departments, Treasury’s staff numbers have been reduced from 476 in 1984 to 353 in 1998. Those of the Department of the Prime Minister and Cabinet have increased during the 1990s from 55 in 1991 to 112 in 1998; but the SSC, which under the terms of the State Sector Act 1988 ceased to be a central personnel agency, saw its numbers reduce from 709 in 1984 to 134 in 1998. Table 3.6 compares 1984 staffing figures for public service policy areas with those for 1998, controlling as much as possible for structural changes within the public service during the reform period.7 (That is, staffing numbers are included for some agencies which by 1998 were officially no longer a part of the public service.) It is important to note that the figures obscure some functional transference among departments that took place during the reforms. For example, between 1984 and 1988 responsibility for enforcing the Transport Act shifted to the Police; and the Department of Labour lost some of its responsibilities to the Department of Education. Some functions included in the 1984 figures were later privatized or corporatized.8 And, as discussed earlier, the numbers of ‘in-house’ employees is only one measure of the resources committed to a policy area, and take no account of outsourcing in its various forms. The figures in Table 3.6 show that the largest staff reductions have occurred in the areas of Transport (76 per cent), Health (70 per cent), Primary Services (70 per cent), and Economic, Scientific and Industrial Services (40 per cent). By far the biggest increase has been in the area of

57

1940

Education Education

3541

1712

Housing Corporation

Health Health

7070 5358

Social security and welfare Social Security

3541

1940

Total

Classification by policy area

1984

Table 3.6 Public service policy areas, 1984 and 1998

Health Ministry of Health Health Funding Authority Health and Disability Commissioner Health Sponsorship Council

Education Ministry of Education Education Review Office Careers Service Early Childhood Development Unit Education Training and Support Agency Learning Media Ltd Special Education Service Teacher Registration Board

Social security and welfare Department of Social Welfare Commissioner for Children Housing Corporation of NZ Housing NZ Ltd Ministry of Housing

Classification by policy area

1998

479 527 41 10

557 163 101 77 230 60 1015 8

5531 8 48 637 146

1057

2211

6370

Total

58

5596 357 36 527 31 13 84 709 626 731

589 24 46 34 1789

Core government services Audit

Ombudsmen Parliamentary Services Parliamentary Counsel Commissioner for the Environment Office of the Clerk State Services Commission Statistics Internal Affairs

Foreign Affairs Crown Law

Prime Minister and External Affairs Environment Lands and Survey

1984 Total

(continued)

Classification by policy area

Table 3.6

Core government services Audit NZ Office of the Controller and Auditor General Ombudsmen Parliamentary Services Parliamentary Counsel Commissioner for the Environment Office of the Clerk State Services Commission Statistics Internal Affairs Casino Control Authority Electoral Commission Film and Literature Classification Foreign Affairs and Trade Crown Law Office Security Intelligence Service Prime Minister and Cabinet Environment Land Information NZ Antarctica NZ Asia 2000

Classification by policy area

1998

112 109 684 49 –

233 48 36 527 31 13 84 134 727 950 8 5 39 614 124

4527

Total

59

674

2776 4278

620

Defence Ministry of Defence  civilians Govt Communications Security Bureau

Transport Transport

Economic, scientific and industrial services Trade and Industry

1210 4077

Police

Customs Justice

Justice, law and order

11 978

4278

2776

5961

Commerce Research, Science and Technology Commerce Commission

Economic, scientific and industrial services

Transport Transport Civil Aviation Authority Land Transport Safety Authority Maritime Safety Authority Transfund NZ Transit NZ Transport Accident Investigation Commission

Defence Ministry of Defence (64)  civilians (2178) Govt Communications Security Bureau

Justice, law and order Corrections Courts Customs Justice Serious Fraud Office Law Commission Legal Services Board Police Complaints Authority NZ Police

34

836

63 267 387 69 41 179 11

2242 –

3668 1815 737 144 35 28 48 9 2097

7212

1017

2242

8581

60

(continued)

1984

Maori Affairs

890

5606 3120

Primary services Agriculture Forest Service

Heritage, culture and recreation

2063

2209 558 6528

Energy Labour Tourist and Publicity

Scientific and Industrial Research

Classification by policy area

Table 3.6

890

8726

Total

Heritage, culture and recreation Conservation Cultural Affairs National Library Pacific Island Affairs Maori Development Women’s Affairs Youth Affairs Te Papa (Museum of NZ) Hillary Commission NZ Sports Drug Agency

Primary services Agriculture and Forestry Fisheries Environmental Risk Management Authority

Crown Research Institutes Foundation for Research Science and Tech. NZ Trade Development Board Labour Tourism Board

Classification by policy area

1998

1482 11 404 31 329 34 24 235 39 4

2236 317 22

3875 45 165 2197 60

2593

2575

Total

61 9577

Subtotal other 57 274

205 1091 415 913 689 6264

Total 1998 public sector equivalent

9577

Other Rural Banking and Finance Corporation State Insurance Government Life Insurance Government Printing Office Valuation Works

10 142

3953

66 851

205 1091 415 913 689 6264

565

476 3477

Total 1984 core public sector

Other Public Trust Office Rural Banking and Finance Corporation State Insurance Government Life Insurance Government Printing Office Valuation Works

Finance Treasury Inland Revenue

Total public sector

Other Public Trust Office Fire Commission Earthquake Commission

Finance Treasury Inland Revenue 445 401 13

353 4479

44 076

859

4832

62

The state at work, 1

Heritage, Culture and Recreation (191 per cent), but this mainly reflects the fact that heritage work previously undertaken in the Departments of Internal Affairs and Lands and Survey was shifted to a new Department of Conservation during the reform period. Perhaps the most significant increase, both substantively and in terms of the validity of the comparison, is the 44 per cent rise in the numbers of public servants employed in the area of Justice, Law and Order. Permanent Public Servants by Policy Areas, 1960–80 It can be seen from Table 3.7 that staff numbers in the core public service departments increased by 62 per cent between 1960 and 1980, those in the economic/developmental agencies increased by 82 per cent, while in the social agencies the increase was 30 per cent. Within the economic/ developmental policy area, the highest proportion of the growth in real staff numbers – as distinct from reclassification changes apparent in transport, for example – came in agriculture/fisheries/forestry (151 per cent), and in labour (127 per cent). The increase in the labour category is largely attributable to a 73 per cent increase between 1975 and 1980, a period when full employment ended in New Zealand and the government was expanding its employment service in efforts to cope with rising numbers of unemployed. Other (Non-Public Service) State Sector Employees by Policy Areas, 1960–98 Table 3.8 shows state sector, but non-public service, employees in the policy areas categorized in Table 3.7, where relevant (many of the policy areas, of course, are staffed by public servants only). These areas are military personnel, police, post and telecommunications, railways, health (mainly hospitals), education (teachers at pre-school through to tertiary levels) and the legislative branch of government. The dramatic reductions in the post-1980 figures for post and telecommunications, and railways, are due to privatization. However, since 1960, the numbers employed in ‘core’ areas of the state sector have remained quite stable, while the total number of employees in the social policy areas of education and health increased by 130 per cent between 1960 and 1998 (notwithstanding a 10 per cent fall in the numbers of health employees between 1990 and 1998, as a result of the health sector reforms introduced in 1993).

63

Government employment in New Zealand

Table 3.7

Permanent public servants by policy areas, 1960–80

1. Core Customs Defence (civilian) Finance • Audit • Inland revenue • Public Trust Office • Treasury • Valuation Foreign affairs (inc. Prime Minister’s Department) Internal affairs Justice Police (civilian) Public Service Commission/State Services Commission Totals 2. Economic and development Agriculture, fisheries, forestry Energy • Electricity • Mines Industry, commerce and trade Labour Lands and survey Public works Scientific and industrial research Tourism and publicity Transportation Totals 3. Social policies Education Health Housing Maori affairs Social security and social services Totals

1960

1965

1970

1975

1980

622 2 752

741 1 738

768 1 924

903 2 010

1 119 2 355

215 2 162 783 314 372

189 2 298 822 344 421

179 2 134 780 431 493

218 2 879 672 403 617

278 3 342 643 467 683

201 514 1 446 94 114

294 509 1 876 141 171

345 500 2 313 225 162

496 581 3 223 411 452

635 6 450 3 889 602 702

9 589

9 544

10 254

12 865

21 165

3 548

4 756

5 831

8 891 5 841

2 725 280 312 745 1 327 4 799

3 085 271 419 729 1 441 6 077

3 849 247 415 755 1 625 6 624

8 039 26 4 641 338 612 977 1 774 7 080

670 1 688 1 842 7 025

1 039 572 836

1 246 669 2 456

1 516 731 3 244

1 814 693 3 953

2 138 646 4 411

16 183

21 149

24 837

29 947

33 152

1 451 5 257 1 165 789

1 930 6 123 1 216 861

2 300 7 021 1 189 925

1 779 3 494a 1 509 933

1 957 3 992 1 934 915

1 355

1 307

1 412

2 988

4 219

10 017

11 437

12 847

10 703

13 017

64

The state at work, 1

Table 3.7

(continued) 1960

4. Others Broadcasting Environment/conservation Printing and stationery Totals Grand

Totalsc

1965

1970

1975

1980

1 233b 504

516

621

795

30 965

1 737

516

621

795

995

37 526

42 646

48 559

54 310

68 329

Notes: a. The sharp drop in numbers between 1970 and 1975 is due to the transfer of mental health employees from the Public Service to the control of the hospital boards. b. The public broadcasting service, run as a government department, was corporatized in 1962. c. There are some minor discrepancies between some of these grand totals and the figures in Table 3.4. These results from some marginal classification choices.

Table 3.8

State sector employees by policy areas, 1960–98 1960

1970

1980

1990

1998

1. Core Defence • professional military Legislative Police (sworn)

12 771 244 2 453

13 287 253 3 119

12 641 313 4 961

11 879 457 5 095

9 462a 611 6 760

Totals

15 468

16 659

17 915

17 431

16 833

23 325

31 773

39 465

8 500

9 000

25 200

20 733

21 794

8 878

–

48 525

52 506

61 259

17 378

9 000

17 986 1 609

26 336 3 707

34 121 6 129

36 884 8 424

48 352a 10 356a

23 135

33 942

48 532

43 767

39 400

42 730

63 985

88 782

89 075

98 108

2. Economic and development Post and telecommunications Transport • railways Totals 3. Social policy Education • Schoolsb • Universities and polytechnics Health • Hospital staff (admin and health professionals) Totals Notes: a. Figures at 1997. b. Pre-school to secondary.

65

Government employment in New Zealand

Table 3.9

Women in the public service, 1950–98 Percentage of permanent staff

1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 1988 1987 1986 1985 1984 1980 1975 1970 1965 1960 1955 1950

54.2 54.0 54.7 54.1 53.0 54.0 50.0 49.9 * * 46.7 40.5 39.6 37.4 36.4 34.8 30.9 30.2 29.3 28.4 28.3 27.7

Note: * The State Services Commission has no gender data available for 1989 and 1990.

SOCIAL DIFFERENTIATION IN THE NEW ZEALAND PUBLIC SERVICE Women in the Public Service Whereas in 1950 only 28 per cent of staff of the official public service (that is, excluding crown entities that have taken over the functions of former departments and ministries) were women, by 1998 the proportion had risen to 54 per cent. As Table 3.9 shows, the increase rate for women quickened substantially from the mid-1970s, slowed down until the implementation of the state sector reforms, and then increased rapidly. There are probably several reasons for this: the implementation of Equal Employment Opportunity (EEO) policies and the reporting requirement established under the State Sector Act 1988;9 the public sector’s flexible work and

66

The state at work, 1

family-friendly policies; and the fact that the government trading organizations that were corporatized and/or privatized tended to be predominantly male in their composition (for example, Works and Development, Electricity, Railways, Mining, Oil and Gas). Also, anecdotal evidence suggests that men are looking for higher wages, available in the private sector, whereas women may be more attracted to the public sector by nonpecuniary rewards. By 1998 men formed 55 per cent of the total employed labour force in New Zealand, almost the precise converse of the position applying in the public service. A comprehensive sample of crown entities, as at 1998, shows that 45 per cent of full-time equivalent staff are women, that is nine percentage points less than for the official public service. Among crown entities, those that perform functions that have been traditionally dominated by males, such as the crown research institutes, continue to have a disproportionate number of male employees. Table 3.10, which gives a male/female breakdown by selected occupational divisions from 1967 to 1987, shows that very few women occupied positions in the upper reaches of the public service (the administrative division) prior to the state sector reforms.10 The percentage of women holding senior positions in the public service had reached 26 per cent by 1998. Time-series figures from 1950 through until the present time are not available, but as at 1999, 24 per cent of the total New Zealand labour force was employed on a part-time basis, compared with only 6 per cent of the public service staff (33 per cent in 1960), and 14 per cent of all state sector employees. However, as at 1999, women comprised the great majority of part-time workers in both the public service (86 per cent) and the state sector as a whole (87 per cent). In the labour force as a whole women filled 73 per cent of all part-time positions. Table 3.11 shows that women are over-represented in the lower salary bands and under-represented in the upper bands, although the numbers of women in the public service earning more than $NZ40 000 increased steadily during the years 1992 to 1998, and those in the lower bands decreased markedly in 1997–98. As at 1998, men in the public service were three times more likely than women to be earning more than $NZ70 000. In the same year the average base salary for women in the public service was 81 per cent of the average base salary for men. Ethnicity The numbers of Maori, the indigenous people of New Zealand (Aotearoa), have been growing at a faster rate than the non-Maori population since the

67

Government employment in New Zealand

Table 3.10

Male/female by public service divisions, 1967–87 Administrative division/class

1987 1986 1985 1984 1983 1982 1981 1980 1979 1978 1977 1976 1975 1974 1973 1972 1971 1970 1969 1968 1967

Clerical division and main succeeding classes

Typing class

Males

Females

Males

Females

Males

Females

191 132 114 107 102 99 109 110 109 114 108 105 104 89 88 86 87 85 41 42 41

7 2 1 – – – 1 – – – – – – – – – – – – – –

11 526 11 146 10 664 10 651 10 651 10 715 10 757 10 632 10 776 10 693 11 251 11 498 11 113 10 718 10 464 10 430 10 063 10 007 11 280 11 214 10 985

16 566 15 375 13 177 12 624 12 302 12 174 11 884 11 334 10 532 9 740 8 990 8 775 7 932 6 869 6 141 6 050 5 632 5 013 5 976 5 820 5 818

3 4 6 4 5 6 4 6 3 6 7 3 2 1 – 1 – 1 2 3 2

3 092 3 181 3 022 3 049 3 029 3 078 3 104 3 085 3 055 2 984 2 899 2 954 2 908 2 786 2 720 2 708 2 704 2 711 2 735 2 690 2 649

1940s, and by 1999 had reached about 15 per cent of the population.11 Maori are expected to comprise 19 per cent of the total population within the first three decades of the 21st century. The Treaty of Waitangi, signed by the Crown and many Maori tribal leaders in 1840 – but thereafter largely ignored by successive New Zealand governments – had by the 1980s become a central component in New Zealand’s constitutional framework, after Maori grievances against European colonization, especially in regard to land matters, became an increasingly pressing dimension to political life.12 For governmental purposes Maori, as the tangata whenua (indigenous people) are officially in a bicultural partnership with the Crown. New Zealand, however, is a multicultural society, in which by far the largest grouping is New Zealand European (80 per cent), with the main ethnic minority groups (other than Maori) being Pacific Islands (5.6 per cent), Asian (4.7 per cent), and Indian (1.2 per cent).13

68

Table 3.11

The state at work, 1

Salary by gender (in percentage) 1991

1992

1993

1994

1995

1996

1997

1998

Under $NZ30 000 Women Men

61 30

60 29

59 31

53.9 22.5

53.9 22.1

51.7 22.0

39.2 16.7

33.6 13.8

Total

44

46

46

39

39.3

38.3

28.9

24.5

Above $NZ40 000 Women Men

11 35

10 32

11 32

12.1 36.4

13.2 38.9

15.9 42.2

20.8 47.5

25.2 50.9

Total work force

24

20

20

23.5

25

27.8

33.1

37.1

Ethnicity in the state services generally, including the public service, came to be measured systematically only with the introduction in the 1980s of EEO policies, administered by the SSC. These policies were developed by the fourth Labour government from 1984, and the commitment to EEO was included in the State Sector Act 1988. The EEO policies are based on the ‘representative bureaucracy’ aspiration to ensure that employment in governmental agencies reflects the social composition of the population at large. The ethnicity component of these EEO policies comprises the three categories of Maori, Pacific Islands, and Asian. Table 3.12 shows that the numbers of Maori in the Public Service increased between 1993 and 1998 (before 1993 they had been declining), and that their employment rates in the public service have been higher than in the labour force as a whole. The trend is less clear-cut for Pacific Islands staff in the public service.14 The proportion of Maori (and Pacific Islands and Asian) staff in the public service earning less than $NZ30 000 per annum has been higher than that for all staff (about 33 per cent to 25 per cent). However, the proportion of Maori in the senior ranks of the public service more than doubled in the period 1987 to 1995, from 3 per cent to 7 per cent (Gregory 1995).

LOCAL GOVERNMENT EMPLOYMENT As noted above, New Zealand has had a unitary system of government since 1876, when provincial governments were abolished. Local government, which derives all its powers from Parliamentary enactments, by the late 1990s comprised a dozen regional councils which are responsible for the management of natural resources and civil defence, and some 15 city

69

Government employment in New Zealand

Table 3.12

Ethnicity composition of public service, 1993–98

Ethnic group

Public service (%)

Labour force (%) 1998

1993 1994 1995 1996 1997 1998 Employed

Maori (all) 11.0 Pacific Islands (all) 4.1 Ethnic minority (all) 4.0 Asian (all) – Other ethnic group (all) –

11.0 11.1

11.4 11.0 13.1

Total (including unemployed)

8.0

9.1

4.4

4.3

4.3

4.3

5.0

4.0

4.0

3.9 –

3.4 –

3.9 –

3.4 –

– 2.6

n.a. n.a.

n.a. n.a.

–

–

–

–

1.5

n.a.

n.a.

councils and 59 district councils administer functions and services such as roads, parks, buildings, sewerage, water and refuse collection and disposal. Local government in New Zealand accounts for about 10 per cent of total public sector expenditure. In the decades after the Second World War, governments from time to time embarked on what were, usually, incremental attempts to restructure (mainly by means of amalgamation and abolition) the large numbers of small territorial local government units and so-called ‘ad hoc’ agencies that had been created to administer single-purpose functions, such as pest and noxious weeds control. This process, carried out under the auspices of the Local Government Commission (an independent statutory body), was given considerable impetus in the 1980s and 1990s when the number of local authorities was reduced from about 700 to 73, excluding community councils and the 12 regional councils (Bush 1995). The reforms of this period also drew from the experiences of central government reorganization, particularly in the establishment of local authority trading enterprises. These represent the local government manifestation of corporatization. Table 3.13 shows the total number of employees in New Zealand local government 1956 to 1997 (the latest figures readily available), demonstrating the steady growth in employment numbers until the sharp decline brought about by the reforms of the 1980s and 1990s.15 In this, the figures show that the drastic reduction of staff in the public service and other components of the central state sector was mirrored in New Zealand local government.

70

Table 3.13

1997 1995 1990 1985 1980 1975 1970 1965 1960 1956

The state at work, 1

Local government employment, 1956–97 Employees

Percentage of population

Percentage of employed labour force

34 890 35 600 40 768 66 152 64 953 62 815 51 996 49 679 44 770 40 685

0.96 0.99 1.21 2.00 2.05 2.02 1.82 1.91 1.86 1.88

2.05 2.25 2.77 4.30 4.96 5.11 4.77 5.01 5.11 5.08

PUBLIC SERVICE PERSONNEL EXPENDITURE 1998 Table 3.14 shows the public service (rather than the wider state sector) personnel expenditure figures for 1998, by policy areas as categorized in Table 3.6. The data have been taken from the annual reports of the individual agencies, and/or have been supplied by Statistics New Zealand. There are some gaps in the data, because of various problems of information availability, and inconsistent methods of reporting in annual reports. The most significant gaps are in the categories of Social Security and Welfare, where about 11 per cent of employees have not been included; Transport, 22 per cent; and Other, 52 per cent. However, the total number of employees not accounted for represents no more than about 4 per cent of the overall employee numbers generating personnel expenditure. It should also be noted that the figures do not include teachers, or those personnel working within the health services who (like teachers) are not public servants in terms of the State Sector Act 1988.

CONCLUSION: ROLLING BACK THE STATE? There can be no doubt that New Zealand’s state sector reforms of the late1980s and 1990s have led to dramatic reductions in the numbers of people employed by the state, at both central and local levels. This reduction, which is mainly a result of the process of privatization, is apparent both in absolute terms and in relation to the total population and the national labour force. However, the financial resources expended in the new ‘outsourcing’

71

Government employment in New Zealand

Table 3.14

Public service personnel expenditure, 1998

Policy area

Social security and welfare Education Health Core government services Justice, law and order Defence Transport Economic, scientific and industrial services Primary services Heritage, culture and recreation Finance Other Total

Total expenditure ($000s)

Personnel expenditure ($000s)

Percentage of total expenditure

487 816 547 309 173 043 620 223 1 383 345 1 409 802 184 614

247 477 110 870 66 860 264 413 802 581 524 534 43 694

51 20 39 43 58 37 24

763 279 130 423

359 066 66 826

47 51

282 882 415 929 212 018

119 720 225 521 149 204

42 54 70

6 610 683

2 980 766

45

environment, the distinction between ‘in-house’ public servant and outside consultant, and the proliferation of crown entities, have together created shadowlands within which it is much harder to gauge the real as distinct from the apparent reconfiguration of the New Zealand state, in terms of employment. The New Zealand experience of radical public policy change post-1984 was virtually sui generis in its rapidity, scope and fidelity to theoretical prescription. State sector reform proceeded further, faster, more suddenly and more technocratically than in any other OECD country. Because it constituted a sharp and violent break with prevailing political tradition and culture it cannot be explained adequately by reference to existing theories of changing patterns of public employment, but perhaps needs to be considered in its own peculiar terms. Wildavsky’s (1985) argument that a country’s political culture best explains the size of the state, and that cultural change precedes and dominates budgetary change appears to fit the New Zealand situation quite well, at least during the ‘nation-building’ period during the middle decades of the twentieth century. The egalitarian ethos and general political consensus that had underpinned New Zealand’s liberal welfare state during the post-Depression decades, and which, as Wildavsky would have predicted,

72

The state at work, 1

saw the inexorable growth of public expenditure, had come under serious strain by the 1970s. Greater political conflict began to emerge in the face of a poorly performing economy with high levels of inflation and unemployment. From 1975 to 1984 politically pragmatic ‘Muldoonism’, while generally consistent with the political culture of New Zealand, held at bay the anti-statist, pro-market, aspirations of an emerging and cohesive policy elite. With Muldoon’s electoral defeat this powerful minority force seized the opportunity to outrun the political and cultural inertia that might have been expected to delay the shift to a more competitive, less egalitarian ethos. The new Labour government headed by David Lange put a rocket under the process of gradual cultural change; and to the present day, wider sections of New Zealand society are still struggling to catch up with the reconfigured state that has resulted.

NOTES *

1. 2. 3.

4.

The author gratefully acknowledges the assistance provided by Malcolm Macaskill and Mark Skilton, of the State Services Commission, and variously by research assistants Nick Grove, Graham Howell, Fiona McDonald, Pauline Ngan, Tamika Simpson and Mark Thornton. In addition, by 1998 there existed in New Zealand, 2664 primary and secondary school boards of trustees, which are also officially designated as crown entities. The figure of 44 600 does not include the board members of crown entities, who are appointed by the government of the day. In the context discussed here there would be a total of 150 to 200 board members. Other measures are, of course, available to gauge the extent to which the state in New Zealand has been ‘rolled back’, but these are tangential to the main focus of the current project. In the mid-1980s, when the state sector reforms were begun, government expenditure was 39 per cent of GDP. Eleven years later, in 1996/97, it had fallen to 34.4 per cent of GDP. In the decade 1970 to 1979 it averaged 33.13 per cent; from 1980 to 1989, 38.64 per cent; and in the seven years from 1991 to 1997, 37.94 per cent. These figures are for government expenditure as a whole, not just for the public service per se; but they indicate that ‘down-sizing’ the public service has not been able to reduce government spending to anything like the levels that pertained before the state sector reforms of the 1980s. Against this, expenditure on what a former Minister of State Services called ‘core government’ – ‘activities generally associated with a nation state such as foreign affairs, a justice system and basic regulation of the economy’ (Upton 1999) – was reduced in real terms by 2.5 per cent per annum between 1993 and 1997, or as a share of total government spending it fell from a high of 35 per cent in 1975/76 to 22 per cent in 1993/94. It should also be noted that in all three core areas of welfare spending – social welfare, health, and education – expenditure as a proportion of GDP has increased since the mid1980s. Expenditure on public education averaged 5.17 per cent of GDP in the decade 1976 to 1985, and has averaged 5.51 per cent since then. The figures for health expenditure are 5.43 per cent and 5.53 per cent, respectively; and for social welfare, 10.2 per cent and 13.17 per cent. See too the section on employment by policy area. This section draws heavily on work done by the Industrial Relations Centre at Victoria University of Wellington. The data in Table 3.5 are from Harbridge, Walsh and Crawford (1998) of the Centre.

Government employment in New Zealand

73

5.

Salaries for ministers of the Crown and members of parliament are set by an independent authority, the Higher Salaries Commission. 6. Through 1950 to 1998 the number of public service departments fluctuated only marginally, from a ‘peak’ of 47 in 1950 to 39 in 1998. In the decade 1950 to 1960 the average number in existence was 44; 1961–70, 41; 1971–80, 39; 1981–90, 39; 1991–98, 38. 7. The policy areas are based on those used in the New Zealand Government’s Budget Economic and Fiscal Update (The Treasury 1998). 8. Between 1988 and 1999, successive New Zealand governments privatized a wide range of state assets for a total price of about $NZ19 billion (a figure equal to around 20 per cent of GDP). The assets sold included Telecom, NZ Rail, the Forestry Corporation, the Bank of New Zealand, State Insurance, the Rural Bank, Air New Zealand, Post Bank, and Petrocorp. 9. However, there is no central collection point for EEO data on the state sector as a whole. See first section. 10. Source: Henderson (1990). The divisions/classes were abolished under the terms of the State Sector Act 1988. 11. In New Zealand there are large numbers of people of mixed ethnicity. For census and other official purposes ethnic identity is determined only by self-declaration. The main ethnic categories identified in New Zealand for official purposes are: NZ European, Maori, Pacific Islands, Asian, and Indian. People may identify as more than one ethnicity. The numbers of those who identify only as Maori have been declining, and those who identify as Maori and at least one other ethnicity have been increasing. The latter group is officially recognized as the total Maori population. 12. Under the Treaty, Maori ceded sovereignty to the Crown as subjects, and the Crown guaranteed to Maori possession of their lands, forests, fisheries and other properties, but with the Crown having an exclusive right of preemption over any land that Maori might determine to sell. 13. As at the 1996 official census. The percentages sum to more than 100 per cent because of the numbers of people identifying with more than one ethnicity. 14. The figures are available for 75 per cent of the public service only, and include some double counting of those who identify as both Maori and Pacific Islands. 15. While these figures show trends clearly, the numbers for 1995 and 1997 are fulltime equivalents rather than head counts, and so may tend to overstate the total reduction.

REFERENCES Bollard, Alan and Robert Buckle (eds) (1987), Economic Liberalisation in New Zealand, Wellington: Allen and Unwin/Port Nicholson Press. Boston, Jonathan and Paul Dalziel (eds) (1992), The Decent Society? Essays in Response to National’s Economic and Social Policies, Auckland: Oxford University Press. Boston, Jonathan, John Martin, June Pallot and Pat Walsh (1996), Public Management: The New Zealand Model, Auckland: Oxford University Press. Brosnan, Peter and Pat Walsh (1997), ‘Why are New Zealand unions stronger at the workplace under the Employment Contracts Act than Australian unions under the Accord?’, in Tom Bramble, Bill Harley, Richard Hall and Gillian Whitehouse (eds), Current Research in Industrial Relations: Proceedings of the 11th AIRAANZ Conference, Brisbane, pp. 78–88. Bush, Graham (1995), Local Government and Politics in New Zealand, Auckland: Auckland University Press.

74

The state at work, 1

Duncan, Ian and Alan Bollard (1992), Corporatization and Privatization: Lessons from New Zealand, Auckland: Oxford University Press. Easton, Brian (1994), ‘How did the health reforms blitzkrieg fail?’, Political Science, 46, 215–33. Gauld, Robin (1996), ‘Revolution or evolution in health sector restructuring? The experiences of New Zealand and Hong Kong’, Hong Kong Public Administration, 5, 87–103. Gregory, Robert (1995), ‘Post-reform attitudes of New Zealand’s senior public servants: A follow-up study’, Political Science, 47, 161–90. Harbridge, Raymond, Pat Walsh and Aaron Crawford (1998), ‘The restructuring of employment relations in the New Zealand public service’, paper presented at Developing Competitiveness and Social Justice: The Interplay Between Institutions and Social Partners, 11th World Congress of the International Industrial Relations Association, Bologna, September. Henderson, Alan (1990), The Quest for Efficiency: The Origins of the State Services Commission, Wellington: State Service Commission. Hood, Christopher and B. Guy Peters (1994), ‘Understanding RHPOs’, in C. Hood and B.G. Peters (eds), Rewards at the Top: A Comparative Study of High Public Office, London: Sage, pp. 1–24. Kelsey, Jane (1995), The New Zealand Experiment: A World Model for Structural Adjustment?, Auckland: Auckland University Press/Bridget Williams Books. Lipson, Leslie (1948), The Politics of Equality: New Zealand’s Adventures in Democracy, Chicago: University of Chicago Press. Peters, B. Guy and Christopher Hood (1994), ‘Conclusion: what have we learned?’ in C. Hood and B.G. Peters (eds), Rewards at the Top: A Comparative Study of High Public Office, London: Sage, pp. 215–27. Report of the Royal Commission of Inquiry (1962), The State Services in New Zealand, Wellington: Government Printer. Roper, Brian (1997), ‘The changing class structure’, in Chris Rudd and Brian Roper (eds), The Political Economy of New Zealand, Auckland: Oxford University Press, pp. 79–99. Rudd, Chris (1997), ‘The welfare state’, in Chris Rudd and Brian Roper (eds), The Political Economy of New Zealand, Auckland: Oxford University Press, pp. 237–55. Rudd, Chris and Brian Roper (eds) (1997), The Political Economy of New Zealand, Auckland: Oxford University Press. Schick, Allen (1996), The Spirit of Reform: Managing the New Zealand State Sector in a Time of Change (A Report Prepared for the State Services Commission and the Treasury, New Zealand), Wellington: State Services Commission. Schick, Allen (1998), ‘Why most developing countries should not try New Zealand’s reforms’, World Bank Research Observer, 13, 123–31. Scott, Graham (1995), Government Reform in New Zealand, Washington, DC: International Monetary Fund. Scott, Graham, Peter Bushnell, and Nikitin Sallee (1990), ‘Reform of the core public sector: the New Zealand experience’, Governance, 3, 138–67. The Treasury (1998), Budget Economic and Fiscal Update, Wellington: The Treasury. Upton, Simon (1999), ‘The role of the state’, IPS Policy Newsletter, Institute of Policy Studies: Victoria University of Wellington, 56, February, pp. 8–15.

Government employment in New Zealand

75

Vowles, Jack, Peter Aimer, Helena Catt, Jim Lamare and Raymond Miller (1995), Towards Consensus? The 1993 General Election in New Zealand and the Transition to Proportional Representation, Auckland: Auckland University Press. Walsh, Pat (1997), ‘From arbitration to bargaining: changing state strategies in industrial relations’, in Chris Rudd and Brian Roper (eds), The Political Economy of New Zealand, Auckland: Oxford University Press, pp. 183–201. Wildavsky, Aaron (1985), ‘The logic of public sector growth’, in Jan-Erik Lane (ed.), State and Market: The Politics of the Public and the Private, London: Sage, pp. 231–70.

4.

Public employment in Australia: in competition with the market Helen Nelson

The Australian system of government is a hybrid, born out of a federal system adapted from the USA and a parliamentary system inherited from the British colonial period. Time has produced local variations in both federal and cabinet system practice, but the inborn incompatibilities remain, as became evident in 1975 when a constitutional crisis highlighted the anomaly of a Senate that has the dual role of being an upper house in both a federal system and a system of responsible government.1 The Australian public services at Commonwealth (federal) and state levels have their origins in that British legacy. The services that evolved out of the colonial period and into the twentieth century comprised bureaucracies that, in the main, conformed in both structure and processes to the Westminster model. Broadly, the public bureaucracies developed around the notion of a career service, neutral in its politics and dedicated to providing ‘frank and fearless’ advice to the government of the day, regardless of its political make-up. Independent public service boards oversaw services in which: recruitment and promotion procedures were by merit and seniority, free of political influence; legislative provisions provided protection against arbitrary dismissal; regulations discouraged the recruitment of ‘strangers’ (outsiders) to positions above the base grade; a distinctive retirement and pension system was in place; and heads of department held the title ‘Permanent Head’. Since the 1970s, a protracted reform programme has reshaped the public sector. The very concept and nature of the public service has been redefined. The reformation has yielded changes in the size and composition, deployment of resources, procedures and priorities of the Australian public services at federal, state and local levels of government. The reform process has proceeded uninterrupted under both Labor and LiberalNational Party Coalition governments, and is on-going. The trigger for change came in the early 1970s with the electoral defeat of long-serving Conservative governments at the Commonwealth and some state levels and the installation of reformist Labor governments. The initial 76

Public employment in Australia

77

focus for reform was at the political executive level, including the upgrading and streamlining of cabinet procedures and provisions for strengthening the hand of ministers in relations with their departments, for instance, with the expansion of ministerial staffs and allowance for some ministerial input into department senior appointments. A further reform phase, commencing in the 1970s, targeted the secrecy typical of Westminster-style bureaucracies and opened up the public services to closer public scrutiny. Reforms associated with the ‘new administrative law’ included establishment of Ombudsmen’s offices, review and revision of the system of administrative appeals, and the introduction of freedom of information legislation. The emphasis in this chapter, however, is on the impact of what might be identified as a third phase in the reform process, specifically, a managerialist-inspired set of changes designed to bring public management more into line with private sector practice. Commencing in the 1980s with the re-election of a federal Labor government, public sector reform became intertwined with a broader agenda of economic reform, geared to meeting the challenges posed by globalization and international trade liberalization. By the 1990s, the agenda had extended further to include workplace relations reforms that sought to establish for the public sector an employment framework that was indistinguishable from that of the private sector. The movement gained further impetus with the return to office of a Coalition government in 1996 and now (2002) appears to be consolidated into a programme of reform that denotes radical revision of the old model – albeit political rhetoric continues the convenience of reference to a ‘Westminster model’. In brief summary, the original emphasis on managerial reform has broadened and evolved into a market-based reform programme. The programme is informed by a view that challenges the notion of a professional career service with a concept of the public service as ‘a business that is judged by its results’ (Halligan 2000, p. 53). While the full content and implications of the redirection are still evolving, the traditional distinctions between the public and private sectors have blurred. Major aspects of the reform programme have included review of the ‘core’ functions of government, utilization of a range of privatization mechanisms, the application of contestability to the delivery of services, a switch in primary focus from inputs to outputs and outcomes, and deregulation of the system of personnel management. The character and composition of the Australian public services are changing accordingly. The on-going programme of extensive cutbacks in employment and services has created leaner bureaucracies, most particularly at the federal level. Privatization programmes have reduced further the

78

The state at work, 1

scope of the public sector, in terms both of the size of the workforce and the range of government functions. The previous hallmark emphasis on continuity of employment and accumulated experience has been all but lost in face of widespread use of outsourcing and contract work. The public service has lost its place as the primary source of policy advice to governments and now competes against a range of new actors. The extensive use of the private and voluntary sectors for delivery of services has promoted further devolution and decentralization of decision-making, and sharpened debate about lines of accountability. The shift to a focus on outputs and outcomes has stimulated a series of radical changes in budgetary practices and financial management generally. Outcome concerns have given client needs more overt prominence. With a deregulated personnel system, the terms and conditions of public sector employment now conform to a market-based model that favours limited-term contracts and performancebased pay. The emerging model is one in which the public sector is more ‘an adjunct to the private sector’ (Halligan 2000, p. 60). Albeit, at federal level, the government has legislated a set of public service ‘values’, there appears to be little left of the notion of a unified ‘career service’: ‘agencies and departments are expected to operate as business-like entities, whose fortunes will be shaped by the way in which they perform in a contestable environment’ (Stewart 1998, p. 1). While the model is still evolving and debate continues about its likely final shape and impact, the following account traces developments in the size, composition and terms and conditions of public sector employment, and the distribution of staffing resources across policy areas.

PUBLIC SECTOR EMPLOYMENT LEVELS, 1954–2000 Australians appear to have an ambivalent attitude towards the public service. The contrariety might be traceable to the origins of the country as a penal colony, established when the convict transports carrying the first white settlers arrived from Britain in 1788. On the one hand, the early reliance on government to provide basic utilities and services, most notably, transportation and communications systems, extended also to expectation of the bureaucracy as a source of subsidy and substantive backup support in time of need. The attitude has evolved so that Australians are said to hold a utilitarian approach to government (Hancock 1930). The extent to which this outlook shapes community tolerance of the contemporary trend towards limited government is yet to become clear. Its lingering potency might be seen in the continuing controversy surrounding current government attempts to wholly privatize Telstra, the 51 per cent government-owned

Public employment in Australia

79

communications company, and in the hostility aroused by withdrawal of government services, particularly in rural areas. The willing reliance on government as the ultimate provider is paralleled by the poor image suffered by the public sector. It is illustrated by common resort to the term ‘Fat Cats’, an often-mocking public attitude towards authority and frequently unfavourable comparisons with the private sector. The call for an injection into the public service of people with ‘business experience’ goes back at least to the immediate post-First World War period and employment in the public sector has been seen generally as second-best to a career in the private sector (Spann 1979, pp. 318–20). The ministerial statements accompanying the current round of reforms commonly voice an apparent lack of confidence in the capacity of the public sector to deliver services and a belief in the private sector as the appropriate guide to ‘best practice’. Growth and Decline At 7.6 per cent of the population and 15.2 per cent of the labour force, public sector employment in Australia in the year 2000 ran at a level not dissimilar to that recorded for other countries included in the present study. The Australian figures, however, are indicative of a sharp change in direction. At their peak in 1986, the equivalent figures were 10.7 and 23.4 per cent respectively (see Table 4.1). The period of most rapid growth occurred in the 1960s, when the level of public sector employment increased by almost half again over the course of the decade. As noted below, although the expansion occurred at all three levels of government, it was most evident at the federal level. The Commonwealth government workforce had expanded considerably during the period of the Second World War, when it exercised its emergency powers to take over a range of functions carried out previously by state governments, for instance housing and transport. In 1946 the Commonwealth was successful in a referendum that established its jurisdiction in the social security area.2 Although a further attempt in 1948 to expand its limited constitutional powers by referendum failed, the Commonwealth used its concurrent powers to continue and extend its scope, most particularly in immigration and transport. The Commonwealth moved into education in 1966 when it established the Department of Education and Science and made the first of the conditional grants that were to become the instrument by which it would achieve dominance in the area. The emergence of newly-politicized movements and new issues, for instance the environment and other quality of life issues, and the

80

The state at work, 1

Table 4.1 Public employment relative to population and labour force, and government expenditure as a percentage of GDP, 1954–2000 Total government employmenta (000s)

Percentage of populationb

Percentage of labour force

Total outlays of government as a % of GDP

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1986 1981

1468.4 1458.9 1464.8 1490.6 1517.1 1572.6 1578.0 1658.0 1684.8 1768.3 1741.5 1750.0 1549.3

7.6 7.7 7.9 8.0 8.3 8.7 8.8 9.4 9.6 10.2 10.2 10.7 10.4

15.2 15.5 16.2 16.1 16.7 17.7 18.1 19.3 19.8 20.8 20.9 23.4 22.9

33.3c

1976 1971

1433.2 1203.2

10.2 9.2

22.9 21.5

1966

946.9

8.2

19.3

37.8 38.4 39.1 39.6 37.7 38.8 33.6d 33.6

24.4e 1961 1954

824.1 695.9

Percentage change 1991–2000 1981–91 1971–81 1961–71 1954–61 1954–2000 1954–91

7.8 7.7

19.5 18.8

16.9 14.1 28.8 46.0 18.4

111.0 154.1

Notes: a. Excludes members of permanent defence forces and employees of overseas embassies, consulates etc. b. Population figures prior to 1966 exclude full-blood Aborigines. c. Estimate. d. Average 1974–79. e. Average 1960–73. Source: ABS, Wage and Salary Earners (Catalogue No. 6248.0); Year Book Australia; OECD (1997); Aulich et al. (2001, p. 8).

Public employment in Australia

81

consequently expanded political agenda of the 1970s, had an impact at all levels of government with employment figures increasing overall by more than a quarter. The expansion of government activity had the added impetus of the election of reformist Labor governments at the Commonwealth and some state levels. New portfolios such as Environment, Women’s Affairs, Aboriginal Affairs, Sport, Recreation and the Arts were created at federal and state levels. Since 1991 the trend has reversed to give way to a pattern of withdrawal. While the cutbacks are evident in ‘core’ government areas, their main impact has been in the shedding of public trading and financial enterprises (see further, below). In summary, during the period 1991–2000 the size of the public sector overall declined by 16.9 per cent; public sector employment as a percentage of the population fell from 10.2 to 7.6 per cent; and, as a percentage of the labour force, fell from 20.8 to 15.2 per cent (Table 4.1). public sector spending has followed the same pattern: growth throughout the latter part of the twentieth century, until the 1990s, when concerns about the efficacy of ‘big government’ saw a turnabout. After peaking at 39.6 per cent in 1992, the estimated total outlays of government as a percentage of GDP fell to 33.3 per cent in the year 2000 (Table 4.1).

EMPLOYMENT BY LEVEL OF GOVERNMENT The constitutional division of powers allocates the day-to-day labourintensive services, such as education, health and transport, to the states. Consequently, the bulk of public sector employment is located at state level. The pattern of recent cutbacks has reinforced this concentration. The decline in the level of public sector employment has been common across all three levels of government. The most dramatic fall has been at the federal level, with a drop of 45.1 per cent during the period 1991–2000. State-government-level employment during the same period declined by 6.7 per cent. The equivalent local government figure is 12.4 per cent (see Table 4.2). The outcome is a further concentration of public sector employment at state level. During the period 1991–2000, state government employment rose from 65.5 to 73.6 per cent of total public sector employment. The federal government share of the total declined accordingly, from 25.4 to 16.8 per cent (see Table 4.2) The changing balance reflects changes in Commonwealth-state administrative relations. Under a process of review that began in 1991 with a round of Special Premiers’ Conferences, intergovernmental arrangements across a range of policy areas have been re-ordered so that, increasingly, the

82

Table 4.2

The state at work, 1

Federal, state and local government employment, 1971–2000 Total Federal Federal government (000s) as % of employment total (000s)

State (000s)

State as Local Local % of (000s) as % of total total

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990

1468.4 1458.9 1464.8 1490.6 1517.1 1572.5 1577.9 1658.0 1684.8 1768.3 1741.5

246.5 241.7 255.2 282.6 302.5 339.2 350.9 370.3 388.5 448.8 400.7

16.8 16.6 17.4 19.0 19.9 21.6 22.2 22.3 23.1 25.4 23.0

1080.8 1080.2 1069.7 1066.0 1068.7 1077.8 1068.1 1126.8 1136.0 1158.5 1180.6

73.6 74.0 73.0 71.5 70.4 68.5 67.7 68.0 67.4 65.5 67.8

141.1 137.0 139.9 142.0 145.9 155.5 158.9 160.9 160.3 161.0 160.2

9.6 9.4 9.6 9.5 9.6 9.9 10.1 9.7 9.5 9.1 9.2

1985

1708.1

434.3

25.4

1119.2

65.5

154.7

9.1

1976

1433.2

391.0

27.3

920.9

64.3

121.3

8.5

1971

1203.2

339.2

28.2

756.5

62.9

107.4

8.9

Percentage change 1991–2000 16.9 1971–90 44.7

45.1 18.1

6.7 56.1

12.4 49.2

Source: ABS, Wage and Salary Earners (6248.0); Year Book Australia.

Commonwealth has taken primary responsibility for funding and the setting of policy guidelines, meanwhile ceding increased administrative responsibilities to the states. Commonwealth government domination of the public policy agenda has evolved over the decades of federation, most particularly since the 1970s. It represents a reversal of the intentions of the early constitution-makers. When the six Australian states federated in 1901, they sought to establish the Commonwealth government as the weaker partner. The Constitution provided for a limited range of powers at the national level, with the residual powers allocated to the states. A record of electoral resistance to constitutional change through referenda has seen the original constitutional agreement remain largely unchanged. Formally the states retain their constitutional superiority. The Commonwealth has achieved its dominant position through the exercise of its financial powers and, in more recent years, a more liberal judicial interpretation of the constitutional provisions.

Public employment in Australia

83

The Australian federal system has a higher degree of vertical fiscal imbalance than any other federation. Under the Constitution, taxation is a concurrent power. The federal government has held the monopoly over income tax revenue since 1942 when it used its war-time emergency powers to gain the upper hand. The move has since survived two High Court challenges. The upshot is a system of federal financial arrangements that defies principles of financial accountability, specifically, that the government that raises the revenue should be also the government that spends it. The federal financial arrangements that have evolved since 1942 have witnessed the Commonwealth using the power of the purse to gain entry into, and then policy control over, a range of policy areas regarded once as state government domains. The main lever applied by the Commonwealth has been the use of conditional grants. Over time, and particularly since the 1970s, conditional grants have both increased as a proportion of financial transfers and become more specific in their conditions. Currently conditional grants comprise over half of total Commonwealth transfers to other levels of government. The Commonwealth has formally taken over responsibility for tertiary education, which it now funds directly. Under present arrangements, primary and secondary education, health, and a range of other responsibilities for which the states retain constitutional responsibility are, in practice, funded predominantly through Commonwealth transfers, with the Commonwealth exercising policy control and the states retaining administrative responsibility. Local government is the poor partner in the Australian federal system. It is a state-level constitutional responsibility. Increasingly local government has come to be funded directly by the Commonwealth, but its form of organization, its electoral arrangements and its functional allocations remain dependent upon state government legislation and ordinances. Although each state has its own local government system, the similarities are greater than the differences. There are over 850 local government councils, municipalities and shires scattered across the nation. Historically, state governments have not promoted expansion at the local level. Services that require large expenditures or large-scale provision or regulation tend to be managed by state governments themselves. Local government growth has been hindered further by not infrequent boundary changes, inspired usually by ruling state-level political party ambitions. Although direct federal transfers have boosted funding and extended its range of services, local government is viewed still as confined mainly to ‘roads, rates and rubbish’. During the period 1991–98, local-level employment increased from 9.1 to 9.6 per cent of total public sector employment, a figure considerably lower than that recorded for other federations included in this study.

84

The state at work, 1

Impact of Privatization on public sector Employment Whereas cutbacks in public sector employment have occurred across the board, the major cutbacks have arisen from a determined programme of privatization. The programme has been pursued by both Labor and nonLabor governments, at federal, state and local levels of government, and has been most evident since the election of the federal Coalition government in 1996: The privatization agenda has been so dominant in Australia that since 1996 the principal driver of the mergers and acquisitions market has been divestment of publicly-owned organizations. In 1997 divestment of state and national publicly owned assets constituted 13 per cent of the total global market, placing Australia second only to Brazil. Three-quarters of this involved infrastructure assets, principally in the electricity, gas, airports and telecommunications sectors. (Aulich 2000, p. 162)

Much of the impetus and justification for the divestment of public assets and termination of government services has been based on revised notions of what constitutes the ‘core’ functions of government and/or the argument that the private sector could deliver alternative supply. Various divestment strategies have been followed, including, particularly at the Commonwealth level, the conversion of many statutory bodies into an incorporated company structure. The approach makes the ownership status of these entities difficult to locate and there is now no reliable, complete inventory of government-owned, or partially-owned, bodies. Most of the enterprises so corporatized have two ‘shareholder’ ministers representing the government’s and the taxpayers’ interests, usually the portfolio minister and the Minister for Finance. The company structure, however, has relieved these bodies of the reporting requirements usually associated with governmentowned entities and, from publicity given recently to certain disputes regarding payment claims from contractors, the ministers concerned are under no pressure to become involved or be in any way answerable for the manner in which these bodies conduct themselves (Tingle 2001; Wettenhall 2000). public sector employment figures shed some light on the impact of privatizations on the size of government workforces, particularly at the Commonwealth level. Table 4.3 distinguishes between ‘general government’ employees, that is, generally, those attached to government departments, and employees in public trading and financial enterprises. The former group constitute the ‘public service’. The latter group are government employees, mostly employed under terms and conditions specific to their particular organization. The latter group has been the main target for cutback.

85

Public employment in Australia

Table 4.3 General government and statutory body employment, selected years General Public Public Total General Public Public government trading financial government trading financial (000s) bodies bodies (%) bodies bodies (000s) (000s) (%) (%) Federal government 1998 145.2 1996 173.1 1995 174.7 1990 190.8

107.6 128.3 123.1 170.3

2.4 1.2 41.4 39.6

255.2 302.6 339.2 400.7

Percentage change 1990–98 23.9

36.8

93.9

36.3

State government 1998 971.0 1995 955.6 1990 961.1

94.4 105.7 183.6

4.3 16.5 38.2

1069.7 1077.8 1182.9

Percentage change 1990–98 1.0

48.6

88.7

9.6

Local government 1998 133.7 1995 143.7 1990 137.5

6.1 11.8 22.7

– – –

139.8 155.5 160.2

Percentage change 1990–98 2.8

73.1

–

12.7

56.9 57.2 51.5 47.6

42.2 42.4 36.3 42.5

0.9 0.4 12.2 9.9

90.8 88.7 81.2

8.8 9.8 15.5

0.4 1.5 3.2

95.6 92.4 85.8

4.4 7.6 14.2

n.a. n.a. n.a.

Source: Data supplied by ABS.

In 1990 the Commonwealth government comprised 47.6 per cent ‘general government’ employees and 52.4 per cent public trading and financial enterprise employees. By 1998 a reversal had taken place, with the relevant ratios being 56.9 and 43.1 per cent respectively (see Table 4.3). The most savage cuts of the period were to the public financial bodies, where the number of employees was cut by 93.9 per cent. As shown in Table 4.3, the deepest incisions occurred in 1996, the first year in office of the Liberal– National Party Coalition government. Similar changes have occurred at the state and local government levels, although in each case, the ‘general government’ component has been traditionally the larger component. At the state level, the public trading and financial enterprises component was cut from 18.7 per cent in 1990 to 9.2

86

The state at work, 1

per cent of total employment in 1998. Again the public financial institutions witnessed the largest withdrawal, with cuts of 88.7 per cent. In the same period, public trading enterprise employment numbers declined by 48.6 per cent. As recorded in Table 4.3, state ‘general government’ employment actually increased slightly, rising by 1.0 per cent. Indicative of local government’s limited status, corporate financial bodies have not operated at that level. In 1990 public trading enterprises accounted for 14.2 per cent of employment at local government level. By 1998, following cutbacks of 73.1 per cent, trading enterprises accounted for only 4.4 per cent of total employment. Contracting Out: Choice and Contestability Privatization has followed other forms besides divestment and withdrawal. Two privatization mechanisms that have had widespread impact on the delivery of government services concern contracting out or outsourcing, wherein private-sector providers replace delivery by public sector organizations, and the alternative approach of opening up the public provision of services to market competition. In the case of the latter, government agencies are required to show that they are competitive in terms of costs and quality of delivery. Both approaches entail some continued government involvement, commonly in terms of decision-making powers and resource input. What criteria determine the appropriate degree of government involvement? According to Halligan (2000, p. 52): The scope of the public sector is defined in three simple ways: by the ‘Yellow Pages’ Test (if a supplier is listed in the telephone book supplement, the public sector role should be questioned), the emphasis on choice, and contestability. The provision of policy advice by departments is also contestable. Choice for consumers and purchasers was a related element.

Strangely, for a reform programme that emphasizes a focus on outputs and outcomes, there appears to have been no systematic attempt to ensure that formal evaluation processes are in place. Data regarding the costs and impact of the changes on service delivery are therefore not available. The changes are also, perhaps, too recent to allow detailed assessment. Outsourcing, however, is now common practice in the management of taxation, health, social security and other records. In January 2001 the Commonwealth government did a turnabout on its earlier decision to outsource information technology (IT) services. Acting contrary to advice from the agencies concerned, the Commonwealth had pursued a policy of outsourcing the supply of all government computer infrastructure, from

Public employment in Australia

87

mainframes to desktops. Savings to the order of $A1 billion in three years were promised (Aulich 2000, pp. 168–9). A subsequent report from a commissioned review exposed a saga of management and financial failures in which ‘implementation risks have been magnified and the management of those risks subjugated in the pursuit of a contracted outcome’ (cited in Kingston 2001). The report stressed the risks associated with the scheme overall, particularly in the delivery of payments to welfare recipients, and queried its cost-effectiveness and its potential to set in train a brain drain from the public sector. Costs of the scheme had included $A15.2 million paid to legal consultants appointed without tender. Contestability opens up the delivery of government services to market competition from private and voluntary organizations. In order to stay in the field, government agencies must demonstrate that they are competitive in terms of cost and quality. The Commonwealth Employment Service has now been replaced by non-government agencies that deliver services to the unemployed. In a similar privatization development, Centrelink has been established as a one-stop-shop delivering government services to 6.1 million ‘customers’. Government departments purchase services from Centrelink. That is, departments, acting on behalf of the government, contract Centrelink to implement policies. In the year 2000 Centrelink serviced a range of departments, including: on behalf of the Department of Family and Community Services, it provided social security payments to millions of Australians; for the Department of Employment, Workplace Relations and Small Business it handled unemployment registrations; for the Department of Education, Training and Youth Affairs it provided assistance for isolated children; for Veterans’ Affairs it delivered services to veterans; for Transport and Regional Services it administered the Tasmanian Freight Equalization Scheme; for Agriculture, Fisheries and Forestry it administered the Farm Family Restart Scheme and Exceptional Circumstances Relief Payments. The Chief Executive Officer of Centrelink could claim that by June 2000 her organization had delivered a three-year dividend of $A361 million to the Australian taxpayer and was on target to deliver $A1 billion at the end of six years (Vardon 2000, pp. 4–5). One of the more tantalizing examples of contracting-out, mentioned by Aulich, concerns the granting of a contract to the Australian Capital Territory government to provide corporate services (including human resources, office services, payroll, public relations functions) to the Commonwealth government Department of Finance and Administration. As Aulich observes, ‘clearly, each level of government had different views about what constituted its “core” business’ (2000, p. 170). The use of consultants is a long-established practice. The direction of the current reform programme has encouraged increased usage. Government

88

The state at work, 1

Purchasing Guideline No. 13 dictates that all agencies must report on their use of external consultants, including information regarding the number of consultants employed, their purpose and cost, and the justification for employing them. In practice there appears to be no standard form of reportage. Nor are collated data available. A selected sample of annual reports reveals that in 1997/98 the Department of Health and Family Services employed 381 consultants, at a cost of $A23.9 million; the Department of Defence utilized the services of 252 consultants ($A9.9 million); the Department of Communications and the Arts, 289 consultants ($A10.9 million); Treasury, 53 consultants ($A1.2 million). The consultancy market appears to be fragmented with consultancies awarded to a variety of players, including accounting firms, legal firms, major financial institutions and academics. The most common justification for their use was the lack of in-house skills. An additional lever in the privatization package concerns the introduction of ‘user-pays’, that is the transfer of responsibility for payment of services to consumers, with non-payers excluded. The approach has seen, for instance, the Higher Education Contribution Scheme amended so that students headed for high income-earning careers in fields such as medicine and the law are now required to make an increased contribution towards the cost of their education. Other examples include the limitations put on government-subsidized childcare places, thereby increasing the burden on users, and the tightening of qualifications for unemployment benefits (Aulich 2000, p. 171). Overall the venture represents a steady transfer of public sector service provision to the private sector, the full extent of which is yet to be documented. In addition, an assessment of the impact on the cost and quality of service delivery generally is yet to be undertaken. Already, however, it confirms a revised role for government. Governments must ‘steer’, not ‘row’. The core functions of government are to provide policy direction, resource allocation and contract management. Part-time Employment ‘Part-time’ applies generally to employees who ordinarily work less than the award number of hours per week. Part-time employees can have permanent, temporary or casual status. Greater flexibility in staffing arrangements has been a reform objective and part-time employment overall has increased accordingly. During the period 1986–98, part-time employment in the public sector increased from 12.6 to 26.0 per cent of the total number employed (see Table 4.4). As might be anticipated, the bulk of part-time employment is located at state level. In the Australian federal system, state governments carry

89

Public employment in Australia

Table 4.4

Total public sector full- and part-time employment, 1986–98 Full-time (%)

Part-time (%)

N (000s)

74.0 76.7 83.4 87.4

26.0 23.3 16.6 12.6

1473.4 1585.3 1731.4 1727.4

1998 1996 1991 1986

Source: Figures supplied by ABS.

Table 4.5 Part-time employment by level of government and gender, 1986–98 Federal

State

Local

Males Females N Males Females N Males Females N (%) (%) (000s) (%) (%) (000s) (%) (%) (000s) 1998 1996 1991 1986

23.7 17.7 22.2

76.2 82.3 77.8

27.9 32.4 27.7 26.6

21.9 21.9 23.5

78.1 78.2 76.5

318.9 298.3 231.0 172.1

23.9 19.3 19.8

76.1 80.7 80.2

36.9 39.3 28.5 19.2

Source: Figures supplied by ABS.

responsibility for the labour-intensive services such as education, health and transport. During 1993–97, for instance, the West Australian public service recorded a drop of 10 per cent in full-time employment. ‘Non-fulltime’ employment rose from 24 per cent in 1993 to 34 per cent in 1997. Of 50 of the larger public sector agencies in that state, six employed more than 50 per cent of their workforce on a non-full-time basis. This latter group included a major hospital complex and secondary schools in rural areas (Profile 97 ). If part-time employment has increased, its gender composition has remained fairly constant. Women continue to account for more than three-quarters of the part-time workforce at each level of government (see Table 4.5). Table 4.5 derives from surveys of the total public sector workforce, at all three levels of government, and includes government employees working in the range of government bodies. Statistics for the federal public service proper, that is, for those employed under the terms of the Commonwealth Public Service Act, suggest a reverse trend at that level. The Commonwealth

90

The state at work, 1

part-time component has increased, but from a relatively small base, and it remains a relatively minor segment. In 1977 the federal public service comprised a part-time workforce of 1.6 per cent; the equivalent figure in 1998 was 5.2 per cent. In each instance women made up between 85–88 per cent of the part-time workforce. At the same time, in the distinction between ‘permanent’ and ‘temporary’ employees, the number of ‘temporary’ employees actually decreased. In 1989 the temporary category accounted for 15.6 per cent of total employment under the Act; in 1998 the equivalent figure was 10.3 per cent (APS Statistical Bulletin 1997–98, p. 12). These figures make no allowance for changes in coverage of agencies under the Act. They could be misleading. On the other hand, they appear to report a public service that in the course of two decades of strenuous cutback and reform has become more ‘permanent’ in composition.

PUBLIC SECTOR EMPLOYMENT BY POLICY AREA The extensive cutbacks in public sector employment have had no dramatic impact on government employment policy area priorities: the Commonwealth workforce is still concentrated mainly in the areas of ‘government administration’ and communication services, only more so; state government employment is still focused primarily in the policy areas of education and health, only more so; and the bulk of local government employment remains predominantly in the government administration area. In so far as the series of cutbacks has entailed a rethinking of the appropriate role of government, the impact has been to limit or terminate government involvement in the more marginal policy areas. The data in Table 4.6 is drawn from Australian Bureau of Statistics (ABS) surveys based on the Australian and New Zealand Standard Industrial Classification (2nd edn, 1993). It shows public sector employment figures by policy area, at Commonwealth, state and local government levels, for the years 1983, 1991 and 1999. The broad category ‘government administration’ encompasses such a range of areas, from ‘Financial and economic management except banking’ to ‘Governor-General’s unit operation’, that a large amount of data is lost for analysis. The limitations of the classification aside, some observations can be made. Public Employment by Policy Area and Level of Government As stated, the ranking of top priority areas in public sector employment has undergone little change; the impact of a diminished workforce is more evident in what might be seen to represent ‘non-core’ government activities.

91

Public employment in Australia

Table 4.6 Public employment by policy area and level of government, 1983 and 1999 Policy area

Government administration. Defence Justice Police Finance and Insurance* Communication services Transport Energy Construction Manufacture Mining Agriculture, fisheries and forestry Education Health Community services Culture and recreation Personal services Other Total

Total

Federal(%)

1983

1999

254 700 23 100 n/a 32 400

305 100 30.0 29.6 16 600 100.0 100.0 14 900 n.a. 15.4 42 100 8.3 5.0

63 800 126 600 148 000 134 700 50 800 54 800 6 800

1983

1999

State(%)

Local(%)

1983 1999 1983 1999 25.3 – n/a 91.7

29.3 – 84.6 95.0

44.8 – – –

41.1 – – –

35.1

38.6

63.5

–

1.4

83 500 100.0 100.0 51 600 27.5 12.0 40 600 1.1 1.0 20 900 16.7 – 17 900 44.5 83.8 1 700 – –

– 71.5 80.0 83.3 53.6 100

– 85.3 89.2 88.0 14.5 82.4

– 1.0 18.9 – 1.8 –

– 2.7 9.9 2.5 1.7 17.6

7 400

61.4

11 200 342 700 254 200

5 900 446 500 274 200

2.7 3.4 6.7

1.7 1.4 0.8

95.5 96.4 93.3

98.3 98.6 99.2

1.8 0.2 –

– – 0.1

14 600

35 300

1.4

3.4

98.6

96.0

–

0.6

26 500 1 200 34 900

27 000 1 100 48 400

40.8 41.7 39.5

30.4 – –

59.2 50.0 60.5

66.3 90.0 99.0

– 8.3 –

3.3 9.1 1.0

1 581 000 1 440 700

25.1

16.4

65.8

74.1

9.0

9.4

Note: * refers to government trading and financial bodies. Financial management (e.g. Treasury) is included under ‘Government Administration’ in this classification. Source: Data supplied by ABS.

The cutbacks in employment have been across the board, but, as noted above, they have been also quite specifically targeted. In 1983 the ‘big three’ areas of public sector employment overall were education, government administration and health. The same three areas dominate in 1999. The only change during the period is that each has increased its ratio of total employment. Education increased from 21.7 per cent in 1983

92

The state at work, 1

to 31.0 per cent of the total in 1999; government administration increased from 16.1 to 21.2 per cent; health from 16.1 to 19.0 per cent (see Table 4.6). Of the remaining policy areas – comprising 46.2 per cent of the workforce in 1983, but only 28.8 per cent in 1999 – the largest single leap was in the community services area. Its workforce expanded by 141.8 per cent. This category includes accommodation for the aged, which might account for the bulk of the increase and if so, would seem to reflect an ageing population (actual breakdowns of the figures are currently not available). During the period, policies anticipating changes in the age demographics have been an expanding item on government agendas. The main areas of public employment cutback arise from privatizations. Specifically, public employment in the area of finance and insurance (for example trading and financial bodies) dropped by 88.4 per cent. Similarly, in the same period, the number of public employees engaged in mining activities dropped by 75.0 per cent, energy by 69.9 per cent and manufacturing by 67.3 per cent. The size of the defence workforce (military and civilian combined) fell by 28.1 per cent. At each governmental level, taken separately, the impact of the changes in employment levels is most evident in the increased concentration in the area of government administration. The area as defined in this classification is too broad to withstand detailed analysis, but the figures perhaps denote the new emphasis on governments’ role in ‘steering’, as opposed to ‘rowing’, most particularly at the Commonwealth level. Government administration claimed 19.2 per cent of the Commonwealth government workforce in 1983, the second largest single area of employment after communication services (31.8 per cent). In 1999, government administration, now the largest single area, accounted for 38.1 per cent of the total, with communication services now the second largest area (35.3 per cent of the total). Continuation of the Commonwealth’s strong presence in the communications area awaits political resolution. The telecommunications organization, Telstra, has undergone two episodes of share offers to the public, but it remains 51 per cent government owned. The current coalition government has repeated its intention to complete the privatization, but, to date, has been thwarted by strong electoral opposition to the move, particularly from rural areas. As it stands at present (2002), the government has given an undertaking that the third and final round of the sale will not proceed until Telstra services meet a set of criteria that includes the provision to rural areas of services equivalent to those available in the cities. At state government level, government administration increased from 6.2 to 8.4 per cent of total employment in the period 1983–99. The most labour-intensive areas, however, remained education (31.7 per cent of the total in 1983; 41.2 per cent in 1999) and health (22.8 and 25.5 per cent).

Public employment in Australia

93

Again reflecting the impact of privatizations, transport employees dropped from 10.2 to 4.1 per cent of the total state-level workforce; energy from 10.4 to 3.4 per cent. Most states have now shed (sold off) the vast electricity, gas and water utilities that not so long ago were powerful and wealthy players in state administrative machines. Despite the reduced scale of direct employment in the transport policy area, road and rail transport remain prominent items on state government agendas. Freeway planning and construction are major state responsibilities and motor registrations and licences provide a substantial source of state government revenue. Most state railway systems have been corporatized, although the need to maintain unprofitable services to sparsely populated rural regions ensures some form of continuing government involvement. The local government workforce remains concentrated in the government administration area, only more so. In 1983 the category accounted for 79.8 per cent of the local government workforce, with the remainder mainly in the energy area (17.8 per cent of the total). By 1999 government administration claimed 92.3 per cent of the total, with energy again the second largest area of employment, but now reduced to 2.9 per cent of the total. The policy areas classification also reflects the changes in federal–state administrative arrangements, outlined above. The Commonwealth government retains its employment monopoly in the areas of defence and communication services. In 1983 state governments had the monopoly in the area of mining; by 1999 17.6 of the mining workforce was engaged at the local government level. All other policy areas identified here continue to operate in a federal context, generally with a reduced Commonwealth component and a concomitant expanded state-level workforce. As noted earlier, major changes in federal–state administrative arrangements during the 1990s had the effect of expanding the Commonwealth’s role in setting the policy ambits and monitoring activities, meanwhile shifting an increased administrative load to state level. The impact is most evident in the changes in the relative sizes of the Commonwealth and state workforces in the areas of transport, health and education. Not shown in Table 4.6, both the Commonwealth and state governments showed growth in employment in the ‘services to transport’ sub-category. The increase was more notable at the Commonwealth level, denoting again the emphasis on the ‘steering’ role. Table 4.6 shows a similar pattern at the respective Commonwealth and state levels for the culture and recreation and finance and insurance policy areas. In each case, however, the changes reflect actual Commonwealth withdrawal rather than a reallocation of functions. There is a different story again in the community services (including, in particular, services to the aged) policy area. It is identified above as an area of employment growth (by 141.8 per cent, 1983–99). Accompanying that

94

The state at work, 1

change, the Commonwealth input has grown proportionately, the equivalent state government role has decreased accordingly, and local government has entered the field. The trend might be anticipated to accelerate as the decade advances. Commonwealth Government Employment by Policy Area The employment statistics reproduced so far are sourced from the nationwide surveys undertaken by the ABS. The following summary of federal government employment draws on reports issued by the Public Service and Merit Protection Commission (PSMPC). As indicated earlier, the data relate only to those employees who come under the coverage of the Public Service Act, that is, about 57 per cent of the total Commonwealth workforce. They lack the comprehensive coverage supplied by the ABS, but they provide a breakdown of the ABS’s broad ‘government administration’ category. Interpretation is constrained by changes in the coverage of the Act. In Table 4.7, for instance, apparent large increases in court and parliamentary staffs can be discounted; these staffs were previously employed under alternative legislation. Again the cutbacks have not changed the employment profile of the federal public service overall. In 1990 social security, defence, the Taxation Office and education together accounted for 58.0 per cent of total Commonwealth public service employment. In 1998 the decrease in education reflects the reordering in federal–state arrangements (the cuts to health occurred prior to 1990), but social security (now including Centrelink, with a total staff of 25 014), defence and the Taxation Office together still account for 53.4 per cent of the total. Of the other policy areas, the increase in the number of employees engaged in the immigration area arises perhaps from the need for government to respond to the more recent influx of persons seeking refugee status in the country. The increase in finance and administration area employment derives from that department’s role in overseeing and monitoring the public sector reform agenda. The large ‘Other’ category classification for 1990 includes the 13 079 staff employed in the Department of Administrative Services. Its services have since been privatized.

HUMAN RESOURCES MANAGEMENT By the 1980s, the Commonwealth government’s Personnel Management Manual ran to nine volumes. It included rules and procedures for 17 different types of leave (O’Faircheallaigh et al. 1999, p. 146). Such cumbersome

95

Public employment in Australia

Table 4.7 Commonwealth Public Service Act employment by policy area, 1990 and 1998 Policy area Prime Minister Defence Foreign Affairs Foreign affairs: AusAID Immigration Customs Justice Justice: courts and tribunals Finance Finance: Australian Taxation Office Communications/transport Labour Other industry Agriculture Education Health Social security Aboriginal affairs Environment Parliamentary services Other Total

1990 (N)

1998 (N)

Percentage change

542 23 275 4 442

452 18 946 4 010

16.6 18.6 9.7

587 2 630 5 215 4 246 268 2 408

600 3 737 4 373 2 294 2 062 4 286

2.2 42.1 16.1 46.0 669.4 78.0

19 417 3 132 1 615 2 172 4 401 10 523 5 464 30 062 1 519 3 341 767 17 660

17 581 2 846 1 523 4 515 3 886 4 511 5 639 28 248 1 800 3 287 1 806 3 913

9.5 9.1 5.7 107.9 11.7 57.1 3.2 6.0 18.5 1.6 135.4

143 491

121 262

15.5

Source: PSMPC, Australian Public Service Statistical Bulletin (relevant years).

arrangements were unlikely to survive in a managerialist milieu. Changes in personnel management have been far-reaching. The intent behind the reforms was summed up by the Secretary of the Department of Employment, Workplace Relations and Small Business (Shergold 2000, p. 19): The government’s public service reform agenda is aimed at promoting a highperformance APS [Australian Public Service] – one that is capable of operating in an increasingly competitive environment and emphasises contestability, value for money and a focus on client service. Key components . . . are the removal of unnecessary prescription and centralized controls, with responsibility for workplace relations ‘devolved’ to individual agencies.

96

The state at work, 1

The earliest, most radical, reform measures were enacted by the Victorian state government in 1992. Under the provisions of the Act, the number of government departments was reduced from 22 to 13, the Public Service Board was abolished and the Premier was made employer of all departmental secretaries, who, in turn, became the employers of their respective departmental staffs. Further, all senior executives were put on limited-term contracts with increased performance pay components, impediments to lateral recruitment were abolished and appeals and grievance mechanisms were rationalized (O’Neill 1998). A not dissimilar agenda has been pursued at the Commonwealth level, at a slightly slower pace. Early Commonwealth reforms set out to streamline procedures. A major restructuring in 1987 introduced the concept of the ‘super portfolio’ by amalgamating 28 departments into 17 portfolios. It brought also shifts in both the work undertaken by the public service and the manner in which that work is done. In brief, the machinery of government changes were accompanied by the devolution of personnel management responsibilities, including recruitment, appointments, retirements, transfers and promotions, to departments and agencies. These changes were followed by a restructuring of federal public service classifications, prompted in large part by changing skill requirements and designed to introduce a flatter administrative structure. Previously the federal public service had a fourtier divisional structure, each with internal class or grade or work levels. A series of government negotiations with staff associations found acceptance for the introduction of generic classifications and multiskilling across the service. Some hundreds of classifications were collapsed into eight bands of officers, plus professional and technical grades. The system of classification has undergone further revisions as the reform programme has progressed. Similar streamlining changes have been introduced at the state and local government levels. Table 4.8 sets out the major classification groups at federal level in 1998. The Senior Executive Service (SES) is roughly equivalent to the former Second Division, comprising deputy secretaries, specialist advisers, heads of divisions, branches or bureaus and regional directors (see further, below); the category Senior Officer (SO) includes both management and professional employees; Administrative Service Officer (ASO) can be read as roughly equivalent to the former clerical division, albeit skill levels are now much more specific and no job involves more than 50 per cent keyboard work; ‘Other’ is defined in official publications as ‘includes staff in classifications not mentioned in the above categories’(!). The Senior Executive Service (SES) was established in 1984. As stated, in composition it is roughly equivalent to the former Second Division. The major differences arise from the conditions of employment, recruitment

97

Public employment in Australia

Table 4.8 Federal government permanent staff: departmenta by major classification group, 1998 SES (%)

SO (%)

ASO (%)

Other (%)

Total (N)

Attorney-General Communications and the Arts Defence Employment, Education, Training and Youth Affairs Environment Finance and Administration Foreign Affairs and Trade House of Representatives/Joint House/Parliamentary Library/Parliamentary Reporting Staff/Senate Health and Family Services Immigration and Multicultural Affairs Industry, Science and Tourismb Primary Industries and Energy Prime Minister and Cabinet Social Security Transport and Regional Development Treasuryc Workplace Relations and Small Business

4.6 2.3 0.6

8.6 16.8 6.0

49.2 53.5 49.8

37.6 27.4 43.5

4 329 2 195 18 752

1.3 1.7 1.8 5.6

17.2 12.3 14.1 25.8

73.7 35.5 58.8 55.1

7.9 50.5 25.2 13.5

4 279 2 870 3 336 2 864

1.9 1.7

15.5 14.8

43.7 50.7

38.9 32.8

1 134 4 489

1.1 1.5 1.9 4.1 0.3

13.8 11.3 13.6 24.6 3.8

77.8 18.9 38.4 63.7 90.5

7.2 68.4 46.1 7.6 5.4

2 814 6 413 3 548 2 407 24 351

4.1 1.2

25.1 14.3

47.5 74.5

23.3 9.9

832 22 852

3.5

22.2

65.7

8.6

1 320

Total

1.4

10.9

64.1

23.5

108 785

Notes: a Includes attached agencies. b Includes Australian Customs Service. c Includes Australian Taxation Office. SES  Senior Executive Service. SO  Senior Officer. ASO  Administrative Service Officer. Source: PSMPC, Australian Public Service Statistical Bulletin 1997–98.

and remuneration. The SES comprises officers employed under contract, albeit they remain under the provisions of the Public Service Act for the period of their contract, which is limited to a maximum of five years, with provision for renewal. They are eligible for performance pay and are

98

The state at work, 1

individually evaluated against formal results-based performance appraisal schemes usually specific to the agency. The intent behind the creation of the SES is to create an identifiable elite with generic skills that can be applied to management across most portfolio areas. The SES has also become a feature of state government administrations. In 1998 the SES comprised 1.4 per cent of the total federal public service (see Table 4.8). Foreign Affairs and Trade, regarded traditionally as the crème de la crème, has the highest ratio of SES staff with 5.6 of officers classified as SES. Other departments with a proportionately higher SES component are: Attorney-General’s with 4.6 per cent; Prime Minister and Cabinet, 4.1 per cent; and Transport and Regional Development, 4.1 per cent. In the 1990s, the Commonwealth government led the endeavour to achieve increased uniformity in transport policies across the states, which might account for the last mentioned. The Commonwealth government is not known for strong regional policies. The relatively low proportion of SES staff at Treasury (1.2 per cent) might be queried, did not that portfolio include the Taxation Office. Other client-oriented departments such as Social Security and Immigration have similarly low SES ratios (0.3 and 1.1 per cent respectively). Each of these departments has a higher than average ASO component. The ASO group comprise about two-thirds of the federal public service overall. Social Security has an ASO component of 90.5 per cent; Immigration, 77.8 per cent; and Treasury, 74.5 per cent. The low SES ratio at Defence (0.6 per cent) is matched by a much higher than average ‘Other’ category, which, one assumes, includes supply and maintenance staff. The relatively high ‘Other’ category in the Department of Industry, Science and Tourism probably includes mainly customs officers. Concurrent with the introduction of the SES, Department Secretaries (heads of department, also known as Chief Executive Officers) were placed on contractual employment and removed from the protection of the Public Service Act. (In a recent court case, the head of the Defence Department challenged his dismissal by his Minister. The court found that Department Secretaries were subject to dismissal by their Ministers provided that due process was observed, that is, that reasons were given in writing and that the Secretary had opportunity to respond in writing. A subsequent further court challenge by the Secretary that the Minister’s loss of ‘trust and confidence’ in him represented insufficient grounds for dismissal was rejected by the court (Richardson 2000, p. 5)). The establishment of a SES, a concept borrowed from the USA, and the now fairly widely accepted expectation that newly in-coming governments will replace serving departmental heads with their own appointees has

Public employment in Australia

99

given rise to concerns regarding politicization of the service. Weller’s longitudinal study of appointments at the Commonwealth level (1989) found little evidence that removal of existing senior staff extended below the level of chief executive officer. Anecdotal evidence suggests that at state level, overt politicization might be more to the fore, but there is no substantive evidence for the claim (O’Faircheallaigh et al 1999, p. 152; on politicization, see also Mulgan 1998). A currently more controversial outcome of the introduction of the SES concerns the question of performance pay. Conceived of originally as a means of closing the gap between public and private sector salaries, performance pay has come to be associated more with the size and complexity of the agency in which the individual executive is located, rather than being awarded on the basis of actual performance as assessed against a set of published criteria. According to O’Fairchealleigh et al (1999, p. 154), ‘The fact that performance pay may be awarded without reference to efficiency or explicit merit criteria, and may involve patronage or favouritism, does not appear to have unduly worried Australian governments.’ The development has, however, given rise to increased demands from other non-executive level employees for similar pay-enhancing opportunities. The broadening gap between executive salaries and base pay rates appears to be another likely emerging problem. The reform agenda has been further consolidated with the enactment of the Workplace Relations Act 1996 and, after some delay, the Public Service Act 1999. The former measure brought public sector workplace relations into line with the private sector. It abandoned the previous service-wide approach in favour of transferring responsibilities for workplace relations to the individual agencies. Henceforth, agencies would establish their own certified agreements: ‘Agencies would be more accountable for improving their people management policies, their overall performance and, importantly, remuneration and other conditions of employment for public servants. This would in turn enable human resources management strategies to be more closely aligned with the organizational goals of individual agencies’ (Yates 1998, p. 86). By the year 2000, 99 agencies – covering virtually the entire Commonwealth public service – had certified agreements in place. A diversity of agreement types has been adopted and the process has been accompanied by minimal recourse to industrial action (Yates 1998, p. 88; Shergold 2000, p. 20). The uniformity in pay and classification matters typical of the traditional career service model has been replaced by a market-based model characterized by limited-term contract employment and performancebased pay. A likely consequence is wide variations in conditions of employment and rates of pay between agencies, including differences in hours of

100

The state at work, 1

work, productivity bonuses and job classifications. Ironically the evolving system is reminiscent of that pertaining at the end of the nineteenth century, when wide inter-agency differences in wages and salaries for apparently similar jobs provided the incentive for the reforms that introduced parities across the service as a whole. The government view, however, appears to be that it will improve performance by promoting a stronger focus on strategic human resource management. The changes have been capped by enactment of the Public Service Act 1999. The legislation saw the previously voluminous, over 250 pages long and much amended Public Service Act 1922 reduced to an easy-to-read 47 pages. Under the provisions of the Act, Secretaries are formally designated as the employer of staff within their agencies, able to determine remuneration packages and made responsible for their employees. A 13-point code of conduct binds employees across the service and establishes a set of ‘Australian Public Service values’, including professionalism, a commitment to apolitical service, merit-based selection, discrimination-free workplaces and a focus on achieving results and managing performance: There is no longer a ‘one size fits all’ culture in the APS. Agencies are now delivering terms and conditions of service that are appropriate for their own employees and agencies’ operating needs. . . . The Act heralds a shift to a values-based APS, rather than one driven by rules and regulations . . . [W]orkplace relations will act as the driver of cultural change within the APS. There will be a recognition that what binds us together as public servants are values and ethics, not a standard set of terms and conditions, which will increasingly vary between agencies. (Shergold 2000, pp. 19, 21)

PUBLIC EMPLOYMENT BY GENDER, ABORIGINALITY, DISABILITY AND ETHNICITY Equal employment opportunity (EEO) legislation and affirmative action policies represent relatively recent developments in the Australian public services. The rule that women must upon marriage resign or become temporary employees was abolished in 1967. The first EEO legislation was introduced by the New South Wales state government in 1980. The Commonwealth followed in 1984. Legislation is now in place in all states. There are concerns that the current emphasis on results-based management practices and the devolution of workplace relations responsibilities to individual agencies will weaken responsiveness to equity policies and diminish monitoring. There is no empirical evidence to support the claims, although there can be no doubt that the managerialist thrust has shifted the spotlight away from equity concerns.

101

Public employment in Australia

Table 4.9 Federal government permanent staff: equal employment opportunity (EEO) data, 1990 and 2000a 1990 (N  134 308)

2000 (N  100 335)

Gender Male Female

73 377 60 931

50 618 49 717

Aboriginality Yes No No data

1 576 91 776 40 956

2 273 74 103 23 959

Disability Yes No No data

6 284 86 050 41 974

4 247 68 320 27 768

Language background Group 1b Group 2c English speaking No data

6 577 10 140 77 415 40 176

3 382 8 214 62 542 26 197

EEO group

Notes: a Data supplied to Department of Finance by individuals on a voluntary basis. b Refers to people born overseas who arrived in Australia after the age of five and whose first language was not English. c Refers to people who arrived in Australia before the age of five, or to Australian-born people with parents who are first generation people of non-English speaking background. Source: Australian Public Service Statistical Bulletin.

EEO data is provided on a voluntary basis, thereby limiting coverage and reliability, as is illustrated by the large ‘no data’ categories in Table 4.9. The table shows, however, a federal public service that remains predominantly Anglo-Celtic in composition. At the same time, the service has become one in which the sexes are almost equally balanced. Comment is tempered by the observation that, during the period under review, the attrition rate for males has been considerably greater than for females (Table 4.9). The part-time workforce still comprises predominantly women. In 1977 part-time employment in the federal public service ran at 2.1 per cent, of which 89.5 per cent were women. In the year 2000, 8.6 per cent of the total Commonwealth workforce worked part-time. Of this group, 85.3 per cent were women (Table 4.10).

102

The state at work, 1

Table 4.10 Federal government permanent staff: gender by employment category, selected years 1977–2000 1977

1980

1985

1990

1996

Males Full-time Part-time

78 389 263

79 625 320

86 308 494

72 927 389

66 941 556

52 875 1 408

Females Full-time Part-time

37 693 2 249

42 755 2 366

54 603 3 240

57 969 2 795

57 235 4 114

48 521 8 150

118 594

125 066

144 645

134 080

128 846

110 954

Total

2000

Source: Australian Public Service Statistical Bulletin.

The period 1988–2000, however, has seen increased feminization of the senior grades of the service (Table 4.11). In the year 2000, women comprise just over half of the trainee and lower grades (54.4 and 54.1 per cent respectively). But whereas in 1988 18.0 per cent of the Senior Officer (now ‘Executive’) class were women, the equivalent figure in 2000 is 33.0 per cent. Similarly, whereas the elite SES was made up of 8.6 per cent women in 1988, the equivalent figure in 2000 is 26.1 per cent.

CONCLUSION It has been a protracted reform process, carried forward by successive governments. Halligan (2001, p. 191) summarizes the emerging debates and issues: ● ● ● ● ● ● ●

How far will the rebalancing of public and private sectors be taken? Will the commitment to a government framework based on outcomes be maintained? Capability of public service: maintain, dismantle or rebuild core capability? Government to governance: how far will this proceed? Employee/employer relationship: will this change with the demise of the career service? Relationship between public sector management and the political context? The Australian model: increasingly isolated or the best future option?

103

Public employment in Australia

Table 4.11 Federal government permanent staff: classification group by gender, 1988 and 2000 1988 (N  145 221)

2000 (N  100 335)

Males SES SO ASO Other Trainee

1 397 7 128 38 544 36 455

1 308 13 030 35 801 – 479

Females SES SO ASO Other Trainee

132 1 561 51 447 8 557

462 6 419 42 264 – 572

Total SES SO ASO Other Trainee

1 529 8 689 89 991 45 012

1 770 19 449 78 065 – 1 051

Notes: Key: SES: Senior Executive Service; SO: Senior Officer (‘Executive’ in 2000); ASO: Administrative Service Officer (APS in 2000). Given changes in the classification system, the 1988 and 2000 categories are not strictly equivalent. Source: Australian Public Service Statistical Bulletin.

For the meantime, the concept of the ‘career service’ has gone. Permanence has been usurped by contract employment and organizational cutback and termination. Neutrality is in a fine balance with responsiveness to the political agenda. Anonymity is lost as public servants become more individually identified with outputs and performance. The commitment to due process must do battle with the rigorous cost-cutting approach. The concept of a unified service has given way to a diversity of individual agency industrial agreements.

NOTES 1. In November 1975, the Senate delayed passing the annual appropriation bills. The Governor-General dismissed the then Labor Prime Minister, installed the Liberal–National

104

The state at work, 1

Party Coalition Opposition Leader as Acting Prime Minister and called an election for December. The election was won by the Coalition. There is no doubt that the Senate was acting within its powers. Controversy continues regarding the Governor-General’s response and the role taken by various other actors. 2. Under section 128 of the Constitution, constitutional change is by referendum, with the proviso that besides a majority overall, proposals require a majority in a majority of states in order to be passed. The Commonwealth’s 1948 proposal included a list of 14 new powers, none of which received electoral endorsement.

REFERENCES Aulich, Chris (2000), ‘Privatization and contracting out’, in Gwynneth Singleton (ed.), The Howard Government, Sydney: University of New South Wales Press, pp. 162–73. Australian Bureau of Statistics (ABS) (1993), Australian & New Zealand Standard Industrial Classification (1993 edition), ABS Catalogue No. 1292.0; NZ Catalogue No. 19.005.0092. Australian Bureau of Statistics (ABS) (1993), Wage & Salary Earners, Australia, ABS Catalogue No. 6248.0. Australian Bureau of Statistics, Year Book Australia, various years. APS (Australian Public Service) Statistical Bulletin, 1997–98. Halligan, John (2000), ‘Public service reform under Howard’, in Gwynneth Singleton (ed.), The Howard Government, Sydney: University of New South Wales Press, pp. 49–64. Halligan, John (2001), ‘Conclusion’, in C. Aulich, J. Halligan and S. Nutley (eds), Australian Handbook of Public Sector Management, Crows Nest, NSW: Allen & Unwin, pp. 186–93. Hancock, W.K. (1930), Australia, London: Ernest Benn. Kingston, Margo (2001), ‘System error’, Sydney Morning Herald, 20 January. Mulgan, Richard (1998), ‘Politicization of senior appointments in the Australian public service’, Australian Journal of Public Administration, 57 (3), September, 3–14. O’Faircheallaigh, C., J. Wanna and P. Weller (1999), Public Sector Management in Australia, 2nd edn, Brisbane: Centre for Australian Public Sector Management. O’Neill, Deirdre (1998), ‘Different roads to similar places?’, Canberra Bulletin of Public Administration, August, (89), 169–73. Profile 97: Employee Appointment Type, Employees Status, 6 April 1998, www.wa. gov.au/psmo/pubs/directory/guides/hrpp/wdu/prof97/section13.htm, accessed 15 September 1998. Richardson, Jack E. (2000), ‘Defenceless secretaries’, Canberra Bulletin of Public Administration, September, (97), 5–6. Shergold, Peter (2000), ‘The Public Service Act and workplace relations: complementary reforms’, Canberra Bulletin of Public Administration, September, (97), 18–21. Spann, R.N. (1979), Government Administration in Australia, Sydney: George Allen & Unwin. Stewart, Jenny (1998), ‘The end of the career service’, Canberra Bulletin of Public Administration, August, (89), 1–3.

Public employment in Australia

105

Tingle, Laura (2001), ‘Passing the ministerial buck’, Sydney Morning Herald, 23 July, p. 29. Vardon, Sue (2000), ‘We’re from the Government and we’re here to help – Centrelink’s story’, Australian Journal of Public Administration, June, 59 (2), 3–10. Weller, Patrick (1989), ‘Politicization and the Australian public service’, Australian Journal of Public Administration, December, 48 (4), 369–81. Wettenhall, Roger (2000), ‘Reshaping the Commonwealth public sector’, in Gwynneth Singleton (ed.), The Howard Government, Sydney: University of New South Wales Press, pp. 65–95. Yates, Bernie (1998), ‘Workplace relations and agreement making in the Australian public service’, Australian Journal of Public Administration, June, 57 (2), 82–90.

5.

Public employment in Canada: downsizing in a multi-layered state James Iain Gow and Sharon L. Sutherland

DISTINCTIVE CANADIAN FEATURES Features that need to be kept in mind in the study of public employment in Canada are natural, cultural, constitutional and technical. Its great size and for the most part, sparsely scattered population, have meant that public services had to be stationed in large numbers outside federal and provincial capitals. With the bulk of its population located in the south, near the American border, the land poses a challenge to the effective presence of governments and their services. The chief cultural fact of Canadian existence is the presence of two founding cultures and linguistic groups. Before Confederation in 1867, politics in the United Provinces of Canada was run strictly on a twocommunity basis, both in the composition of the government and of the civil service. After the creation of Canada until the introduction of the merit system in 1918, the division of civil service employment was largely handled by favouritism. Fifty years of the new merit system led to a sharp decline in the presence of French Canadians. After a royal commission on the question, the Official Languages Act of 1969 declared French and English to be official languages. The exact status of the two communities within the Canadian polity has never been fully determined. In the 1960s, the emergence of an independence movement in Quebec led to seemingly endless constitutional negotiations, and referenda on sovereignty in Quebec in 1980 and 1995. The second referendum led to a very narrow rejection of the idea of a sovereign Quebec, linked to the rest of Canada by a partnership. From the beginning, the Canadian constitution combined two features that were to some degree contradictory (Verney 1986): its system of government was drawn from the British parliamentary model, but it also created a federation, with the distribution of powers enshrined, however imperfectly, in the constitution.1 Thus, for more than a century, the principal legal limit to the powers of the federal government was the division of powers. The addition to the constitution in 1982, of a Charter of Rights 106

Public employment in Canada

107

and Freedoms was a further step away from the classic Westminster model, since it gave the courts a vast new area of supervision of the legality of acts of parliament. Even allowing for these constraints, the Canadian system creates a strong executive. The Prime Minister who is leader of a majority party in the House of Commons has the potential for a unified control of the administration and the possibility to have adopted all but the most controversial of legislative proposals (Savoie 1999). Another feature of decentralized government in Canada concerns the variety of employers and consequent legal status of local employees. Although the governments of the ten provinces and the three territories have final control over the organization and jurisdiction of all forms of local government, the latter have, for the most part, legal autonomy within their spheres of competency. Thus local governments have their own employees, who do not fall within the category of provincial civil servants or employees. Furthermore, most employees in health, education and welfare institutions are not civil servants of any level of government, federal, provincial or local, even though they come under the jurisdiction of provincial governments. The universe of the public sector in Canada is represented in Figure 5.1. This scattered nature of public employment has long caused problems for those who wished to know the full extent of public employment in Canada. Statistics Canada only began to compile a consolidated portrait of public sector employment in the 1990s. The problem was that the Statistics Act compels individual citizens but not other governments. In this sense, it is not a central statistical office for the country, but only a federal agency. As a result, statistics for provincial government employment as collected by Statistics Canada are only relatively complete and reliable after 1970, and comprehensive public sector data only from 1990.

TOTAL PUBLIC EMPLOYMENT For reasons previously noted, until the 1990s, there was no single series of statistics concerning the broadest definition of the public sector. Since the 1950s, Statistics Canada has used a basic distinction between government employees and those working for public enterprises. From a base of very limited government in the 19th century, public employment historically developed gradually in the 20th century into a welfare state presence after the Second World War. The general rule has been that the federal government grows in times of crisis, war or economic, whereas the provinces have grown in times of peace. Canada does not have

108

The state at work, 1 2002 Public Sector FMS Statistical Universe

Government business enterprises

Government

Federal government

General government

Ministries, departments, agencies and non-autonomous funds

Autonomous funds and organizations

Federal non-autonomous pension plans

Canada and Quebec pension plans

Canada pension plan

Quebec pension plan

Provincial and territorial government

Local government

General government

General government

Ministries, departments, agencies and non-autonomous funds

Municipalities and non-autonomous funds

Autonomous funds and organizations

Autonomous funds and organizations

Provincial non-autonomous pension plans

School boards

Universities and colleges

First nations and Inuit government

General government

Financial enterprises

Non-financial enterprises

Self-governing jurisdictions, Indian bands and Inuit and non-autonomous funds

Monetary authorities

Other financial enterprises

Autonomous funds and organizations

Universities and colleges

Universities

Colleges, vocational and trade institutions

Health and social service institutions

Provincial and territorial government business enterprises

Financial enterprises

Non-financial enterprises

Local government business enterprises

Financial enterprises

Universities

Colleges, vocational and trade institutions

Federal government business enterprises

School boards

Non-financial enterprises

First nations and Inuit government business enterprises

Health and social service institutions

Health boards Health and social service institutions

Social service organizations and community boards

Social service organizations and community boards

Other health and social service organizations

Other health and social service organizations

Source: Minister of Industry 2002. Financial Management System (FMS) 2002, S. 130, in: http://www.statcan.ca/english/IPS/Data/68F0023XIB.htm, accessed 27 May, 2003.

Figure 5.1

The public sector in Canada

Public employment in Canada

109

a strong state tradition, but it has had a greater reliance on state intervention than has the United States.2 Public services remained relatively small however, until the Second World War, because a major mode of intervention was transfer payments, that involved various grants and contributions to other governments, businesses and individuals, without the need of great numbers of civil servants. This mode of government also made possible the practice of clientelism or ‘patronage’ as it is known in Canada, in the distribution of money, the passing of contracts and the hiring of public servants. The pioneer analyses of public sector employment began only 30 years ago (Hodgetts and Dwivedi 1969, 1974). The first comprehensive historical study was that of Bird et al. (1979). Using census data, they found that public sector employment had grown steadily by 3 per cent per year since 1911, from 7 per cent of total employment that year to 12 per cent in 1971. This growth had its ups and downs however, the wartime high of 11 per cent of total employment was followed by a lower figure (10 per cent) in 1951 and then a swift rise to 14 per cent in 1961. If the share of total employment fell to 12 per cent in 1971, it was not because the number of public sector employees declined, but rather that total employment expanded so rapidly. Using Statistics Canada data in a comprehensive way, Sutherland and Doern (1986) found that public employment increased in absolute terms from 1960 to 1982 (Table 5.1). In terms of the labour force, however, it took an increasing share until 1975, after which it declined gradually until in 1982 it was at the same relative level (17.9 per cent) as it had been in 1960. In terms of the population, the relative growth continued until 1980 (Siegel 1988, p. 161). Between 1960 and 1982 the population grew by 42.9 per cent (17.5 million to 25 million), and the labour force by about 80 per cent. In this same 22 years, the total federal public sector grew by about 47 per cent, a figure much closer to population growth than to labour force growth. After this, no comprehensive figures are available until the new series introduced by Statistics Canada in the 1990s. Public sector employment statistics released in 1992 show that from 1987 to 1991, employment increased by 1.3 per cent per year. Growth in public employment and the progressive expansion of the coverage by the agency meant that the grand total of public sector employment jumped from 2.1 million in 1982, to 2.7 in 1991 (Statistics Canada 1992, Public Sector Employment, p. 14). As measured against the growth of the population, however, the public sector total peaked at 10 per cent in 1989, and declined to 9.7 per cent in 1992. The most recent and comprehensive numbers show that public sector employment began to decline in 1992, passing from 3 063 234 to a low of

110

2.7

7.6 3.4

11.0

2.5 1.6

4.1

173 121

491 850 218 280

710 130

160 800 96 696

260 496

Total

Total government General Enterprise

Total

Education sector Teaching Non-teachingd

Total

3.2

202 878

Total

2.3 0.4

2.2 1.0

139 434 63 444

Provincial Generalb Enterprise

149 403 23 187

5.2

334 131

Total

Localc General Enterprise

3.2 2.0

203 013 131 118

Federal Generala Enterprise

1960 % LF

342 570

211 463 131 107

769 407

543 350 226 057

188 761

162 901 25 860

238 817

168 536 70 281

341 829

211 913 129 916

1965

4.7

2.9 1.8

10.8

7.6 3.2

2.7

2.3 0.4

3.3

2.3 1.0

4.8

3.0 1.8

% LF

472 431

291 624 180 807

920 539

669 137 251 402

233 401

201 425 31 976

311 995

216 475 95 520

375 143

251 237 123 906

1970

5.7

3.5 2.2

11.0

8.0 3.0

2.8

2.4 0.4

3.7

2.6 1.1

4.5

3.0 1.5

% LF

507 612

313 341 194 271

1 165 839

860 038 305 801

286 441

247 199 39 242

423 450

288 937 134 513

455 948

323 902 132 046

5.1

3.1 2.0

11.8

8.7 3.1

2.9

2.5 0.4

4.3

2.9 1.4

4.6

3.3 1.3

1975 % LF

Table 5.1 Public sector employment, 1960–1982, with share of labour force

532 940

328 975 203 965

1 270 270

921 135 349 135

317 645

274 126 43 519

459 739

311 634 148 105

493 363

335 375 157 988

4.7

2.9 1.8

11.0

8.0 3.0

2.8

2.4 0.4

4.0

2.7 1.3

4.3

2.9 1.4

1980 % LF

509 006

314 201 194 805

1 298 923

955 805 343 118

332 680

287 103 45 577

476 667

317 407 159 260

489 576

351 295 138 281

1982

4.4

2.7 1.7

11.0

8.1 2.9

2.8

2.4 0.4

4.1

2.7 1.4

4.2

3.0 1.2

% LF

111

2.8

17.9

183 189

1 153 815

1 363 488

251 511 19.0

3.5 1 712 796

319 826 20.5

3.8 2 012 968

339 517 20.3

3.4 2 103 073

299 388 18.3

2.6 2 105 124

297 195

2.5 17.9

Source: S.L. Sutherland and G.B. Doern, Bureaucracy in Canada: Control and Reform (Toronto, University of Toronto Press, 1986) table 34–26.

Notes: a. Except federal hospital workers, included in hospital sector. b. Except provincial hospital workers, included in hospital sector. c. Local government employment figures are estimates for 1960 and 1965. Local enterprise figures are estimates from 1960–82. d. Estimates for 1960–82. e. 1979–80 hospital statistics used for 1980; 1982 figures are estimates.

Grand total

Hospital sectore

112

The state at work, 1

2 798 061 in 1998, before rising slightly to 2 798 970 in 1999 (Table 5.2). While government employment has always constituted over 80 per cent of total public employment, its share has increased as privatizations reduced the numbers of employees working for public enterprises. In 1999 government employment accounted for 90.6 per cent of the total. Public sector employment declined from 10.8 per cent of population in 1992 to 9.1 per cent in 1999. As a share of the labour force, total public employment fell steadily from 26.0 per cent in 1992 to 19.3 per cent in 1999. In spite of the fragmentary nature of these figures, it seems clear that public employment grew in absolute terms in Canada until 1992. In terms of the population, growth peaked earlier, in 1989, and declined steadily thereafter. However, in terms of its share in overall employment, public sector employment continued to grow with absolute numbers also peaking in 1992, at 26.0 per cent. The subsequent decline reflects the widespread move of governments of the 1990s to reduce spending and eliminate their deficits. It is interesting to note that there is a time lag in the growth or decline of public sector compensation. Wages and salaries continued to grow for two years after employment stopped growing, but by 1996 the decline of remuneration was a whole percentage point greater than that of employment (Statistics Canada 1998, Public Sector Employment, p. 34). Even after this adjustment, in 1999, public sector wages and salaries constituted 23.2 per cent of total wages and salaries in Canada, but public sector employment only accounted for 19.3 per cent of total employment, the difference presumably being caused by higher public sector remuneration.3 In comparative terms, Canadian public sector employment places the country right where its tradition of extreme moderation would lead one to expect it: in the middle of the pack of developed countries. Whether one uses Rose (1985) as Sutherland (1987b) does or Heller and Tait (1983) as Siegel (1988) does, Canada has dramatically fewer public employees than many continental European countries in relative terms, but somewhat more than the United States and Japan (see also Bird et al. 1979, p. 45). When downsizing of personnel is compared to budget cutting overall, some differences emerge. While reducing or eliminating the deficit was the top priority for the largest jurisdictions in Canada from the mid-1990s, some achieved it sooner than others. Sakatchewan and Alberta have had surpluses since 1994, the federal government achieved one in 1997, while Quebec and Ontario followed in 1998 and 1999 respectively (Department of Finance 2000). British Columbia ran deficits throughout the 1990s under the direction of a social-democratic (NDP) government. While public employment in Canada fell during the 1990s from 26 per cent to 19 per cent of the labour force, public expenditure fell from a high

113

50.8 29.5 1.7 31.2

50.9

29.1 1.7

30.8

Total

Local Government Enterprises

Total

Source:

Total

89.0 11.0 100.0

88.5 11.5

100.0

100.0

89.3 10.7

100.1 3037

31.7

30.0 1.7

50.6

46.0 4.6

17.8

13.3 4.5

1993

100.0

89.2 10.8

100.0 3003

32.0

30.3 1.7

50.1

45.8 4.3

17.9

13.1 4.8

1994

Statistics Canada Series Public Sector Statistics 68-213-XIB, April, 2001.

Type (%) Government Enterprises

100.0 3063

46.0 4.8

45.9 5.0

Provinces/Territories Government Enterprises

100.1 3057

18.0

18.4

Total

Public sector (%) Public sector (N 000s)

13.4 4.6

13.6 4.8

1992

Level (%) Federal Government Enterprises

1991

32.3

30.7 1.6

50.6

46.3 4.3

17.1

12.5 4.6

1995

100.0

89.6 10.5

100.0 2958

Table 5.2 Public employment by level of government and type, 1991–9

100.0

90.5 9.5

99.9 2871

33.2

31.6 1.6

50.8

46.5 4.3

15.9

12.4 3.5

1996

100.0

90.8 9.2

100.0 2803

33.3

31.8 1.7

51.2

46.9 4.3

15.3

12.0 3.3

1997

100.0

90.7 9.3

100.1 2798

33.6

31.9 1.7

51.4

47.0 4.4

15.1

11.8 3.3

1998

100.0

90.6 9.4

100.0 2802

33.6

31.9 1.7

51.4

46.9 4.5

15.0

11.8 3.2

1999

114

The state at work, 1

of 50.1 per cent of Gross Domestic Product in 1992 to 38.8 per cent in 1999. The main reason for these higher shares of GDP than of the labour force is presumably the importance of transfers of money from government to government, and from governments to individuals and enterprises. For the government of Canada alone, such transfers (pensions, unemployment insurance, grants, transfers to the provinces etc.) represented 37.5 per cent of budgetary expenses in 1999 (Department of Finance 2000, p. 16). The activities that governments carried out for themselves, final consumption expenditure, cost about 20 per cent of GDP in the late 1990s, thus approximating quite closely, public employment figures. In international terms, in 1992 Canadian governments spent a much greater share of GDP (50.1 per cent) than did the average of OECD members (39.8 per cent) and a little more than the members of the European Union (48.8 per cent). By 1999, Canada was close to the OECD average (37.5 per cent) having lowered its share to 38.8 per cent, while EU countries were now ahead of Canada, with 45.9 per cent of GDP (OECD 2000, p. 264). The gap between Canadian governments and the OECD average remained wider for government final consumption expenditure, since it was almost 20 per cent in 1997, while the OECD average was 14.6 per cent (OECD 1999).

GROWTH OF EMPLOYMENT BY LEVEL OF GOVERNMENT The overall trend of public employment masks differences in the rates of growth at the federal, provincial and municipal levels. The federal government’s share of public employment has been declining for some time, while that of the provinces has expanded and municipal government has remained relatively steady. Table 5.1 showed that the share of the labour force taken by the federal government (general government and enterprises combined) declined from 1960 to 1982. Figures for the next decade for general government only show the federal share of government employment declining from 40 per cent in 1974–75 to close to 31 per cent in 1990–91, while that of the provinces increased from just over 36 per cent to over 42 per cent, and local government went from 24 to 28 per cent (Statistics Canada 1991, Public Sector Employment, p. 19). More complete and revised figures for the years 1991–99 appeared in Table 5.2, since they once again include government enterprises. They confirm the relative decline of the share of the federal government in public employment, which has fallen from 18.4 per cent in 1991, to 15.0 per cent in 1999. The provinces and territories have the largest share of public

Public employment in Canada

115

employment, rising slightly from 50.9 per cent in 1991 to 51.4 per cent in 1999. Local government and enterprises have gained relative ground also, with more than twice the share of the federal government, at 33.6 per cent in 1999. As we shall see, these figures reflect the fact that the provinces are responsible for the most labour-intensive services in health, education and welfare. As well the federal government began to reduce its workforce before the provinces which, in turn, began to cut before local governments did. From 1960 to 1999 the federal share of total public sector employment fell from 28.9 per cent to 15 per cent (although some of this decline is accounted for by the increased statistical coverage of the later period). The federal government sector had its greatest growth from 1965 to 1975, whereas public enterprise employment grew most rapidly from 1975 to 1980. In the 1990s much of the decline of public enterprise employment came from privatizations. The federal government has relatively few programmes that involve the population directly. Looking at the components of its growth in 1979, Bird and colleagues wrote: ‘At least in terms of federal employment, it seems more accurate to refer to the growth of the “police state” than the “welfare state” ’ (Bird et al. 1979, pp. 57–8). Indeed, in the interval, protection and general government services have increased their share from 35 per cent to 66 per cent of federal employment (taking 44 per cent and 22 per cent respectively in 1996). Where the expansion of the welfare state in the 1960s had its personnel impact was particularly in the provincial and territorial governments. The provinces, being constitutionally responsible in the fields of health and welfare, experienced most of this increase, but the division of power between them and municipalities means that employees of primary and secondary schools are considered as being in the local government universe, while welfare employees may in some cases belong there also. Table 5.1 charted this growth for the years 1960–82. The greatest relative growth occurred between 1960 and 1975; the same applies to hospital employees, who are now counted among provincial employees by Statistics Canada. The figures for the 1990s are given in Table 5.3. They show that, at the provincial level, government employment accounts for less than 25 per cent of total employment, health institutions employing just under onehalf, and post-secondary education and public enterprises accounting for just over 25 per cent. Growth has not been uniform across provinces and territories. Rates reflect several factors: size, since larger provinces seem to achieve economies of scale; previous history, since several cases of catching up occur; and finally degree of government intervention, but this is not

116

281.72 138.91 1536.08

281.62 147.03

1556.28

Total

Source:

Total

1502.25

276.15 130.28

360.37 735.45

1994

1491.00

274.80 126.37

358.42 731.41

1995

Statistics Canada Series 72-209-XPB until 1996, afterwards 68-213-XIB.

6.64

Cumulative change 1992–99 (annual average) General government 10.80 Health/ social services 3.87 Universities, colleges, trade institutes 5.84 Enterprises 12.37

371.08 744.37

378.45 749.18

1993

Provinces/Territories General government Health/social services Universities, colleges, trade institutes Enterprises

1992

Table 5.3 Provincial and territorial public employment (000s) in the 1990s

1455.44

272.61 124.54

341.38 716.91

1996

1421.38

270.74 119.69

338.71 692.24

1997

1430.98

268.99 122.76

335.19 704.04

1998

1448.07

265.18 125.14

337.57 720.18

1999

117

Public employment in Canada

Table 5.4 Percentage rates of change in provincial and territorial government employment, 1992–9 General government

Health and social services

Universities and colleges

All government employmenta

Newfoundland Prince Edward Island Nova Scotia New Brunswick Quebec Ontario Manitoba Saskatchewan Alberta British Columbia Yukon Territory Northwest Territoryb

21.16 3.89 18.20 3.54 6.59 18.38 6.58 7.84 32.07 5.66 10.26 23.55

4.19 13.01 7.34 0.36 1.63 11.60 8.25 0.86 13.67 1.58 1238.0 32.43

1.78 3.50 3.87 93.92 8.85 12.90 0.97 9.59 3.38 14.60 – –

5.27 2.28 9.70 1.09 3.98 13.68 1.99 4.25 16.76 5.83 19.01 25.80

Average

11.32

4.94

6.01

6.64

Notes: a All Government Employment  first three columns combined. b The figure for the Northwest Territory is lower because of the creation from its eastern territory of the new territory Nunavut, which employed 1964 people in 1999. Source: Statistics Canada, Public Sector Employment and Wages and Salaries. Series 72-209-XPB 1998, and Public Sector Statistics 68-213-XIB, April, 2001.

automatic. Bird and colleagues (1979, p. 54) found that from 1961 to 1975 the small maritime provinces had the highest rate of growth, but this only brought them close to the national average. The same was true for Quebec, which during and after its Quiet Revolution had been engaged in the 1960s in ‘rattrapage’ with respect to the other provinces. One surprise was that the Saskatchewan public sector experienced the lowest growth rates, in spite of its being under socialist governments for half of these years. This trend was also evident in financial matters, where the socialist governments of the CCF and then the NPD practised fiscal restraint and budget deficits were the result of Conservative governments. Sutherland and Doern (1986) found that the most rapid growth in a provincial public sector occurred in Alberta, while Ontario and Quebec, the largest provinces, had the slowest growth rates. A look at provincial and territorial government employment during the downsizing of the 1990s (Table 5.4) shows that the rates of change varied

118

The state at work, 1

considerably around the average reduction of 6.64 per cent from 1992 to 1999. The leader in cutting jobs is the government of Alberta, whose Conservative Prime Minister Ralph Klein was the first head of government in Canada to make eliminating the deficit a priority after his election in 1992. Next comes another Conservative premier, Mike Harris, elected in Ontario in 1995. Among the larger provinces, prior to the 1995 referendum on Quebec sovereignty, Quebec was not an eager downsizer, probably because the Parti Québécois government of Jacques Parizeau was reluctant to alienate its supporters among public employees. After 1996 his successor made eliminating the deficit a priority, which was achieved in 1998. The growth of large and wealthy British Columbia’s government sector stands out as an exception. Two factors are probably relevant in this case: from 1991 the social-democratic party was back in office under Premier Mike Harcourt, and British Columbia came into the 1990s following a series of surpluses. The movement in the other, smaller provinces was less consistent, and the tiny territorial services grew quite rapidly. Two trends about downsizing in the provinces should be noted. First, the overall cutting continued until 1998, after which time it increased modestly by less than 2 per cent in two years. Moreover, the rate of decline varied greatly among components of provincial public sector employment: the deepest decrease occurred in public enterprises, followed closely by general government services. The decentralized fields of higher education and health and social services, with many more employees, suffered much less decline. Local government occupies more than twice the share of public sector employment than does the federal government. Local governments experienced a rapid growth in employment in the years 1970–75 (Siegel 1988, p. 178). The most recent figures (Table 5.5) show that local government and municipal enterprises underwent modest downsizing from 1993 to 1997, with a slight increase since. Local or municipal government figures have declined slightly, but local enterprises have lost over 10 per cent of their employees. What keeps local government employment figures up are those for school boards, that offer primary and secondary education in Canada. With the exception of New Brunswick, where teachers and school employees are employees of the provincial government, school boards are unifunctional local governments with their own elected governing bodies. Their autonomy is nowhere near as great as that of municipalities, however, since they rely heavily on provincial grants (as well as on property taxes) and they fall under the overall supervision of a single government department that controls curriculum, teacher certification, acceptable texts, teaching load and so forth. In many ways their status is closer to that of the hospital employees, who are counted by Statistics Canada among

119

Source:

52.14

52.77

956.92

Total

960.59

51.53

543.03

366.03

1994

956.01

48.70

540.94

366.37

1995

952.64

47.15

548.21

357.28

1996

Statistics Canada series 72-209-XPB annual, until 1996, then 68-213-XIB.

962.04

541.09

368.81

1993

533.61

370.54

1992

938.86

46.43

542.00

350.43

1997

938.35

46.79

546.38

345.18

1998

Local governmental public employment in the 1990s, annual average (000s)

Local government School boards Municipal enterprises

Table 5.5

941.12

47.41

542.99

350.72

1999

1.65

10.15

1.76

5.34

Cumulative change

120

The state at work, 1

provincial employees. Without the education employees, local governments have followed closely the leads of the federal and provincial governments in cutting back. Presence of Public Sector Employees by Area Turning to the presence of public sector employees of all levels in the provinces, Sutherland and Doern (1986) found that they were distributed according to population, except in the largest province, Ontario, which had less than its share. For the 1990s, recent figures show that the largest provinces, Ontario, Alberta, British Columbia and since 1999, Quebec, have a larger share of national population than they have of public employment (Statistics Canada 2001, Public Sector Employment). The others have a larger share of public employment than of population. As a general rule, they are the smaller and poorer provinces. Ontario and Alberta are large, rich and governed by right-wing governments. British Columbia has only very recently begun to cut its employment, as it was governed in the 1990s by a social-democrat government and came to budgetary deficits later than the others. Quebec is the exception, since it is the second largest province in population. It has a number of structural economic problems, including higher long-term unemployment and lower wage levels than the national average, and especially than those of its immediate neighbour, Ontario (Noel 1993). Since 1960 Quebec has developed a unique model of government in Canada, sometimes called Quebec Inc. This involves greater state intervention than elsewhere, a separate pension fund, greater use of public corporations, wider labour union rights in an enlarged public sector, and higher wages and benefits than in most provinces or the federal government. Even so, in 1997, in pursuit of a zero deficit, the Parti Québécois government conducted a massive buyout of public sector employees over the age of 55, which led to 37 000 departures. With respect to the presence of public employees in Canadian cities, we do not have consolidated figures. The size of the country has long meant that federal employees were dispersed throughout it, only around one quarter working in the Ottawa–Hull national capital area. Figures for 1996 show that, among government employees only, almost 40 per cent were in the national capital region, Ottawa–Hull (Statistics Canada 1998, Public Sector Employment, p. 53). The fact that some provinces (Quebec, Alberta and British Columbia) do not have their capitals in their largest cities should induce some deviance from a simple population correlation.

Public employment in Canada

121

CENTRAL GOVERNMENT The federal government, as we have seen, has lost ground progressively relative to provincial and local governments to the point where today it accounts for only about 15 per cent of total public employment in Canada. Nevertheless, it remains the most important public employer, not only because it is the largest, but because it has been the leader in introducing public personnel reforms since at least the beginning of the 20th century. The Civil Service Commission was created in 1908 to deal with recruitment to the inside civil service (those working in Ottawa). In 1918, as a product of a coalition wartime government under Sir Robert Borden, its jurisdiction was extended to the entire civil service and its autonomy expanded. It has been the champion and the guardian of the merit system since that time. Most of the provinces followed this lead in the following decade. By virtue of its size, the federal government has led in the process of modernizing the machinery of government for purposes of supporting the executive and assisting coordination across general government functions. This development began in the Second World War. Then the expansion of Canadian government activities in the post-war years (particularly from 1960), and the introduction of collective bargaining in 1967 led to a much more bureaucratic administrative machine than the federal service had previously known (Granatstein 1982). The structure of the Canadian government, like that of the British, has become quite complex since the Second World War. In part, this has been because of the possibility for the Cabinet (and thus the Prime Minister) to create and transfer units without legislation (Tardi 1992). While the only department without a statutory basis is that of the Prime Minister (the Privy Council Office), transfers of function are regularly made between departments without statutory authorization. In the past, also, several public corporations were created by letters patent, a number of which were quite important (e.g. Atomic Energy of Canada Limited). Another reason for this complexity was the establishment, in 1951, of a tripartite classification of Crown corporations (departmental, mandatory and proprietary) only one of which had the functions and the autonomy of real commercial or industrial corporations. Departmental corporations were simply units enjoying a certain autonomy within a government department. Finally, Crown corporations were able to create subsidiaries, joint ventures and other subunits, so that their exact number was unknown even to the government, prior to corrective legislation in 1984. Sutherland and Doern (1986) describe federal government organizations as belonging to one of three kinds: primary, or classic departments headed by a minister, depending on parliament for their funding and accountable

122

The state at work, 1

to parliament directly through the minister; secondary, being agencies, boards, commissions and public corporations that have more or less autonomy from ministerial supervision in order to implement relatively clear and stable mandates; tertiary, being agencies of planning, coordination and control for the entire federal administration. Tertiary administrations divide in turn into two types. The larger number are relatively autonomous bodies with advisory, consultative or control functions, such as the Canadian Human Rights Commission. Some of them are watchdog agencies on behalf of parliament: the Auditor General of Canada, the Public Service Commission, the Official Languages Commissioner, the Privacy and the Access to Information commissioners. A smaller group of tertiary organizations are central agencies, that provide support directly to the government or to departments in its name; it includes the Privy Council Office, the Treasury Board Secretariat, for some functions the Department of Finance and some functions of the Public Service Commission. In comparison to Britain, the Canadian structure of executive support agencies has been more complex in recent years: throughout the 1980s we had 6.5 agencies to two in Britain; since 1993 the number is 4.5. Since budgetary control became a preoccupation at the beginning of the 1990s, the government has abolished about 40 small agencies, including most of the advisory bodies, such as the Economic Council of Canada, the Science Council, and the Law Reform Commission. The emphasis has clearly been on control and not on generating new programmes, at least until a budgetary surplus was achieved in 1997–98. When looking at downsizing, it will be interesting to see if the hypothesis of Herman Schwartz (1994) is correct, that central agencies (‘fiscal bureaucrats’) do not cut their own staffs to the same extent that they cut operating departments and agencies. The core of the government is made up of the government departments that are directed by ministers, all members of the House of Commons. Most government agencies from the secondary group report to the House of Commons via one of these ministers. The demands of regional representation in Cabinet have meant that it has in the past been much larger than was necessary for effective government. As part of the streamlining measures of the 1990s, Progressive Conservative Prime Minister Kim Campbell managed, during her brief term in office in 1993, to reduce the number of ministers from 39 to 25, and the number of departments from 32 to 23 (Aucoin 1995, p. 120). Under Liberal Prime Minister Jean Chrétien the Cabinet grew back to 33 members in December 1999, but the number of departments has continued to decline to 20. One reason for this discrepancy is the creation of a number of Special Operating Agencies. All told, there are 20 Special Operating Agencies, the largest being the Canada Customs and Revenue Agency, created in

Public employment in Canada

123

November 1999, to replace the former Department of Revenue and made up of 44 000 public servants. There has been some ambiguity about SOAs: they were inspired by the British agencies of the Next Steps programme, but they have never been given the autonomy that the British agencies have (Aucoin 1995, p. 146; Savoie 1999, p. 216). Agencies remain part of the classic structure in that they report to the deputy minister of their originating department, and they have limited powers to manage their personnel as they wish. They fall under all government-wide ‘decisions on budget cuts, salary freezes and . . . restrictions of management numbers and levels’ and by department-wide policies also (Rogers 1996, p. vii). The fear of disorder or anarchy that public corporations posed two decades ago has now resurfaced regarding autonomous agencies and new collaborative arrangements that the federal government has with other governments and non-governmental organizations. In recent years, at least six chapters of the Auditor General of Canada’s annual reports have had as their subject, accountability in non-departmental structures that either deliver programmes themselves or work with other governmental and nongovernmental actors to deliver programmes and services. In Chapter 5 of his annual report for 1999 the Auditor General claims that ‘Citizens may not have the same recourse to the minister and members of parliament in these cases [of shared responsibility]’. In a follow-up report to the House of Commons in November 1999, the Auditor found that 25 departments and agencies had set up 77 arrangements of shared responsibility, while the government had no central inventory nor even an overall framework and no rules about reporting to parliament. The overall trend of federal public employment has been seen in Tables 5.1 and 5.2. However, because it only divides general government employees from those of public corporations, it mixes those who fall under the Public Service Act and those who have other status. Table 5.6 gives more detail for the 1980s for the three ‘universes’ mentioned in the annual reports of the Public Service Commission. The first universe is that of employees covered by the Public Service Employment Act (PSEA), who fall under the jurisdiction of the Public Service Commission (before 1967, the Civil Service Commission). This is the core civil service with which we deal below. Here let us note only that the Progressive Conservative government of Brian Mulroney was successful from 1985 to 1989 in reducing the public service by about 3.9 per cent, but that the upward trend had returned in their second mandate with the difficult years of the early 1990s, so that the overall decline of the public service from 1985 to 1991 was less than 2 per cent. The second, bigger universe takes in an additional number of employees who also have the Treasury Board (the government’s management

124

Table 5.6

The state at work, 1

The three federal government employment ‘universes’ 1983

1985

1987

1989

1991

Change (%)

Public Service Commission 222 044 223 173 214 930 214 494 217 818 Treasury Board 13 033 14 079 13 160 13 630 17 109 Statistics Canada Government enterprises 207 122 212 622 204 817 164 031 155 828 Military personnel 86 381 87 017 87 866 88 993 86 319 Royal Canadian Mounted Police 18 603 18 428 19 735 19 163 19 521 Other corporations/ agencies 33 759 32 760 31 985 34 032 33 109

24.77 0.07

Total

245 865 350 827 344 403 306 219 294 777

14.77

Fed. government employment

580 942 588 138 572 493 534 343 529 704

8.82

1.90 31.27

1.93 1.93

Note: Figures are for December of each year. Source: Annual reports of the Public Service Commission.

committee) as their employer, but whose positions do not fall under the PSEA. They are mostly occasional or temporary employees. Although their number jumped by almost 30 per cent from 1983 to 1991, they represented a relatively small group, less than 10 per cent of the Public Service Commission universe. The third universe, that of Statistics Canada, includes employees of autonomous agencies that neither fall under the Public Service Commission, nor have the Treasury Board as their employer. The figures for the armed forces, the RCMP and a number of non-commercial, nonindustrial corporations and agencies remained stable or showed slight increases through the 1980s. The principal cause of the 9.48 per cent decline of public employment from 1983 to 1991 is the privatization programme of the Conservatives. From 1984 to 1992, the Mulroney government had 24 privatizations, of which 15 had assets of over Can$100 million and 12 brought in prices of over $1 million (Stanbury 1994, pp. 204–5). The result, as seen in Table 5.6, is that employment in public enterprises declined 24.77 per cent during the 1980s. For the 1990s, Table 5.7 shows the further decline of the importance of public enterprise employment. From 1992 to 1999, employment in federal

125 327 474

136 429 138 910 52 135

1993

325 196

143 784 130 281 51 131

1994

308 936

135 763 126 371 46 802

1995

272 697

101 210 124 542 46 945

1996

259 426

93 311 119 690 46 425

1997

260 903

91 359 122 758 46 786

1998

262 452

89 903 125 135 47 414

1999

22.91

36.08 14.89 –10.15

Change (%)

Statistics Canada Series Public Sector Employment and Wages and Salaries, annual, until 1994, then Public Sector Statistics 68-213-XIB.

340 448

Total

Source:

140 647 147 030 52 771

1992

Employment in public sector enterprises, 1992–9

Federal Provinces/Territories Local

Table 5.7

126

The state at work, 1

public enterprises declined by 36.08 per cent, with the largest decline being in 1996, with the sale of Canadian National Railways and Petro Canada. Thus, since 1983, federal public enterprise employment fell by just over 57 per cent, in shares roughly divided between the Progressive Conservative government of Brian Mulroney (1984–93) and the Liberal government of Jean Chrétien (in the years 1993 to 2001). The provinces and municipalities have followed the trend but in proportions of 15 per cent and 10 per cent respectively, and have even seen some increase in the last two years of the 1990s. Stanbury considered all privatizations in Canada to have been modest in scale, those of provinces even more so than those of the federal government (Stanbury 1994, pp. 171–3). Even so they account for the largest share of employment reduction in the 1990s, and the share of public enterprises in total public employment in Canada has fallen from 11.5 per cent in 1991 to 9.4 per cent in 1999 (Table 5.2). In closing, a good example of the value in using a large interpretation of the public sector as we have in this volume comes from the Canadian Post Office. It was transformed from a department into a public corporation in 1981. The public service (universe of the PSC) lost 62 000 members overnight, while the public enterprise section of the Statistics Canada universe gained the same number. No real change, but the impression was given that the public service had shrunk considerably. In the Canadian system, core personnel are usually considered to be those working for departments and agencies falling under the jurisdiction of the Public Service Commission. This is the world of the career service, appointed by competitive competition, protected from the arbitrary impact of changes of party in power. It is thus the government bureaucracy, but it is also the branch of the administration directly responsible to parliament, by the means of a minister. Finally, as this branch is almost entirely dependent upon funds voted by parliament, it falls into the classic model of parliamentary control of public expenditure. This Public Service Act universe has the most complete statistics, going back to 1912. Recent changes in legislation, however, have led to the abandon of this longstanding series. The agency that acts as the government employer and manager, the Treasury Board, began a replacement series in 1993, but it is not identical, so we are once again obliged to present two overlapping series in order to get some overall picture. Table 5.8 gives these figures for the years 1985–99. They are more complete than those of the Public Service Commission. The latter’s universe appears to cover only the first three columns of Table 5.8, that is, continuing or indeterminate employees (full and part-time) and term employees whose contract was over three months (over six months before 1994). The indeterminate group are the Canadian equivalent of the career service. What Table 5.8 includes

127

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

204 747 202 234 196 287 191 400 188 865 189 653 192 905 192 532 195 014 192 152 187 851 172 968 158 107 150 086 146 774

N

83.88 83.41 82.34 80.76 79.22 79.12 80.08 79.24 81.17 83.04 83.26 83.17 81.30 80.18 78.78

%

Full-time indeterm.

3 353 3 649 4 043 4 290 4 236 4 339 4 496 4 661 4 811 4 996 5 399 5 104 4 711 4 410 4 476

N 1.37 1.51 1.70 1.81 1.78 1.81 1.87 1.92 2.00 2.16 2.39 2.45 2.42 2.36 2.40

%

Part-time indeterm.

15 543 16 683 16 353 17 703 19 076 20 223 19 921 21 048 21 289 26 066 23 051 21 554 22 267 24 186 25 923

N 6.37 6.88 6.86 7.47 8.00 8.43 8.27 8.66 8.85 11.26 10.22 10.36 11.45 12.92 13.91

%

Term over 3 monthsa % N

%

Seasonal

N

%

Casualb

N

% 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Total

19 166 7.85 1 280 0.52 n.a. – 244 089 18 580 7.67 1 248 0.51 n.a. – 242 394 20 418 8.57 1 279 0.54 n.a. – 238 380 22 137 9.34 1 457 0.61 n.a. – 236 987 24 719 10.37 1 519 0.64 n.a. – 238 415 23 905 9.97 1 588 0.66 n.a. – 239 708 21 884 9.08 1 697 0.70 n.a. – 240 903 23 106 9.51 1 611 0.66 n.a. – 242 958 17 709 7.36 1 639 0.68 n.a. – 240 462 6 094 2.62 1 426 0.62 666 0.29 231 400 4 739 2.10 1 483 0.66 3 096 1.37 225 619 2 550 1.23 1 383 0.66 4 418 2.12 207 977 3 026 1.56 1 293 0.67 4 992 2.57 194 396 2 512 1.34 1 149 0.61 4 844 2.59 187 187 2 868 1.56 1 165 0.63 5 108 2.74 186 314

N

Term 3 months or less

Table 5.8 Federal public service employment, 1985–99

241 170 239 287 235 125 234 007 235 237 236 486 237 713 239 668 237 202 228 318 222 362 204 923 190 979 183 942 183 068

Total FTEc employment

128

N

N

7.21 1.63 21.73

%

Part-time indeterm.

Indeterminate

%

Full-time indeterm.

(continued)

N

21.16 27.12 3.59

N

%

Term 3 months or less

Term

%

Term over 3 monthsa %

18.67 10.21 21.44

Seasonal

N

Seasonal

%

– – 64.99

Casual

N

Casualb

Total 2.32 3.47 17.42

N

Total

%

2.46 3.45 17.67

FTE

Total FTEc employment

Treasury Board, Employment Statistics for the Federal Public Service, annual.

Prior to 1994, terms employees were either ‘over six months’ or ‘six months or less’. Casual Employment was introduced by legislation in 1993; it involves those hired by departments for no more than 90 days, with possible extensions up to a limit of 125 days in a given 12 months. FTE means full-time equivalent.

Source:

c.

a. b.

Percentage change 1985–89 1990–94 1995–99

Table 5.8

Public employment in Canada

129

are other categories of employment, short-term, seasonal and casual. New totals are given for total employment and FTE or Full-Time Equivalent. FTE is a smaller number than a head count because the numbers of people working part time are treated as fractions of an FTE, so that two half-time jobs become the equivalent of one full-time job. Table 5.8 also gives in summary form the changes in these different categories since 1985. These show that the reduction efforts of the latter half of the 1980s were concentrated in the career group and that additions of various term, seasonal and casuals caused the overall impact to be much smaller (just over 2 per cent). From 1990 to 1995 the situation was almost the opposite, with the career service remaining stable and the non-career service declining. The biggest changes have occurred since 1995, the year that the Chrétien government’s Programme Review began. Although our figures only cover four years, they show how deep have been the cuts in these years, with more than one-fifth of the career employee positions having been cut. Of the 64 729 people who left the Canadian public service during these years, roughly 55 per cent took advantage of various early retirement or bonus schemes, 27 per cent were due to natural attrition, and 18 per cent to devolutions and privatizations. These were very large cuts, greater than any post-war government had made. However, they were concentrated in certain departments only (Treasury Board 1999, Employment Statistics, pp. 62–3). Most of them are or were departments and agencies actually doing things for the public or other departments. Five departments were responsible for over 91 per cent of the decline from 1995 to 1999: the civilian branch of National Defence, Transport, Human Resources Development, Agriculture and Agri-Food, and Public Works and Government Services. The Transport department, before becoming an agency, lost a large number of employees when air traffic control was privatized to a non-profit company composed of the same personnel. Human Resources transferred its manpower training personnel to the Quebec government under an agreement of 1997. If we look at the most prominent central agencies, we see that with one exception, they were cut lightly, or not at all. The most affected departments (MADs) lost an average of 31.2 per cent of their employees from 1995 to 1999. The Privy Council Office, the Prime Minister’s department, has increased its staff by 29 per cent since 1993, and by 3.67 per cent since 1995, under the Programme Review. The Treasury Board Secretariat suffered almost a 5 per cent loss since 1995, but it is up 10.5 per cent since 1993. The Finance Department is down 8.14 per cent since 1993 and 1.67 per cent since 1995. The only central agency to suffer a fate similar to most MADs was the Public Service Commission, with a decline of 38 per cent since 1993 and of 29 per cent since 1995. In terms of Herman Schwartz’s

130

The state at work, 1

(1994) analysis, the PSC staff were not ‘fiscal bureaucrats’ who might be expected to protect their own, while cutting elsewhere. The PSC has been moving away from its combined role as central management agency and as audit and control unit, in favour of the latter role. So there is some support for this hypothesis. The results of the cutbacks in terms of occupational categories give us further information about their nature. As Table 5.9 shows, for an overall decline of 22.52 per cent, the technical and implementing categories suffered much more than their share. The Operational category (bluecollar), has been declining for many decades (Sutherland and Doern 1986; Siegel, 1988, p. 166). The technician category has declined rapidly also, but in terms of sheer numbers, the clerical group, the Administrative Support category has declined the most, about 25 000 jobs or almost 35 per cent. The Treasury Board notes that there has been a migration of people from Administrative Support to the Administration and Foreign Service category. This represents an upgrading of these people, since Administration and Foreign Service along with the Scientific and Professional Category constitute the professional public service, those with post-secondary education. Another reason for the decline of the Administrative Support numbers is the contracting out of these jobs to temporary help agencies. While not among the most severely cut, nevertheless the Executive category, at the summit of the public service, has lost almost 18 per cent of its members from 1993 to 1999. In relative terms it lost ground to the other categories until 1997, but it has regained all of the lost ground and even increased in absolute numbers from 1998 to 1999. The federal public service has thus changed in the direction of an upgrading of its qualifications. Those who executed its policies have lost ground, while the planners, coordinators, controllers and managers have gained. Bird and colleagues (1979, pp. 64–5) noted that the public sector in Canada employs over two-thirds of all professionals (as defined above). They said that the federal government had the greatest share of managers and clerks, while the provinces had the most professionals. This is no doubt still true of the provinces, since they employ professionals in health, education and welfare. However, Table 5.9 suggests that the federal government is cutting back on clerks and upgrading its public service. Public servants increased in number in the National Capital region in 1998–99, while those in the other regions declined. The change is particularly notable in the professional categories, two-thirds of whom work in the capital region (Treasury Board 1999, Employment Statistics, p. x). While this may corroborate the Public Choice theory, it also fits in with the Public Management doctrine, promoted by Osborne and Gaebler (1993) and widely adopted at Ottawa, that government should ‘steer, not row’. Thus

131

Source:

Note:

Treasury Board of Canada, Employment Statistics for the Federal Public Service, annual (Ottawa, Treasury Board).

Figures are for March, taken from total employment.

75 424 40.48 15 922 8.55

1.02 40.86

9.88

22.51

17.66

240 454 100.0 231 391 100.0 215 525 100.0 207 917 99.98 194 394 99.81 187 183 99.99 186 313 99.98

49 721 26.56 22 633 12.09

72 816 38.90 16 180 8.64

1.84

%

23 027 12.36

3 421

N

Change 1993–9 %

Grand total

53 920 27.74 24 708 12.71

70 430 36.23 18 504 9.52

1.71

%

22 630 12.09

3 203

N

1999

34.70 43.90

28.06 56 771 27.30 13.98 28 043 13.48

33.40 71 670 34.46 11.53 23 933 11.51

1.68

%

23 574 12.13

3 258

N

1998

48 232 25.89 20 287 10.89

29.25 63 307 14.39 31 540

1.61

%

1997

30.72 67 671 15.04 33 292

32.12 75 359 11.45 26 007

3 339

N

11.38 24 161 11.62

1.65

%

1996

73 864 36 159

3 735

N

11.13 25 671

1.68

%

1995

30.69 74 317 11.20 26 488

3 878

N

10.63 25 745

1.73

%

1994

73 802 26 923

25 551

4 155

N

1993

Occupational categories in the federal public service, 1993–9

Executive Scientific and professional Administration and foreign service Technical Administrative support Operational

Table 5.9

132

The state at work, 1

privatizations, contracting-out and devolution have led to the decline of the rowers, or implementors.

PUBLIC EMPLOYMENT BY POLICY AREA AT ALL LEVELS OF GOVERNMENT When the three levels of government are combined, a few functions stand out as employing the greatest number of employees. Education and health each account for over one-quarter of public employment in Canada (28.56 per cent and 26.77 per cent respectively in 1995). Protection of persons and property come next, each absorbing more than one-eighth (12.8 per cent and 12.6 per cent in 1995). Social services, at 4.9 per cent in 1995 and transportation and communications, with 3.9 per cent, constitute the other important employers. Only one of these functions has a direct economic content, transportation and communication, a reminder of the tradition of public infrastructure activity by governments in Canada. It is probably lower in importance today, following the privatization of air traffic control in 1995. The other significant functions concern protection or social functions. These overall figures reveal much about the Canadian state, but they mask specialization by level of government. At the federal level, in 1996 protection dominates with 44.4 per cent of all employment, followed by general services at 21.66 per cent. In the provinces and territories, health dominates with 43 per cent, followed by education (post-secondary) at 23 per cent, and social services at 15 per cent. Less than 4 per cent of provincial public employment goes to general services. Local level public employment is overwhelmingly dominated by primary and secondary education in local school boards (59 per cent in 1995). The only other function of importance is general government at 17.5 per cent. As we said earlier, this picture is skewed somewhat, because provincial governments keep a very tight control of school boards, so that, in many ways, they are in the same situation as health services, that appear so important in provincial employment.

ASPECTS OF SOCIAL DIFFERENTIATION Legal Status As we saw earlier, the Canadian government has been using a tripartite distinction among those who fall under the Public Service Employment Act (and the Public Service Commission), those who have the Treasury Board as employer, but do not fall under the PSC, and those who have a separate

Public employment in Canada

133

employer, such as the armed forces, the RCMP and the public enterprises. What is thought of as the career public service is composed of the employees in the first group. They are the ones who work for a minister responsible to the House of Commons, in units that receive their funding from money voted by parliament. The provinces also make this distinction. Table 5.8 also presented employment types as shares of the federal public employment. There are six types of jobs in Table 5.8: the indeterminate (full time and part-time) are the career service with full rights of public servants. Term employees are those with a contract of fixed duration with their department or agency; their terms are divided into those that are three months or over, and those that are under three months. Seasonal employees work, as the name suggests, during a specific time of year. They may be indeterminate or, more likely, term employees. The final category, casual employees, was introduced in 1993. They may not work in the same department for more than 125 days in any 12-month period. While temporaries, seasonals and casuals have been used since the civil service system began to escape the rigidities of the career merit system, only the new casual category gives managers freedom to manage their staffs. Both temps and seasonal workers have to be referred to the departments and agencies by the PSC. Casuals alone are hired from departmental pools of candidates (Gow and Simard 1999, p. 78). To understand Table 5.8, we need to know also that the temporary categories were divided by six-month periods until 1993. Only then were they made more or less than three months. So the abrupt decline in the short time-term appointments after 1993, and the rise of casuals appears to show managers switching either to longer-term appointments (to catch those who were needed for close to six months), or to the freer casual appointments. If we look only at the evolution of these categories since 1995, we would conclude in the gradual decline of the career service, a full 51/2 per cent since 1995. However, looking back to the late 1980s, the decline is only less than one-half per cent. The Treasury Board (1999, Employment Statistics, p. viii) noted that there is a cyclical movement to these appointments and that the 1999 figures mark the fourth time in 20 years that temporaries and casuals counted for more than 18 per cent of the federal public service. The casual group appear to be the real novelty here, since they are completely outside the merit system. Their numbers are small, but the growth has been very rapid. One reason for the fluctuations of the numbers of temporaries and casuals in the past is that many of them were doing the work of permanent employees. Their contracts would be renewed from time to time (often in violation of the rules) and pressure would mount for their integration into the career service. So temporaries and casuals were integrated, either by

134

The state at work, 1

individual competitions, or by occasional group integrations. In fact, temporary personnel represent ‘an old solution to an old problem. From the pre-war period to the 1990s, temporary personnel have been an answer to emergency situations, such as war, or other types of crisis, cutbacks of “permanent” staff and cumbersome hiring procedures’ (Gow and Simard 1999, p. 77). Only indeterminate employees have all the benefits of public service status. Term employees have more benefits than casuals, the longer terms more than the shorter. Casual employees have paid public holidays, sick leave, unemployment insurance and severance pay, but they do not have annual leave, union membership or a pension. In recent years, young people have been over-represented in this group, and almost absent from the career group. Whether or not the present situation indicates a deep move away from the classic career model to a new precarious, flexible workforce will have to be seen from a longer perspective. Gender and Ethnicity The introduction of a system of recruitment and selection based on individual achievement went largely unchallenged in the federal public service until the 1960s. There was some pressure in the 1930s to obtain services in the French language for French Canadians, but it did not lead to substantial changes in policy until the 1960s. The rise of a nationalist sentiment, and even of a separatist movement in Quebec, led the Federal government to appoint a Royal Commission on Bilingualism and Biculturalism in 1963. The commission’s preliminary report in 1965 sounded the alarm bells and the final report, in 1969, led to the adoption of the Official Languages Act in 1969. The law offered service in either official language wherever numbers warranted it, and created different categories of language requirements for public service jobs: unilingual English or French, and bilingual. The 1960s being a time of increased awareness and organization of women, the federal government followed the same path. A Royal Commission on the Status of Women reported in 1970, and the first positive action measures were taken in 1973. These were much less demanding than those concerning language, and women’s representation progressed much more slowly. Measures became gradually more coercive and in 1984 the government adopted ‘targets’, that it refused to call quotas, although many considered the difference cosmetic. Tables 5.10 and 5.11 show the results of these programmes in the 1990s. With 24.6 per cent of the Canadian population, francophones are overrepresented in the Canadian public service, with 30 per cent of all positions (Table 5.10). Their participation is slightly less in the two higher echelons,

135

Public employment in Canada

Table 5.10

Francophones and occupation in the federal public service (%) 1993

1994

1995

1996

1997

1998

1999

Executive Scientific and professional Administrative and foreign service Technical Administrative support Operational

22.9

22.8

23.2

24.3

25.4

25.6

29.0

22.6

22.8

23.9

24.1

23.9

23.9

25.0

31.6 21.5 34.0 23.1

29.6 21.5 32.8 22.2

30.6 21.4 33.1 22.3

31.6 22.0 34.0 23.1

31.3 23.4 34.3 22.6

31.2 24.2 34.4 20.8

32.0 25.0 36.0 24.0

Grand total

28.0

27.5

28.0

29.0

29.2

29.2

30.0

Source: Commissioner of Official Languages, Annual Reports, (Ottawa: Public Works and Government Services) and Treasury Board of Canada, Annual Report on Official Languages 1998–1999 (Ottawa, Treasury Board, 1999).

executive and scientific and professional, but they are at least proportionately represented in all categories. This is quite a reversal, because before the official language programmes were introduced, the complaint was often heard that French Canadians were largely confined to the lower-paying menial jobs. Where the battle is far from won, however, is in actual conditions of work and service. Each year the Official Languages Commissioner draws attention to problems of this kind in his annual report. The presence of francophones is concentrated in a few areas: the national capital region (41 per cent), the province of Quebec (94 per cent) and, to a lesser degree, in New Brunswick (38 per cent). Elsewhere, it varies from 2 per cent to 6 per cent. For a long time women’s organizations cited the progress of French Canadians as a model they would like to emulate. Table 5.11 shows that they have progressed to the point of being slightly over-represented in 1999. They had farther to come. In 1967, women comprised 27 per cent of the public service and held only 1 per cent of executive positions. While they remain far from being equally represented at that level, their progress is constant and stands at 26.9 per cent in 1999. The same is true for participation in the scientific and professional and administrative and foreign service, the two groups that serve as feeders to the executive category. They are still dominant in the clerical administrative support category, but, as we saw earlier, this group is one of the most reduced in recent years. It experienced a considerable migration of its members to the administrative and foreign service category.

136

Table 5.11

The state at work, 1

Women and occupation in the federal public service (%)

Executive Scientific and professional Administrative and foreign service Technical Administrative support Operational Grand total

1993

1994

1995

1996

1997

1998

1999

17.6

18.3

19.1

21.3

23.0

25.1

26.9

28.9

29.2

29.9

30.6

31.6

32.6

33.6

47.1 18.3 83.1 14.5

48.0 18.6 83.5 14.0

49.0 19.2 84.1 13.8

50.1 20.2 84.4 13.6

50.8 23.8 83.8 14.5

52.1 25.4 84.0 14.8

53.4 27.6 83.9 15.8

47.6

47.5

47.8

48.4

50.0

51.0

52.1

Source: Treasury Board of Canada, Employment Statistics for the Federal Public Service.

While women have made great progress in the federal public service, their position is still far from as favourable as that of men. They compose 60–70 per cent of full time temporaries and casuals in 1999, and 83 per cent of part-time employees. In 1997, over 80 per cent of women employees earned less than $45 000 per year, while for men the figure was close to $55 000. Some of this is due to difficult access to the best-paid jobs, but it is also a matter of age. In 1997 women were over-represented in the under-40 age group, but under-represented at 45 years and older. In 1998 the federal government lost a massive pay equity case before the Canadian Human Rights Tribunal. Having also lost in appeal in the Federal Court, the government has negotiated a settlement with the Public Service Alliance of Canada, the union federation that brought the complaint. Some 230 000 present and former employees, largely women, will receive a total of Can$3.6 billion.

WHAT DOES WORKING FOR GOVERNMENT MEAN? The Canadian public sector has been under attack from conservative or neoliberal governments since the mid-1980s, as over-privileged, inefficient and ineffective. The rhetoric of the New Public Management movement has had its impact here, and downsizing, contracting out, privatization, partnerships and increased accountability for results have been introduced as a result. Our brief survey shows that growth of public employment was greatest in areas popular with the public, health and education, as it was in other industrial countries (Rose 1985, p. 10). Bird and colleagues (1979,

Public employment in Canada

137

pp. 127–32) concluded that the public sector had grown partly for demographic reasons, but above all because of political factors, the bureaucratic state creating a ‘huge web of interdependence’ among the middle classes (p. 132). The downsizing has occurred, however, most notably at the federal level and in the large and wealthy provinces of Alberta and Ontario. Sutherland noted in 1987b (p. 74) a ‘mood of scorn and distrust of public bureaucracies had developed’, which she found unwarranted. In several cases, union rights have been curtailed during this process in the 1990s. The result has been discouragement and disillusionment for many public servants. Having first attacked the federal bureaucracy and threatened to give federal public servants ‘pink slips and running shoes’ (Zussman 1986) the government of Brian Mulroney tried to restore public service morale with a study group and policy paper called Public Service 2000. However, the Programme Review of Jean Chrétien’s Liberals swept its humanistic rhetoric away. The need for a new moral contract led the Liberals to launch, in their turn, an operation called La Relève, that has evolved into the Leadership Network. Whether these initiatives can restore public service morale is not clear. Senior executives have, in the past, been motivated by a sense of public service and the intrinsic satisfactions of their work. In this respect they differed from their private sector counterparts, who were more motivated by the extrinsic rewards of pay, promotions, advantages and the like (Zussman and Jabes 1989). If private sector type motivation should become the norm in the public sector, we should expect to see an increased demand for autonomy in personnel matters, by agencies and other bodies, in order to offer more competitive conditions in comparison with the private sector. For the main group of employees, they have begun to exercise once more the union rights that were suspended, usually by legislation, for more than half of the 1990s. With returning budgetary surpluses, the unions are wanting to make up for lost revenues. It is unlikely, however, that the conditions of full bureaucratic security and advantages can be recovered. The career service may continue to decline in relative numbers, but if the past is any indication, not in dramatic fashion.

NOTES 1. It is typical of Canadian political tradition that from 1867 to 1982 this constitution was a law of the United Kingdom, the British North America Act. The Constitution Act of 1982 added important new clauses, but did not abolish the old act. 2. The tradition of state intervention in Canada has been called Red Toryism. Whitaker (1977, p. 43) calls it ‘private enterprise at public expense’.

138

The state at work, 1

3. Bird et al. (1979, p. 76) established that public sector employees enjoyed higher remuneration than those in the private sector. In 1971, the gross difference of 9.3 per cent was caused for one-third by the higher qualifications of public employees and for two-thirds by higher wages. We expect that the weight of higher qualifications would be greater today, in the wake of considerable contracting out for many manual and clerical jobs and the upgrading noted later in the federal government.

REFERENCES Aucoin, Peter (1995), The New Public Management: Canada in Comparative Perspective, Montreal: Institute for Research on Public Policy. Auditor General of Canada (1999), Report of the Auditor General of Canada to the House of Commons, April 1999, Ottawa: Public Works and Government Services Canada. Bird, Richard, Meyer W. Bucovetsky and David K. Foot (1979), The Growth of Public Employment in Canada, Toronto: Institute for Research on Public Policy and Butterworths. Commissioner of Official Languages (annual), Annual Reports, Ottawa: Public Works and Government Services. Department of Finance (2000), Tableaux de références financières, Ottawa: Department of Finance, September. Dwivedi, O.P. and James Iain Gow (1999), From Bureaucracy to Public Management. The Administrative Culture of the Government of Canada, Perterborough: Broadview Press and Institue of Public Administration of Canada. Gow, James Iain and François Simard (1999), ‘Where old and new management meet: Temporary staff in the Canadian federal administration’, International Review of Administration Sciences, 65, 71–86. Granatstein, Jack L. (1982), The Ottawa Men, Toronto: Oxford University Press. Heller, Peter S. and Alan A. Tait (1983), Government Employment and Pay: Some International Comparisons, Washington, DC: International Monetary Fund. Hodgetts, John E. and O.P. Dwivedi (1969), ‘Growth of government in Canada’, Canadian Public Administrative, 12, 224–38. Hodgetts, John E. and O.P. Dwivedi (1974), Provincial Governments as Employers, Montreal: McGill-Queen’s University Press. Noel, Alain (1993), ‘Politics in a high unemployment society’, in Alain-G. Gagnon (ed.), Québec: State and Society, 2nd edn, Scarborough, Ontario: Nelson Canada, pp. 422–49. OECD (Organization for Economic Cooperation and Development) (1999), National Accounts. Main Aggregates 1960–1997, Paris: OECD. OECD (Organization for Economic Cooperation and Development) (2000), Perspectives économiques de l’OCDE, Paris: OECD, number 68, December. Osborne, David and Ted Gaebler (1993), Reinventing Government, New York: Penguin. Public Service Commission of Canada (annual), Annual Report, Ottawa: Public Works and Government Services. Rogers, Betty (1996), Special Operating Agencies: Human Resources Management Issues, Ottawa: Canadian Centre for Management Development and Supply and Services Canada.

Public employment in Canada

139

Rose, Richard et al. (1985), Public Employment in Western Nations, Cambridge: Cambridge University Press. Savoie, Donald J. (1999), Governing From the Centre. The Concentration of Power in Canadian Politics, Toronto: University of Toronto Press. Schwartz, Herman M. (1994), ‘Public choice theory and public choices: bureaucrats and state reorganization in Australia, Denmark, New Zealand and Sweden in the 1980s’, Administration and Society, 26, 48–77. Siegel, David (1988), ‘The changing shape of public service employment’, Canadian Public Administration, 31, 159–93. Simard, Carolle (1994), ‘L’Etat réduit: le cas de la fonction publique fédérale’, in Robert Bernier and James I. Gow (eds), Un Etat réduit?, A Down-sized State?, Sainte-Foy: Presses de l’Université du Québec, pp. 55–73. Stanbury, William T. (1994), ‘Privatization by federal and provincial governments in Canada: an empirical study’, in Robert Bernier and James I. Gow (eds), Un Etat réduit? A Down-sized State?, Sainte-Foy: Presses de l’Université du Québec, pp. 165–219. Statistics Canada (monthly), Canadian Socioeconomic Information Management System (CANSIM) at www.statcan.ca/cgi-bin/Cansim/cansim.cgi. Statistics Canada (annual 1991–96), Public Sector Employment and Wages and Salaries, Ottawa: Minister of Industry, Series 720209-XPB. Statistics Canada (annual), Public Sector Statistics, Ottawa: Ministry of Industry, Series 68-213-XIB. Sutherland, Sharon L. (1987a), ‘Public employment in Canada’, Canadian Public Administration, 30, 110–22. Sutherland, Sharon L. (1987b), ‘Federal bureaucracy: the pinch test’, in Michael J. Prince (ed.), How Ottawa Spends 1987–8, Toronto: Methuen, pp. 268–92. Sutherland, Sharon L. and G. Bruce Doern (1986), Bureaucracy in Canada: Control and Reform, Toronto: University of Toronto Press. Tardi, Gregory (1992), The Legal Framework of Government: A Canadian Guide, Aurora, Ontario: Canada Law Book. Treasury Board of Canada (annual), Employment Statistics for the Federal Public Service, Ottawa: Treasury Board. Treasury Board of Canada (annual), Annual Report on Official Languages, Ottawa: Treasury Board. Verney, Douglas (1986), Three Civilizations, Two Cultures, One State: Canada’s Political Traditions, Durham, NC: Duke University Press. Whitaker, Reg (1977), ‘Images of the state in Canada’, in L. Panitch (ed.), The Canadian State: Political Economy and Political Power, Toronto: University of Toronto Press, pp. 28–68. Zussman, David (1986), ‘Walking the tightrope: the Mulroney government and the public service’, in Michael J. Prince (ed.), How Ottawa Spends 1986–87, Toronto: Methuen, pp. 250–82. Zussman, David and Jak Jabes (1989), The Vertical Solitude. Managing the Public Sector, Halifax: Institute for Research on Public Policy.

6.

Public employment in the United States: building the state from the bottom up B. Guy Peters

The conventional stereotype of the United States is of a society with a minimalist government, especially when compared with its counterparts in Western Europe, and its neighbour to the north Canada. A second and related stereotype has been that American government is concerned almost entirely with defence and the other ‘defining functions’ of government (Rose 1976), with only a residual welfare state. Both these stereotypes have a certain amount of validity. The public sector in the United States is ‘smaller’ by most definitions than are governments in most European states, and there is a large defence component at the federal level. There is no comprehensive health care system and the social safety net is substantially less developed than is that in European welfare states. On the other hand the stereotypes also make it far too easy for the casual observer to underestimate the importance of the public sector in the United States and to misunderstand the size, complexity and impact of American government. Americans complain about ‘big government’ and it is big in absolute terms – a total of more than 20 million people now work for governments at all three levels. Those millions of people provide a wide range of public service, with the largest areas of employment being education and health. There is a major ‘warfare state’, and it is increasing under the second Bush administration, but domestic employment and spending are overshadowing the defence side of government. It is also crucial to remember that, characteristic of federal regimes, all three levels of government are relatively autonomous and powerful in the United States.1 Even if the central (federal) government is not as significant an actor as are central governments in most other readily comparable regimes, the totality of the public sector is substantial and indeed continuing to grow. Further, somewhat paradoxically, the relatively small size of the federal government makes reducing government more difficult, even in an era in which the role of government has come to be questioned in almost all countries of the Western World – when we went to cut government there were relatively few easy targets. 140

Public employment in the United States

141

Associated with the rather insignificant opportunity to downsize the economic and social activities of government has been the failure of the end of the Cold War to have the significant impact on government employment and spending that it might have been thought to have. The ‘peace dividend’ simply did not materialize in the way that advocates of higher domestic spending or lower taxes have expected. The United States continues to be the major military actor in the world, and although forces have been used for peacekeeping and for deterrence (largely against perceived sources of terrorism) as well as for actual military action, there is still a major military establishment. As well as being interested in public employment in its own right, like the authors of the other chapters in this volume, I will be using public sector employment as an indicator of the size of the public sector. As has been pointed out, public employment has its problems as an indicator, but it also provides an excellent insight into government. As well as being an indicator of the overall volume of activity and the influence of government, public employment is an interesting and important variable in itself. Examining the size and nature of public employment reveals a great deal of the nature of the state, as well as its priorities. This utility as an indicator is especially evident in a political system that is as complex as that of the United States with its multiple levels of government, the different types of governments at the local level, and the rather convoluted interactions among those various types of governments.

THE PUBLIC SECTOR IN THE UNITED STATES Before beginning to document changes in public employment in the United States it may be useful to provide a very brief description of the actors involved, and some of the basic structural characteristics. The first point to be made is that there are three levels of government, each with some degree of autonomy in making decisions about revenue, expenditures, and employment. Local governments are ‘the creatures of state government’, but yet do retain some independence and some own-source revenues.2 This relative independence of local governments means that although we will be looking at the aggregate figures, those figures are the result of a large number of independent decisions. There are a very large number of governments, especially at the local level (see Danielson et al. 1977). In addition to the one federal government, the 50 state governments (plus the District of Columbia), in 1992 there were over 84 000 local governments (see Table 6.1). These local governments may themselves be nested; counties may contain municipalities and townships,

142

Table 6.1

The state at work, 1

Local government structure 1992

Municipalities Counties Townships Special purpose associations School districts

19 279 3 043 16 656 31 555 14 422

and there may be special district governments cutting across other local jurisdictions.3 The most important local function – elementary and secondary education – may be supplied by special-purpose school districts, by county and city governments, and may in some instances also be supplied directly by the state governments. Further, each state will have its own pattern of organization for local government, so that the pattern of government in the United States is indeed a ‘crazy quilt’. One relatively small state (in population terms), Nebraska, has almost 3000 local government units. The principal distinction that should be made, however, is between general-purpose governments and single-purpose governments such as school districts and special district government – our friends in Nebraska have 1130 special district governments and 681 school boards. Special district government is the fastest growing category, with over 3100 new districts created from 1992 to 1997. We should also note here that American government tends to be decentralized within each government, as well as across governments. Just as the final numbers of employees reflect a number of decisions taken in a number of governments so too do the decisions made within each government reflect a complex process of bargaining and conflict between branches, and even among different executive agencies.4 This internal conflict is often especially evident in state governments, where a number of executive officers may be elected independently and thus have some claims for making their own decisions about budget and employment, independent of the governor of the state.5 Few governments have a clear strategy for managing the size of the public sector, but the absence of a coordinated policy about employment is even more apparent in the United States.

TOTAL PUBLIC EMPLOYMENT The United States began the post-Second World War era with a very small public sector. This was true even though the country was at that time

143

Public employment in the United States

Table 6.2

1993 1992 1991 1990 1985 1980 1975 1970 1965 1960 1955 1950

Public employment overview, 1950–93 (000s) Total government employment *

% of population

% of labour force

20 683 20 661 20 604 20 497 18 879 18 280 17 102 16 093 13 243 11 357 10 613 8 635

8.0 8.1 8.2 8.2 7.9 8.0 7.9 7.9 6.8 6.3 6.5 5.7

16.0 16.1 16.2 16.2 16.1 16.8 17.9 19.0 17.3 15.9 15.8 13.6

Percentage change 1980–90 12.12 1970–80 13.59 1960–70 41.70 1950–60 31.52 1950–90 137.37 Note: * All figures include military personnel (without conscripts).

heavily involved in the Korean War, and had adopted a good portion of its own limited version of the welfare state just before the Second World War. In 1950, the year at which we will begin this analysis, only about one American in 18 was an employee of government at any level, and only one worker in seven worked for the public sector (Table 6.2). This rather meagre level of public employment was found even though there were 2.3 million men and women in the armed forces and already well over one million employees in education at the local level. It represents a major shift from 1945 when, at the end of the Second World War, the public sector had employed almost 16 million people, or 34 per cent of the labour force. This rapid shift away from public employment reflected the common pattern in American history of demobilizing quickly after a major conflict and then being able to expand the public sector quickly in times of crisis (see Peters 1985). As we will point out, however, the post-war period was to be exceptional in maintaining a large military establishment during a time without overt international conflict, and in creating the ‘warfare–welfare’ state that has come to characterize the American public sector.

144

The state at work, 1

Even in 1950, the federal government was a relatively small part of total public sector employment. Further, if the military component of federal employment were removed, federal employment would be only 1.2 per cent of total public employment, and 3.1 per cent of the labour force. This basic characteristic is perhaps the most crucial point about public employment in the United States. There is a substantial public sector, but most of it, especially for public employment, resides at the sub-national level. Even when the federal government finances an activity it may not actually hire the people delivering the service – programmes such as social welfare, housing and highway construction make that point very clearly. The United States has grown its government from the bottom up, and continues to add to local employment on an almost annual basis. Total public employment in the United States developed gradually during the last half of the 20th century, but it tended to grow rather steadily, albeit for different reasons at different times. What is perhaps most notable in these data is the steady growth of the public sector when considered in terms of the absolute number of employees. Government now employs well over 12 million more people than it did in 1950, and in almost every year since 1950 there have been more employees in the public sector than in the previous year. With the exception of adding almost 5 million new employees between 1960 and 1970 – a period of both warfare and the expansion of federal domestic programmes – there have been few rapid increases, but rather an incremental climb in American public employment. The share of the labour force employed by the public sector grew substantially for the first portion of the period but has remained almost constant since the early 1980s. This stability has occurred despite the rhetoric of cutting government from the Reagan administration, and to a lesser extent from the Bush and Clinton administrations, and a significant Republican presence in Congress and the state houses. Despite the scepticism about government among much of the public, large-scale public employment has been a continuing feature of American life and appears unlikely to be reduced further. This also points to the schizophrenia about government in this country, with the public opposing government in the abstract but being very favourable toward most specific programmes, especially those of local governments. Using employment figures tends to count each person who works for government equally whether he or she works ten hours a week or 40. One of the trends worth noting in American public employment is the increase in part-time employment at all levels but especially in local governments (Table 6.3). This trend reflects in part the increasing number of women

145

Public employment in the United States

Table 6.3

Part-time employment, 1950–93 (000s) Federala

Stateb

Total Part-time Total Part-time (%) (%) 1993 1991 1990 1989 1988 1987 1986 1985 1984 1983 1982 1981 1980 1977 1975 1973 1972 1970 1965 1962 1960 1955 1950

3 112 3 128 3 124 3 113 3 088 3 022 3 021 2 910 2 875

16.4 16.4 14.4 15.0 15.0 13.4 14.3 12.8 12.0

2 860 2 876

13.5 13.2

2 890

10.7

2 881 2 588

2.6 7.3

2 439 2 372 2 068

8.3 4.9 21.3

Localb Total

State and local non-education

Part-time Total Part-time (%) (%)

4 116

25.2

10 005

21.5

3 788

24.8

9 249

18.8

3 491

23.4

9 120

21.8

2 957

21.8

8 007

21.7

1 680

16.4

5 169

17.5

8 239 8 086 7 355 7 525

32.6 25.9 26.9 27.4

6 802 6 675 6 565 6 471 6 369

28.7 28.4 27.8 29.1 29.1

6 335 6 474 6 493 6 272 5 893

26.4 38.4 29.6 32.9 35.4

Notes: a. Source: Office of Personnel Management, Pay Structure. b. Source: US Bureau of the Census, Census of Governments, Vol 2, Compendium of Government Employment (Washington, DC: USGPO, quinquennial).

entering the labour market who seek (or perhaps are only offered) part-time employment. It further may represent managerial choices and attempts to reduce costs for themselves as employers. For whatever reason the increase in part-time employment has arisen, when public employment is calculated in terms of full-time equivalent employees rather than as a simple head count, the actual number of hours worked has been dropping, as has the overall involvement of the public sector in the labour market.

146

The state at work, 1

Level of Government The federal government has been a relatively small portion of the total public sector in the United States, and also continues to decline (Table 6.4). There has been some slight decline in the absolute numbers of employees since the mid-1980s and early 1990s, but the decline of federal employment as a percentage of total public employment is marked. At the beginning of the time period we are investigating, the federal government employment was almost one-third of the civilian total, and well over one-third of total employment. At the end of the period, federal employment was less than one-sixth of the total. Table 6.4

Public employment by level of government, 1950–98 (000s) Total Federal As % of government total employment*

1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1975 1970 1965 1960 1955 1950

19 854 19 540 19 533 19 521 19 420 18 939 18 815 18 563 18 391 16 690 16 191 14 974 13 028 10 589 8 817 7 480 6 346

Percentage change 1990–98 7.95 1980–90 13.59 1970–80 24.28 1960–70 47.76 1950–60 38.93 1950–90 189.80 Note: * Civilians only.

State

As % of total

Local

As % of total

2 765 2 807 2 842 2 895 2 952 3 012 3 131 3 112 3 128 3 021 2 876 2 890 2 881 2 588 2 439 2 372 2 068

13.9 14.4 14.6 14.8 15.2 15.9 16.6 16.8 17.0 18.1 17.8 19.3 22.1 24.4 27.7 31.7 32.6

4 758 4 733 4 729 4 719 4 694 4 659 4 586 4 521 4 503 3 984 3 753 3 271 2 755 2 028 1 538 1 199 1 045

24.0 24.2 24.2 24.2 24.2 24.6 24.4 24.4 24.5 23.9 23.2 21.8 21.1 19.2 17.4 16.0 16.4

12 271 12 000 11 962 11 906 11 775 11 268 11 098 10 930 10 760 9 685 9 562 8 813 7 392 5 973 4 840 3 909 3 233

61.8 61.4 61.2 61.0 60.6 59.5 59.0 58.9 58.5 58.0 59.1 58.9 56.7 56.4 54.9 52.3 50.9

11.60 8.76 0.17 18.12 17.94 51.26

– – – – – –

+5.66 +19.98 +36.22 +79.13 +47.18 +330.90

– – – – – –

+14.04 +12.53 +29.36 +52.73 +49.70 +232.82

– – – – – –

Public employment in the United States

147

This declining role of the federal government is in sharp contrast to the significant development of employment at the local level. At the beginning of the time period, local governments accounted for approximately half of all public employees in the United States. By the end of the period these governments had added 8 million employees and were approximately 60 per cent of total employment, and were continuing to increase. There is a big government in the United States, but it is at the local level, not in Washington as the political rhetoric generally argues. As already noted, this growth of employment represents a whole series of decisions to employ one teacher here and one policeman there, rather than a coordinated strategy. At the federal level the principal source of variance in employment has been international affairs. The expansion and contraction of the armed forces, as well as the civilian component of defence employment, has been responsible for a good deal of the change in employment. The military has expanded and contracted to meet the Korean War, the Vietnam War, and various less overt examples of international tension. At the end of each of the conflicts there was some reduction of forces, but not to the extent that had been true historically in the United States.6 As we consider the role of the armed forces in public employment, we should not overlook the Department of Veterans’ Affairs (formerly the Veterans’ Administration) that now employs almost 250 000 people (the second largest Cabinet department in employment terms) in a number of hospitals and other programmes providing benefits reserved for former members of the armed forces. The federal government has also expanded employment by virtue of taking on a number of new functions, and has also radically reduced other activities. For example, the Environmental Protection Agency was not formed until 1970 but now employs over 18 000 people, and NASA now employs almost 20 000. On the other hand, several economic regulatory agencies have been downsized or even (Civilian Aeronautics Board, Interstate Commerce Commission) eliminated entirely; the Federal Deposit Insurance Corporation has lost half its staff since 1990. There was a secular trend upward in non-defence federal employment for most of this period, with some slight reversal in that trend after the 1980s, with the most marked downward movements actually occurring in the Clinton administration. The second Bush administration actually has had an increase in federal employment, almost entirely in defence and in homeland security. The state level of government increased its level of employment more than any other level of government during this time period. The number of public employees increased over 300 per cent and the share of total employment more than doubled. State governments added almost 4 million new employees during these 45 years, and this level of government now constitutes one quarter of total public employment. This increase in

148

The state at work, 1

employment has come about because of several reasons. One is the various versions of the ‘new federalism’ (Conlan 1998) proffered by several presidents, with most changes designed to enhance the power of state governments and to assign additional functions to it. In addition, the federal government has given the states some major new programmes to administer on its behalf, the most important Medicaid or medical care for the indigent – like Germany the United States has ‘implementation federalism’ (Derlien and Peters 1998). Also, several of the major activities of state governments such as higher education, transportation, and other health programmes have become of increasing importance.7 State governments might have added even more employees had they not been at the forefront of administrative reform and increased their operating efficiency probably more than the other two levels of government. The real increase in the number of public employees has been in local governments – there are more than 8 million more public employees at the local level than there were in 1980. This change is in part by design – much of the ‘new federalism’ actually had the effect of increasing employment at the local level rather than at the state level. Further, two of the most popular government programmes in the United States – education and policing – are primarily local functions. These programmes have continued to increase rapidly, as have other local services. In the post-war period more and more Americans have moved to urban and suburban environments and there has been an increased need for services such as sanitation and water, activities that would be handled privately in rural settings. Although local government employment as a whole has been increasing markedly, and all types of local government have increased employment, the change has not been uniform across the different types of local systems. As might be expected given the general increase in primary and secondary education, school districts have added personnel at the most rapid rate, although only slightly more so than special district governments. Growth in the latter category represents a very large increase in the number of local governments of this type (but rather little growth in each), as generalpurpose governments have shed many functions to these more limited units. The number of school districts has actually declined, as these districts have consolidated to provide services more efficiently. Municipalities have added employees the least rapidly of all types of local governments, as cities have lost some of their centrality in American life. Employment in Geographical Areas Although employment in state and local governments is large and continues to grow, it is not evenly distributed across the 50 states. This is true

Public employment in the United States

149

of the level of employment per capita in the most recent data, as well as in the rate of increase in employment during the time period (Table 6.5). At the state level there is a more than 400 per cent difference between the lowest level of per capita employment (Pennsylvania) and the highest (Hawaii). There was an even greater level of variance in local government employment, with Hawaii having fewer than 20 per cent as many employees per capita as does Arizona. Given that different states choose to allocate functions to different levels of government in their own ways, it is more appropriate to consider the two levels of government together. When this is done, the highest level of employment (Arizona) is less than double the lowest (Pennsylvania). Although the level of variation is reduced, this is still a substantial difference given that all aggregations of state and local government tend to do approximately the same things. Again we must emphasize that the United States is highly decentralized, even for a federal political system, and there is no single pattern of service provision and public employment at the subnational level. This variation appears attributable to a number of different factors. One is political traditions and values (Kincaid 1982), with some northern industrial states such as Pennsylvania and Connecticut, which might have been expected to be major employers, actually having rather low levels of employment by depending more on the private sector than might have been otherwise expected. Similarly, the southern states tend to provide fewer public services than do most other states and hence also tend to employ relatively few people. The states of the upper Midwest, on the other hand, have a history of policy activism that has produced rather high levels of public employment. The western states also tend to have somewhat higher per capita levels of public employment than might be expected, given that they have been dominated politically by the Republican party. One possible explanation for this observation of markedly different levels of employment is that the states with large geographical areas also tend to have sparse populations, so that providing basic public services may be more difficult and more labour-intensive than in more densely populated states (see, for example, California). Still, when the geographical area of a state is correlated with its level of public employment, no significant correlation emerges in the United States.8 There is, however, a strong correlation between the population of the state and its level of public employment. Employment by Executive Departments At one level the structure of the federal executive branch is relatively simple. There are 14 Cabinet departments, some almost as old as the Republic itself and one formed within the past decade, albeit from a previously existing

150

The state at work, 1

Table 6.5 State and local government employment: population ratios by states, 1991 State and local employment per 1000 inhabitants

Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachussetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island

State

Local

Total

23.1 43.8 25.8 21.4 12.7 21.2 20.1 35.0 13.9 18.6 53.1 22.8 14.5 19.9 21.6 23.1 23.7 24.2 21.3 30.6 17.3 18.5 18.9 20.6 16.8 29.2 22.5 16.1 19.1 16.3 33.4 16.1 18.6 32.6 16.2 25.1 21.9 12.4 23.4

40.6 44.3 65.9 39.8 43.5 47.3 33.6 27.3 42.7 44.9 12.8 46.2 45.0 41.5 50.0 53.4 35.6 40.9 44.0 37.6 36.0 43.3 49.6 45.7 39.4 55.7 59.5 38.2 39.4 44.0 42.5 54.7 44.1 53.5 42.9 42.5 44.3 34.8 29.5

63.7 88.1 91.7 61.2 56.2 68.5 53.7 62.3 56.6 63.5 65.9 69.0 59.5 61.4 71.6 76.5 59.3 65.1 65.3 68.2 53.3 61.8 68.5 66.3 56.2 84.9 82.0 54.3 58.5 60.3 75.9 70.8 62.7 86.1 59.1 67.6 66.2 47.2 52.9

Public employment in the United States

Table 6.5

151

(continued) State and local employment per 1000 inhabitants

South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming District of Columbia

State

Local

Total

24.9 24.8 17.9 15.4 25.5 25.6 22.3 35.4 21.7 18.8 28.4 –

37.1 53.7 38.5 45.7 38.6 39.3 39.4 38.6 35.4 47.9 57.4 95.7

62.0 78.5 56.4 61.1 64.1 64.9 61.7 74.0 57.1 66.7 85.8 95.7

organization. Although the heads of all these organizations are members of the President’s Cabinet, these organizations are of very different sizes and have a range of powers. Most of these organizations are, in fact, composites performing a variety of functions using a variety of policy instruments; no one organizational model fits all executive departments well. The federal executive branch does not consist of the Cabinet departments alone. There are, in addition to the 14 departments, almost 50 independent executive agencies, again with a variety of different organizational formats and policy missions. The leaders of these executive organizations report to the President, but most are not in the Cabinet.9 They also vary even more in size than do the Cabinet departments. Further, there are a number of independent regulatory agencies which, although governed by autonomous commissions and not directly responsible to the president, do perform important federal functions. Then, beyond these are a range of other organizational formats and quasi-governmental organizations (Seidman 1999). In addition to the directly federal organizations there are also some quasi-governmental organizations such as the Post Office that has been made into a public corporation. Finally, the legislative branch also employs a substantial number of people as staff for individual congressmen and for the committees (16 000 in 1995), in the Library of Congress and the Congressional Research Service, and the General Accounting Office (roughly another 16 000). Table 6.6 provides some information about the current levels of employment in the Cabinet departments and in the major independent executive

152

23 497 124 663 960 116 56 327 77 357 129 139 48 563 23 400 155 662 16 964 72 361 21 557 7 364 –

State Treasury Defense Justice Interior Agriculture Commerce Labour Health and Human Services Housing and Urban Development Transportation Energy Education Veterans’ Affairs*

25 254 130 084 1 084 549 64 433 77 485 117 750 35 150 18 260 140 151 12 289 62 227 16 749 4 889 –

60.38

1 789 270

1 716 970 60.86

0.05

1 526

1 886 0.02

2 963 542

2 820 978

As % of Total Executive Branch

Executive departments

As % of Total Executive Branch

White House

Total Executive Branch

1985

1980

25 288 158 655 1 034 152 83 932 77 679 122 594 69 920 17 727 123 959 13 596 67 364 17 731 4 771 248 174

67.34

2 065 542

0.06

1 731

3 067 167

1990

Table 6.6 Federal civilian employment by department and agency, 1980–99

25 892 143 088 690 706 124 885 72 541 104 661 60 765 15 913 61 403 10 273 64 135 15 820 4 770 219 245

60.08

1 614 097

0.07

1 914

2 686 690

1999

– 2.39 9.81 33.21 48.79 6.61 14.63 13.09 10.23 50.47 24.44 4.79 10.78 0.00 11.66

7.62 27.27 7.71 49.01 0.42 5.07 43.98 24.24 20.37 19.85 6.91 17.75 35.21 (8.71)

21.86 –

20.30

10.57

8.22

–

12.40

8.73 –

Percentage change 1990–99

Percentage change 1980–90

153 6 152 247 156 750 021 33 646

6 152 228 285 660 014 39 504

* Departmental Status since 1990; before: Veterans’ Administration.

22 562 3 605 6 353 4 960 4 757 32 035 8 851

23 714 3 283 8 280 5 804 4 403 51 714 8 138

Note:

13 788 3 222 6 723 3 133 25 782

14 715 3 515 3 520 3 427 37 654

39.57

39.07

1 172 746

Environmental Protection Equal Employment Opportunity Comm. Federal Deposit Insurance Corporation Federal Emergency Management Agency General Services Administration National Aeronautics and Space Administration Nuclear Regulatory Commission Office of Personnel Management Small Business Administration Smithsonian Institution Tennessee Valley Authority U.S. Information Agency U.S. International Development Cooperation Agency Veterans’ Administration* U.S. Postal Service Other

1 102 122

As % of Total Executive Branch

Independent Agencies and Commissions

4 698 – 816 886 32 841

24 872 3 353 6 636 5 128 5 092 28 392 8 555

17 123 2 880 17 641 3 137 20 277

32.60

999 894

– – 866 342 64 008

18 647 2 862 3 689 4 575 5 271 13 321 6 318

18 590 2 945 7 492 5 521 14 199

38.48

1 033 780 – 8.57 2.26 57.53 76.00 29.98 25.03 14.64 44.41 10.78 +3.52 53.08 26.15 – – 6.05 94.90

–

4.88 2.13 19.86 11.65 15.65 45.10 5.12 23.63 – 23.77 16.87

3.39

16.36 18.07 401.16 8.46 46.15

9.28

154

The state at work, 1

agencies, and about the changes in employment in these organizations across time. The Department of Defense is by far the largest of the Cabinet departments, even after the downsizing during the 1990s. Again, if we consider the Department of Veterans’ Affairs as reflecting part of the defence task of the federal government, then more than half of federal employment is devoted to defence.10 At the other end of the size dimension the Department of Education employs only around 5000 people. This is, however, only one of several federal departments concerned with providing social services. There are over 130 000 employees in the Department of Health and Human Services, and another 13 000 in Housing and Urban Development. The economic activities of government are even bigger employers, including departments such as Agriculture and Commerce, and a number of independent and regulatory organizations as well. Again, however, we should point out that these employment figures may understate the involvement of the federal government in social and economic activities, given that much of the money goes directly to the public in social benefits, and that many federal programmes are actually administered through state and local governments. Just as the allocation of functions to state and local governments differs markedly in the 50 states, so too do the structures of their state governments. These differences make any real comparison of employment by organization fiendishly difficult. These structural differences among these state governments are, however, important for understanding what happens at this level of government, and we will make such observations as possible about the employment implications of these structural differences. One way to describe the structures of these governments is to point out the variation in the number of departments, ranging from six (Vermont) to over 30 in Illinois and Ohio. Further, state governments often rely on independent or partially independent commissions to administer programmes, especially economic and professional regulation. Types of Employment in Federal Departments The federal departments are rather obviously significant in policy and political terms, and there is a consequent desire for the president and Cabinet secretaries to exert their control over these organizations. Most public employees at the federal level are members of the career civil service, whether the white-collar General Schedule or the blue-collar Wage System pay schemes (see Table 6.7). There is, however, a large and growing tier of political appointees in each major federal organization. This is an important source of political patronage and also can be seen as a means of exercising control over organizations that might otherwise be too willing to

Public employment in the United States

Table 6.7

1999 1998 1997 1996 1995 1993 1992 1991 1990 1985 1980 1975 1970

155

US federal employment by class of employment 1970–99 (000s) Executive

White-collar

Blue-collar

Other

15.5 15.0 14.9 14.7 14.6 14.1 14.6 13.9 10.0 7.2 8.4 8.3 7.8

1310.6 1342.3 1358.7 1398.9 1446.5 1550.3 1593.8 1609.6 1574.3 1578.7 1455.5 1382.4 1259.0

213.5 222.0 232.6 245.7 263.9 310.1 328.9 336.8 343.9 422.6 440.9 501.5 574.9

110.7 106.9 105.1 107.8 112.9 135.3 114.5 104.5 106.9 75.3 21.4 54.2 105.1

follow their own paths. As noted, the number and percentage of political appointees in the federal government has been increasing. This increase in political appointment is to some extent general; as Paul Light (1995) has argued, there has been a general thickening of government with accountability coming through these appointments. There has, however, also been some tendency to increase political appointment more in the organizations that are more politically sensitive (Ingraham et al. 1995). In all these cases, however, presidents are attempting to ensure that their programmes are being adopted and implemented. Another means that presidents have for exercizing control over the remaining parts of government is the Executive Office of the President (EOP) (see Table 6.8). The EOP is the structure that supports the president, including providing advice on budgeting (The Office of Management and Budget), economic policy (Council of Economic Advisors), advice of foreign policy issues (National Security Council), and other political services for the president through the White House Office (see Hart 1997). Although most presidents have sought at the outset of their terms to decrease, or certainly stem the increase, of employment in the EOP, they generally have expanded the office by the time they leave office. Employment by Policy Area Although the analysis of federal organizations provides a good deal of understanding of the policy priorities of American government, we need

156

Table 6.8

The state at work, 1

Executive Office of the President, personnel 1965–95 Executive Office of the President

1995 1994 1992 1990 1985 1980 1975 1970 1965

Of which: White House Office

OMB

CEA

386 378 392 396 367 406 408 311 333

534 519 586 574 566 677 619 633 524

33 32 34 34 30 34 23 59 46

1623 1568 1866 1731 1526 1866 1918 4742 2849

Notes: OMB = Office of Management and Budget. CEA = Council of Economic Advisors. Source: U.S. Bureau of the Census, Statistical Abstract of the United States, annual.

to go further to gain a more complete understanding. First, all three levels of government are involved in almost all policy areas. The Constitution was once argued to prevent federal involvement in policy areas such as education, but there are now very few policy areas that are the province of a single level of government.11 Second, most Cabinet departments have activities that cut across several functional policy areas. We have already pointed out that the Department of Energy has important defence functions. Similarly, the Department of Veterans’ Affairs provides health, educational, housing and some social services to veterans. The list of these multipurpose organizations could be extended, but the basic point of the multifunctional nature of most organizations, and the multi-organizational nature of most functions should be clear. The most obvious conclusion to be drawn from the employment figures in Table 6.9 is that government in the United States is not dominated by defence so much as it is by education. In 1999, 43.8 per cent of total public employment, and closer to one-half of civilian public employment, is involved in providing education. If the uniformed military and civilian public employment in defence are considered together they comprise 9.7 per cent; without the uniformed services defence represents 3.3 per cent of total employment, and involves less employment than several other policy areas. In addition to the very large level of employment in education, there is also substantial employment in other social policy areas. Taken together,

157

Public employment in the United States

Table 6.9 Public employment by policy area, 1992 and 1999 (full and part-time) Policy area

1992

1999

N (000s)

%

N (000s)

%

Total Post, telecommunication Professional military

20 661 774 1 807

100.0 3.7 8.7

21 690 876 1 384

100.0 4.0 6.4

Subtotal Education and research Health and recreation Revenue administration Police, fire Social services Housing Defence, civilians Legal services Transport Other services

18 080 8 239 1 319 483 1 202 496 – 984 374 335

45.6 7.3 2.7 6.6 2.7 – 5.4 2.1 1.9 20.0

19 430 9 498 2 131 549 1 394 683 141 713 455 569 3 297

39.9 6.4 2.3 5.8 2.4 – 4.8 1.8 1.6 22.6

social policies (education, health, social services) account for over half of all public employees in the United States. The bulk of this employment is at the state and local government levels, although a good deal of this employment is actually involved in implementing federal policies. This level of employment is not, however, as significant as the social policy employment encountered in many other developed democracies. This difference is perhaps most obvious in the area of health care, given that most government involvement in the United States is through insurance, rather than direct provision of services.12 There have been a number of substantial changes in the relative shares of employment across the time period we have been examining. The most obvious of these changes is the reduction in the overall share of employment in the defence sector. In 1950 defence employment was over 30 per cent of total public employment and 20 per cent of civilian employment. Although education was already over 20 per cent of civilian employment, the characterization of the American public sector as dominated by defence did make some sense at that time. The shift out of defence employment occurred rather steadily over the period, albeit with some fits and starts at the time of the various international conflicts in which the United States was involved. The share of public employment in defence declined over time, and so too did the share of employment going into the ‘economic mobilization’

158

The state at work, 1

activities of government (Rose 1976). The declining relative importance of employment in economic activities was primarily a function of the more rapid increase in the social activities of government, especially education. Toward the end of the time period, there was some more purposive decline in the economic involvement of the public sector, as deregulation and downsizing of other activities became ideologically necessary. The United States did not have the large public enterprises to privatize that most European countries did, but there was still a retreat from economic involvement.13

EMPLOYMENT IN POLICY SECTORS OVER TIME We will now proceed to examine the major policy areas and the change in their levels of employment over time. As well as looking at the numbers of employees in total, we will be interested in changes in employment among the levels of government. As already noted, almost all policy areas have some employment at all three levels of government in the United States, so that looking at the shifts of employment among these levels will tell us a good deal about the changing nature of government. Education American government is more involved in education than in any other policy area. Although this dominant position of education is evident in 1999, it has not been true for the entire time period under review. That having been said, education has been and will continue to be a major source of employment for government. From 1960 to 1996 over 6 million employees were added to the educational sector (Table 6.10). This represents a growth of over 60 per cent in these almost 40 years. As well as being a political priority for governments, this expansion of employment also represents the expansion of the US population and the youth of that population compared to those of most other industrialized democracies. Education is, and has been, primarily a local government function. In 1960 five employees in six in education were at the local level, with the federal government having only about one half of one per cent of educational employment. In relative terms the principal expansion of education employment has been at the state level, going from 15 per cent to one quarter of employment in this policy area. This shift has something to do with education itself, and the massive expansion of higher education in the United States during this period, much of it publicly provided at the state level. This shift in the location of education employment also represents some realities of public finance. Most local revenue comes from property

Public employment in the United States

Table 6.10

1999 1993 1992 1991 1990 1985 1980 1975 1970 1965 1960

159

Public employment in education and research, 1960–99 Total (000s)

Federal (%)

State (%)

Local (%)

9 498 8 372 8 239 8 087 7 921 7 119 6 857 6 294 5 316 3 974 2 930

0.16 0.17 0.17 0.16 0.21 0.38 0.35 0.36 0.35 0.41

25.22 24.88 24.72 24.29 24.78 23.17 22.24 22.23 18.60 16.18

74.62 74.95 75.11 75.55 75.01 76.45 77.41 77.41 81.05 83.41

7.69 50.00 36.84 58.33 8.33

5.65 21.08 34.43 149.37 305.91

2.85 14.15 27.39 68.37 144.84

Percentage change 1991–99 17.45 1991–93 3.52 1980–90 15.52 1970–80 28.99 1960–70 81.43 1960–90 170.34

tax which has not been as buoyant as revenues at either the federal or state levels. Thus, it was easier for the public sector to meet increasing educational demands at the state level than at the local level.14 The federal government has taken a greater interest in education over this period, but since it provides very little education directly,15 it has not kept pace with the expansion of educational employment at the other levels. Health A second important area of social policy that employs a significant number of people is health care. American government provides health care for only a limited number of its citizens, but even then there are almost 2 million people employed in health care. This number has increased significantly since 1960, with the most rapid growth occurring during the 1960s. This was the period when the federal government adopted Medicare and Medicaid, providing health insurance to the elderly and the medically indigent. Although most of the actual provision was provided by private physicians, there was a substantial increase in the number of employees responsible for supervising these programmes.

160

The state at work, 1

Despite the increased federal role in funding health care, the major increase in health employment has been at the local government level. In fact, both state and federal health care employment declined as a percentage of the total in the area. Several factors appear to explain this shift. First, a good deal of state involvement in medical care has been in mental hospitals, and the movement toward deinstitutionalization reduced the size of these facilities significantly. Also, local governments (especially in urban areas) have taken on a somewhat larger role in providing medical care, especially for emergency medicine, and as urbanization and suburbanization increase, volunteer rescue squads and the like are being replaced by professional, public organizations providing that service. Defence and Foreign Affairs As already noted, employment in defence and foreign affairs has been more variable than in most other policy areas in the United States. After downsizing the military significantly after the end of the Second World War (going from 12.1 million in uniform in 1945 to 1.45 million in 1948), the opening of the Korean conflict produced another rapid increase, with the armed forces increasing to 3.7 million in 1952). The continuing Cold War limited downsizing of the military after Korea, although there certainly was some. The military then expanded again to meet the demands of the Vietnam War, and then again to address some of President Reagan’s foreign policy goals.16 The ‘peace dividend’ from the end of the Cold War has not materialized to nearly the extent expected, or at least hoped for. The size of the armed forces has been reduced, and especially civilian defence employment has been reduced, but new tasks have been found to occupy the armed forces. These tasks – peace-keeping around the world and some limited role in the ‘war on drugs’ – are not the roles for which most of the armed forces have been trained, and this change has required a commensurate change in the force structure and in the mix between civilian and uniformed employees in this policy area. As we discuss employment in defence, we should also remember the large number of jobs created in the private sector from military purchases.17 This secondary employment in defence has increased as a proportion of the total as the armed forces have continued to substitute technology for manpower in an attempt to increase effectiveness while minimizing the exposure of troops. The second Bush administration has expanded defence employment somewhat following the attacks of September 11, and following the promulgation of a more aggressive foreign policy involving preemption. At the same time, however, the leadership in the Pentagon (at least the civilian leadership) is emphasizing technology as a replacement for personnel.

Public employment in the United States

161

Social Services The United States is certainly not the heartland of the welfare state, but there is still a substantial level of employment in providing these services. By the mid-1990s the level of employment in social services had more than quadrupled since 1960, and (like health employment) grew especially rapidly during the 1960s. During this decade President Johnson pushed his ‘Great society’ programmes through Congress, rapidly increasing the role of government in social affairs. In contrast, there was very little growth in employment in the 1980s and early 1990s, a period in which the involvement of government in social policy came under increasing attack. There is substantial employment in social policy at all three levels of government, although most is found at state and local levels. The participation of the federal government in social services is primarily as a funder for the other levels, and also as a provider of pensions and other social insurance benefits. While extremely important, these programmes involve relatively low levels of employment relative to the amount of money spent. Thus, if we were looking at spending rather than employment, the federal government would be by far the major actor, but in employment is only a minor player. Even the Great Society programmes mentioned above had their major employment effects at the sub-national level; as with so much policy in the United States these policies were federal but were implemented almost entirely through the state and local levels. As a percentage of total public employment in the social services, local governments still employ over half of the total number of workers in this area, but have a declining share of employment in this field. The growth area has been at the state level, although the division between the two levels has been largely unchanged since the 1970s. The shift in the 1960s and 1970s came primarily in response to federal programmes that used the states as the conduits for funding, and to some extent for direct service provision. The federal role has been declining despite its major (and increased) role in funding. Employment in the Social Security Administration actually has declined over time, in part because of the ability to use computers for record-keeping and cheque-writing. Police and Fire Protection In addition to defence the other major protective services of government are police and fire protection. In the United States these have grown significantly over the period under consideration. For much of this time period, crime was a major concern of citizens, and government felt obliged to respond by hiring more policemen and women. Although discrepancies

162

The state at work, 1

in the data make comparisons across time difficult, it appears that even in this policy area, when popular concerns appeared to have come later, there was a great deal of growth in the 1960s. This is perhaps because of the greater availability of public funding for almost all programmes during that time, and perhaps in response to the urban unrest and anti-war unrest of the time. Police and fire protection is one of the few policy areas in which there has been an increasing federal role over the time period in question.18 State and local governments are by far the dominant employers in this area, with more than five out of six employees, but the federal role has been increasing and now almost equals the level of employment of state governments. This is in large part because of the expansion of federal activities in drug intervention, as well as in terrorism and other major crimes. In this policy area, as in others, there is also an increasing use of state and local governments to implement federal policy. For example, although not showing up yet in these data the federal government has provided funds for up to 100 000 new policemen for local governments. We should also note that there is another significant employment area – prisons and justice – that is related to the policing function. These data have been disaggregated from general public employment in the United States only in recent years, so there is little way to judge the longer trends. Still it appears that there has also been a shift toward greater federal involvement, for many of the same reasons that there has been a shift in policing. Again, the fight against drugs and other federal crimes has made the federal government more of a policing institution than it has been in the past. Transportation The one economic area in which the United States does have a very large level of employment is transportation. Again, this may not be as high as might be found in many European countries, even after the privatization of rail and air transportation in those countries. A good deal of public transportation activity in the United States has been, and continues to be, provided by private sector personnel. For example, a number of cities and counties have their (often limited) public transportation provided by private firms operating under contracts. Further, almost all highway construction is done by private firms, although the design and engineering may be public. The data we have is concentrated on highway employment, a category that includes local streets. In these data the federal government – again despite being a major funder of highway construction – is a very minor employer. The state and local governments, and especially the localities, are the major

Public employment in the United States

163

employers. The distribution of employment between these two levels has remained roughly the same across the time period. This policy area has not experienced the major policy or technological changes found in others and has continued to provide these services much as it has in the past. The pattern found for total employment tended to be reproduced in the individual policy areas discussed here. For a number of these areas there was a relative shift toward local provision and away from federal government employment. The federal government has not lost employment in absolute numbers in most of these policy areas during the time period in question, but local and to some extent state employment simply have increased more rapidly. These increases at the subnational level are often funded by central government programmes, but ‘implementation federalism’ means that the individuals are actually paid by the lower levels. The other obvious point here is that none of these policy areas, with the exception of defence, is employing any fewer people. The public sector is continuing to expand in the United States, despite the rhetoric about cutting government. The public may want to reduce the size of government in the abstract – and with that the amount of taxation – but when it comes down to services, the average citizen wants at least as many, and generally wants more. This is the conundrum that public leaders face as they attempt to manage revenues and expenditures, and public employment.

EMPLOYMENT BY OCCUPATIONAL GROUP As well as examining the functions performed by public employees we should also look at the structure of employment in the public sector according to the occupational categories and grades used to manage and reward these employees. Shifts in these categories may also say a great deal about the political and economic power of the public service. These shifts may also say something about the impacts of technology on the public service and the changing nature of work more generally. At the federal level there are five broad categories of public employees. First, the executive grades are either senior public servants (the SES) or political appointees responsible for major policy or management areas in government, or perhaps providing policy advice. Second, the bulk of the public service is contained in the General Schedule (GS), covering primarily white-collar positions. Third, the Wage System (WS) contains bluecollar employees. The Postal Pay System contains all the members of the US Postal Service. Finally, there is an ‘other’ category for employees not covered by the other four, for example the Central Intelligence Agency, the Tennessee Valley Authority, and a number of small employment groups.

164

The state at work, 1

Table 6.11 Federal general schedule employment by grade and sex (percentage female), 1970–98 Grade

1970

1980

1985

1990

1994

1996

1998

GS 1–6 GS 7–10 GS 11–12 GS 13–15 GS 16–18*

72.1 33.4 9.5 3.0 1.4

74.1 43.6 19.3 8.2 4.4

74.5 50.3 27.5 12.1 6.1

74.7 51.1 29.6 13.7 8.7

74.6 54.9 32.1 15.7 13.2

74.0 55.2 33.6 17.7 15.3

73.2 56.0 37.4 20.1 17.7

Total

40.3

45.1

47.7

48.6

Note: * This category is used decreasingly as the Senior Executive Service is more fully institutionalized.

The general pattern of movement among these five categories has been a marked reduction in the number of WS employees, an increase in the number of executive and GS employees, and limited movement in the ‘other’ category. Employment in the postal service decreased as the movement to a semi-public corporation generated pressures for greater efficiency. It later began to increase again as the volume of mail has continued to increase and the postal service has expanded the variety of services offered to the public. Another way to classify employees is by the so-called PATCO system, meaning broad categories of the type of work performed. These data are available for both federal (GS) and subnational employees, although the data use somewhat different categories. At the federal level the most obvious point is the marked decline in the number of clerical employees since the 1980s. From the mid-1980s until 1997, the number of clerical employees has been cut almost in half at the federal level. The same general observation can be made for employment at the subnational level of government, although the rate of decline of ‘administrative support’ personnel as a proportion of total employment is not as rapid as that found in the federal government.19 The public sector historically has been a major employer of clerks, typists and similar workers, given that a major function of government is collecting, processing and storing words and numbers. This has meant that rapid advances in computers and other information processing equipment has made it possible for governments to reduce their employment levels substantially. Executives now send their own e-mails rather than depending upon someone else to type letters, and files are stored automatically on disk rather than in metal cabinets.

165

Public employment in the United States

EMPLOYMENT BY SOCIAL GROUP The final general question we will address is how public employment is distributed among various groups in society, as well as how the employment of government itself is structured. One of the old arguments in public administration is the importance of ‘representative bureaucracy’ (Kingsley 1944; Seldon 1997), and the extent to which political resources should be used to ensure that the composition of public employment mirrors as closely as possible the society. The assumption is that government will be more democratic, and potentially even more effective, if it more closely mirrors society. In the United States, the representative bureaucracy question arises with respect both to race and to gender. The public sector is often argued to have the responsibility of being a ‘model employer’ and to demonstrate to the private sector that it is possible to create greater social equality in the workplace. Historically the public sector has fulfilled that role in many ways. Federal employment was a way for minorities to obtain stable, reasonably well-paying jobs when the private sector continued with overt discrimination. The Post Office, the military, and then somewhat later the general civil services at federal and state levels have been avenues for minorities and for women. Table 6.12 provides a summary of the development of federal employment broken down by race and by level within the civil service. It is clear that the federal government employs a substantial number of minority group members, more on average than their representation in the population as a whole. On the other hand, these members of minority groups are generally not in the highest levels of the civil service. Their numbers in the Senior Executive Service and in Executive level positions have been increasing, but there is still a marked difference between the level of minority members in the total service (and especially in the blue-collar Wage Service) and in the Table 6.12 Federal government employment by race and status, 1985 and 1993 White

Executive White-collar Blue-collar Other* *

Black

Hispanic

1985

1993

1985

1993

1985

1993

7 179 1 196 578 283 107 66 358

13 182 1 190 705 215 122 96 711

343 248 732 84 492 4 868

716 277 264 69 976 12 823

92 69 241 30 177 2 911

274 87 200 24 515 5 946

‘Other’ refers to pay systems not in the general schedule, largely white-collar jobs.

166

The state at work, 1

Table 6.13 State and local government employment by sex and race (000s)

1997 1995

Male

Female

White

Black

Hispanic

Other

2898 2960

2307 2355

3676 3781

973 993

392 379

164 162

executive positions. Among minority groups there is little difference in the degree of relative success in achieving the top level positions, although Asian Americans are beginning to achieve those positions at approximately the same or higher rates as they exist in the population as a whole. At the state level the story is much the same (Table 6.13). The states have, in general, been somewhat slower to extend employment opportunities to minority group members, but they appear to have moved somewhat more rapidly than the private sector.20 As with federal employment, minority group members tend to be concentrated in lower paying positions. In 1995, blacks earned on average 82 per cent of what whites earn; Hispanics earned an average of 91 per cent of white salaries. The salary gap between white and black employees of state and local governments actually has increased slightly since the early 1970s, while that for Hispanic employees has narrowed. The story for employment by gender is much the same as it is for minority group members. The federal civil service and the public sector as a whole is heavily female – now just over half of federal general schedule (whitecollar) employment is female. That having been said, however, federal employment as a whole is not as heavily female as the total civilian labour force (OPM 1998). Further, there is still some of the same concentration of women in the lower echelons of organizations that is found for minority group employees. Women do now comprise a larger proportion of senior positions in the federal government than do members of minority groups, but given that they are half the population this might be expected. Still, the rate at which women are increasing their involvement in the higher levels of the federal civil service is somewhat greater than that of minority groups. For example, in 1999, 21.9 per cent of the Senior Executive Service was female while 12.4 per cent of that service were members of minority groups. In the state and local governments there is a pay gap similar to that found for minority group employees. In 1995, women in state and local governments on average earned 81 per cent of what men did. This has improved from earning only 73 per cent of male salaries on average in the early 1970s, but still represents a significant pay difference. Further, women tend to earn less than men within various categories of employment in state and local governments. For example, in 1999, women in top administrative positions

Public employment in the United States

167

earned on average 88 per cent of what men earned. In service and maintenance positions women earned less than 80 per cent of the average male salary. The above data indicate that the differences in wages are a function of something more than simply the concentration of women in certain, lowerpaid positions. In fact, the one broad category in which there are more women than men employed in state and local level is the category of professionals. These are primarily school teachers and nurses who are often not as highly paid as other professions, but they are still rather well paid when compared to the labour market in general. There may, therefore, be some discrimination in pay categories based on gender.

SUMMARY The United States does have a substantial and growing public sector. Despite the ideological and political rhetoric about the need to reduce the size of government, the public sector has added more employees almost every year since 1950. Most of this growth has been found at the state and local government levels, providing popular public services such as education and policing, but these are public employees nonetheless. Much of the political debate has been about the power of government in Washington, but the American public sector has been locally grown. The nature of that public sector has also been changing. Defence employment has been declining relative to total public employment, while education and the social services have become more important. Public employment is also becoming even more concentrated at the state and local levels. Public employment also is becoming increasingly devoted to providing social services (especially education) to citizens rather than being primarily concerned with the protection of the country against foreign enemies. Thus, some ‘path dependent’ aspects of employment continue along unabated, while others have experienced a substantial shift in priority.

NOTES 1. 2. 3.

In all countries, whether federal or unitary, there is increasing decentralization of government employment and service delivery. The principal difference may be in the autonomy of subnational governments to make their own employment choices. This is the so-called ‘Dillon’s Rule’, named after an Iowa court case in which the phrase was first used. The author of this chapter regularly pays taxes or fees to county and city governments, to the school board, and to one special district government. When he rides a bus or light rail car he pays fees to another special district government.

168 4. 5.

6. 7.

8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

The state at work, 1 This term is used here in a more general manner than in the UK (Hogwood in this volume). It refers simply to the various components of Cabinet departments (state and federal or even local). For example, in both North Carolina and North Dakota there are seven executive officials elected in addition to the traditional three of Governor, Lieutenant Governor and Attorney General. These officials are responsible for important public sector activities such as agriculture and public education. The experiences with George III left an enduring suspicion of standing armies, something that was overcome only in the Cold War years. For example, at the end of the First World War the US Army dropped from over 2 million to 30 000 troops. One of the persistent misunderstandings about the public sector in the United States is that there is little or no public involvement in health care. In reality about half of total health spending in the US is now public and approximately one-quarter of the population depend upon the public sector for health insurance. There was, however, a very strong correlation between area and level of employment in Germany (see Derlien and Peters 1998). Most recent presidents have included the Director of the Environmental Protection Agency in the Cabinet. A significant share of employment in the Department of Energy is concerned with nuclear weapons technology, as became apparent during recent spying scandals in that Department. Foreign Affairs is perhaps the only good example, but even there state governments are beginning to have their own foreign economic policies. State governments are involved in defence policy through having state militias in the form of the national guard (although not counted as public employees). State and local governments do provide some health services directly, as does the federal government through the Department of Veterans’ Affairs, the Indian Health Service, and some programmes of the Department of Defense. There were some such industries at the state and local level, some of which have been privatized, and there are huge tracts of public land owned by the federal government, some of which have been sold. Much of this problem has been addressed by state funding of local education, but there also has been the shift into state-level employment. The federal government does run the four military academies and Gallaudet College for the Deaf, as well as being heavily involved in Howard University. Whether Reagan himself was clever enough to conceive of the idea of forcing the Soviet Union to spend itself into oblivion is very doubtful. Certainly he presided over growth in the armed forces and in civilian defence employment. All policy areas generate some secondary employment – education involves buying school buses, paper and so on – but it is especially pronounced in defence. This rather significant role for the federal government in policing (almost 8 per cent of police personnel) is in contrast to some other federal systems that have sought to prevent the central government from having any major role in policing. These findings are confounded somewhat by being based on different samples of local governments, along with all state governments, in different years. There have been marked differences among the states with some southern states moving very slowly to hire minorities and some northern and western states (notably Hawaii) having very large minority involvement in the public service.

REFERENCES Conlan, Timothy J. (1998), From New Federalism to Devolution: Twenty-Five Years of Intergovernmental Reform, Washington, DC: Brookings Institution.

Public employment in the United States

169

Danielson, Michael N., Alan M. Hershey and John M. Bayne (1977), One Nation, So Many Governments, Lexington, MA: D.C. Heath. Derlien, H.-U. and B.G. Peters (1998), Who Works for Government and What Do They Do? Common Trends and Structural Differences of Public Employment in the United States and Germany, Verwaltungswissenschaftliche Beiträge, 32, Bamberg: University of Bamberg. Hart, John (1997), The Presidential Branch, 2nd edn, Chatham, NJ: Chatham House. Ingraham, Patricia W., James R. Thompson and Elliot F. Eisenberg (1995), ‘Political management strategies and political/career relationships’, Public Administration Review, 55, 263–72. Kincaid, John (1982), Political Culture, Public Policy and the American States, Philadelphia: Institute for the Study of Human Issues. Kingsley, J. Donald (1944), Representative Bureaucracy, Yellow Springs, OH: Antioch University Press. Light, Paul C. (1995), Thickening Government: Federal Hierarchy and the Diffusion of Accountability, Washington, DC: The Brookings Institution. OPM (Office of Personnel Management) (1998), Fact Book, Washington, DC: OPM. Peters, B. Guy (1985), ‘The United States’, in Richard Rose et al. (eds), Public Employment in Western Democracies, Cambridge: Cambridge University Press. Rose, Richard (1976), ‘On the priorities of government’, European Journal of Political Research, 4, 247–89. Seidman, Harold (1999), Politics, Power and Position, 5th edn, New York: Oxford University Press. Seldon, Sally C. (1997), The Promise of Representative Bureaucracy, Armonk, NY: M.E. Sharpe.

7.

The German public service: between tradition and transformation Hans-Ulrich Derlien

INTRODUCTION When Max Weber stated ‘political domination, in everyday life, means public administration’, he referred to law and order as core state functions that were and are executed by civil servants. Since the beginning of the 20th century however, the scope of public tasks has undergone redifinition, both expanding with the incorporation of functions indicative of the continental European welfare state and contracting following the privatization of major utilities and services. Concomitant with the expansion of the welfare state, the public service is no longer dominated by civil servants but by employees and labourers. As well as the changes in the social composition of the public service, incremental reforms of the legal framework and the vertical distribution of the public service and its distribution among policy areas have occured. Nevertheless, the image of the public service in Germany is still strongly shaped by the classical civil service ideal and its traditional principles enshrined in Article 33 Section 5 of the 1949 federal constitution: the orthodoxy of tenured full-time employment and recruitment to one of the four career groups according to level of education. The tradition includes the vertical division of labour between federal and (since 1990) 16 state (Land) governments and 14 000 local governments that are regarded as part of Land administration. The until-recently unchallenged position of the German civil service should be put in historical perspective in two ways: the German professional civil service is a product of the 18th century absolutist state, notwithstanding several adjustments during its roughly 200 years of existence (Derlien 1987). Second, the reunification of Germany in 1990 posed secular problems for the public service owing to the divergent developments in the East and West over the previous 45 years. Germany for the first time after a regime change in the 20th century, faced the problem of the party-politicized incompetence of public functionaries (Derlien 1991). Furthermore, the unitary eastern public service was transformed to match 170

The German public service

171

the western model differentiating between civil servants and public servants on contract (employees, labourers) as well as between career groups. Whereas most countries such as the UK, New Zealand, Australia and Canada have, since the 1980s, experienced comprehensive administrative and civil service reforms, administrative modernization only became a topic in Germany in the 1990s. Until then government and opposition politicians in Bonn repeatedly acknowledged the professional performance of the ministerial bureaucracy in Bonn in drafting the complex unification treaty with East Germany (Derlien 1993) and appreciated the western public service’s role in helping to transform the eastern part of the country. Since 1991 though, a change of guiding ideas also took place in Germany. Today the public service and the costs of the welfare state they are working for are questioned. The service no longer retains the self-legitimization arising from its alleged superior effectiveness and efficiency. In particular, at the local government level, New Public Management albeit in a reduced version, has gained ground. As in the Anglo-American countries, politicians took over administration policy, replacing what I have called in relation to the previous 30 years of incremental reforms, ‘the intelligence of bureaucracy’ (Derlien 1996a), and began driving for public sector reforms under the banners of ‘lean state’ (1995–97) and ‘administrative modernization’ (federal government programme 1999) (Derlien 2000). Apart from changes to the administrative structure and welfare policies, the public service became the target of reform or was indirectly affected by organizational reforms such as privatizations. Reform measures were motivated basically by fiscal pressure and to a lesser extent induced by a comprehensive philosophy. Paramount are the costs of unification and the mounting burden of pensions that will be felt heavily when large cohorts of civil servants who were recruited into the expanding public service of the 1970s start retiring after the year 2000. Thus, public service policy at the turn of the millennium includes reducing the size of the public service, abandoning public tasks, having civil servants work longer than the previous retirement age, confining premature retirement, making civil servants contribute to their pensions and, last but not least, substituting public employees on private law contracts for civil servants. A new element, in fact, is the determination of some Länder to reshuffle the public service from the allegedly more expensive civil service to public employee status, thus harmonizing the different status groups of civil servants, public employees and labourers. The first section below documents the development of the size of the public service relative to population and labour force. Of particular interest is the increase in part-time employment among the 4.9 million public servants in 2000 (for a full statistical report consult Derlien et al. 1998). This is followed by an analysis of the allocation of the public service to the

172

The state at work, 1

various policy areas; the major growth areas are identified in retrospect. The third section examines the distribution of public employment among the three tiers of government, revealing a ‘lean’ federal government and the importance of the Länder in implementing federal policy. The fourth section deals with various aspects of social differentiation of the public service: the status group system of civil servants vs. employees and labourers, the size of the four career groups and gender representativeness – aspects that vary by public task area and thus by level of government. Furthermore, characteristics of the administrative elite are elaborated. Finally, personnel management is outlined, in particular its decentralized nature within a common legal framework. From the analysis, the German public service, despite privatization and downsizing, appears to be heavily concentrated in policy domains indicative of continental welfare. Despite its traditional formal differentiation, the public service is undergoing a change toward uniformity of status groups, temporarization of employment and feminization, in particular since national unification. All of these dimensions cluster, for women prefer to work in education, health and the social services and mostly do so as employees on part-time jobs at the level of subgovernments. The traditional civil service, although in 2000 only one-third of the overall public service, clusters in general administration, including the ministries, and in the law and order branches where predominantly male civil servants are serving in full-time positions.

PUBLIC SERVICE GROWTH: REVISING WAGNER’S LAW? In 2000 the public service comprised roughly 4.9 million people (Table 7.1). Since 1950, total employment has increased by 115 per cent, with the highest growth rates during the years 1950–60 (38 per cent). Due to national unification in 1990, another 38 per cent increase occurred within one year (amounting to 6.7 million in 1991), but by 2000 absolute numbers were back to where they were before unification and the population ratio even approached that of 1960 (60 public servants per 1000 inhabitants). This downsizing is less an indicator of a general thinning of service intensity than a result of massive privatization, both of central government operated telecom and railway services as well as the dismantling of the eastern German comprehensive state economy and services related to the socialist industrial conglomerates. As a percentage of the labour force, public employment moved between 16 and 17 per cent between 1980 and 1993, with a short-term increase after

173

The German public service

Table 7.1

Total public employment absolute and relative, 1950–2000 Public employment (000s)

Female (%)

Part-time (%)

Percentage of labour forcea

Public servants per 1000 inhabitantsb

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991

4909 4969 5069 5164 5276 5371 6094 6503 6657 6738

50.8 50.5 50.2 50.0 50.0 49.8 48.8 47.6 47.5 46.8

24.7 23.4 22.0 20.9 20.3 19.9 19.3 17.4 16.2 15.8

12.0 12.3 12.6 12.9 13.2 13.4 15.1 16.9 16.6 16.8

59.7 60.5 61.8 63.0 64.4 65.8 74.8 80.1 82.6 84.2

1990 1985 1980 1970 1960 1950

4940 4845 4658 3876 3152 2282

41.8

17.7 15.9 13.3 9.8 6.2 2.9

15.7 16.7 16.9 14.3 12.0 10.4

77.8 79.5 75.7 63.9 56.9 45.7

Percentage change 1991–00 27.1 1990–91 36.4 1980–90 6.1 1970–80 20.2 1960–70 23.0 1950–60 38.1 1950–00 115.1 Notes: a. Labour force is defined as those in labour plus those seeking labour, the problem being to determine who is seeking labour (labour office registration and/or census data). Source for 1998: http://www.statistik-bund.de/basis/d/erwerb/erwerbtab1.htm. b. Population is defined as German plus foreign inhabitants; Source: http://www.statistikbund.de/basis/d/bevoe/bevoetab4.htm for 1998 and Statistisches Jahrbuch für die Bundesrepublik Deutschland 1997, CD-ROM.

unification, and then gradually declined to 12 per cent in 2000. Thus, the German public sector, despite taking 45 per cent of GNP, is smaller in terms of its labour market share than in many industrialized countries. Although Germany is regarded as nurturing a massive welfare state, in fact, it is no larger than the US public sector (Derlien and Peters 1998). The shrinking of the public service in the 1990s, however, both in absolute and

174

The state at work, 1

in relative terms, by no means proves that Wagner’s law of ever-expanding state functions (Wagner 1911) was being reversed; as will be shown below, the figures indicate only that relatively fewer public servants are spending ever more money, in particular for transfer payments. A particularity is the category of semi-state employment (mittelbare Staatsverwaltung) in self-governing public pension, accident and health care funds (federal and Land public law corporations) and in the federal labour market administration. In these areas federal and state governments are confined to legal supervision.1 About 5 per cent of the public service used to be employed in these public law corporations; since 1992 their share has grown (488 000 or 9.9 per cent in 2000) due to the transfer of this type of administration to the eastern part of the country. Total public service growth conceals a substantial portion (24.7 per cent) of part-time employment in the public sector, emerging noticeably only in the 1960s. It was always most widespread at the local and state levels, for education, social and health care are the preferred task areas for part-time employment. National unification had a considerable impact on the further increase of part-time work in the 1990s, for these terms of employment were often used as a more humane way of downsizing the eastern full-time public service (see Table 7.1). In general, with contracting total public employment and a slowly but constantly growing number of part-time personnel, the German public service shows a trend towards temporarization of employment, moving away from full-time orthodoxy (Derlien 1999). Another unification-induced change is the growing share of women in public service; since 1996, parity has been achieved. This development and its association with growing part-time employment will be examined more closely below.

THE DIFFERENT WEIGHT OF POLICY AREAS A rough indicator of the relative importance of public tasks is their share of the 4.9 million (2000) employed in the public service (Table 7.2). As in most OECD countries, school and university education including research (31.7 per cent) is by far the most important task-area, followed by the social services (13.9 per cent, including the semi-state compulsory pension and accident insurance systems). A German particularity is the general administration category (13.1 per cent) that includes personnel in the core administrative units such as the ministries, regional authorities, revenue offices and central local government services. Police and public safety (8.9 per cent) and the health sector (8.5 per cent) occupy the fourth and fifth rank. While general administration employment (644 363), popularly perceived as the bureaucracy proper, increased only disproportionately and less than

175

The German public service

Table 7.2 Policy areas

Task areas by personnel and expenditures in Germany, 2000 Personnel

Personnel expenditures

%

€m

%

€m

%

Percentage of gross expenditures

13.1 0.1 6.5

34 063 22 641 23 123

3.5 2.4 2.4

23 730 445 11 840

13.0 0.2 6.5

69.7 2.0 51.2

8.9 4.0

19 860 10 108

2.1 1.1

14 653 6 310

8.1 3.5

73.8 62.4

31.7

98 569

10.3

52 316

28.8

53.1

13.9 512 942 8.5 14 355 3.4 27 469 0.6 11 226 0.5 26 072

53.4 1.5 2.9 1.2 2.7

17 743 3 474 5 692 1 039 807

9.8 1.9 3.1 0.6 0.4

3.5 24.2 20.7 9.3 3.1

2.3 1.5

3 042 3 756

1.7 2.1

13.5 26.9

12.9 37 014c

20.4

29.9

4 908 883 100.0 960 788 100.0 181 861 100.0

18.9

N

General administration 644 363 Foreign service 6 044 Defence 320 465 Police and public safety 437 850 Legal services 193 974 Education and research 1 554 676 Social service/ semi-statea 684 446 Health service 414 837 Housing 167 803 Agriculture 31 191 Energy/water 22 935 Transportation/ railways 172 522 Public enterprises 75 415 Un-accounted/ othersb 182 362 Total

Gross expenditures

3.5 1.5

22 570 13 977

3.7 123 813

Notes: a. Personnel belonging predominantly to social insurance agencies. b Including personnel of local special purpose associations, respective gross expenditures contained in other categories. c. Excluding local services. Source: Statistisches Bundesamt Wiesbaden 2003. Finanzen und Steuern, Fachserie 14. Reihe 3.1, Rechnungsergebnisse des öffentlichen Gesamthaushalts 2000. Stuttgart: Metzler-Poeschel.

doubled since 1960 (370 800), education, social and health services are the true growth areas since the 1960s. Educational employment, owing to the expansion of secondary and tertiary education in the 1960s and 1970s, increased from 361 800 (1960) to roughly 1.6 million (2000), thus expanding its share from 12.2 to 31.7 per cent in 2000. The expansion of these sectors was compensated in the 1990s by shrinking defence employment

176

The state at work, 1

thanks to the peace dividend and by privatization of the federal and local government transport sector. Thus there are countervailing developments within the public sector that explain why growth of employment has not only stalled during the last decade but was actually turned around after the unification boom; the level of public employment in 2000 is back to where it was in 1990. This decline of total employment, however, does not allow for a revision of Adolph Wagner’s law of ever-increasing public activity and expenditure, for personnel costs amount only to 18.9 per cent of total public expenditure (Table 7.2). Typical of a welfare state in the continental European tradition, in Germany services are often not provided by the public service but transfer payments to citizens enable them to purchase, for instance, health services from private general practitioners; the exception is the almost entirely publicly operated education system. Table 7.2 demonstrates that personnel intensity of a task area is not necessarily related to specific expenditure levels; social services including pensions, comprise 53.4 per cent of gross expenditure, but this amount is transferred by only 13.9 per cent of public employees, whose salaries amount to merely 3.5 per cent of sectoral expenditure. Thus, relatively few people handle the by far largest public budget at low cost. On the other hand, where services are provided by the public sector, the relationship is reversed and personnel costs amount to 50 per cent and more of total sectoral expenditure: this is so in education and research where 53.1 per cent of all education and research expenditures accrue to personnel costs; in legal services (62.4 per cent) and in the police (73.8 per cent) personnel costs consume even more of total respective expenditure. In the defence sector however, where equipment is expensive, personnel costs make up only 51.2 per cent of sectoral expenditure.

PUBLIC EMPLOYMENT BY LEVEL OF GOVERNMENT The Federal Republic of Germany is not a unitary state but a system with a vertical separation of powers between federal and Land governments. Local governments, although legally part of the Länder, are constitutionally granted autonomy ‘within the confines of the laws’ (Article 28, Section 2, Basic Law). Each level is autonomous, in particular with respect to internal organization and personnel policy, although local government’s political organization (constitution) is legislated by the individual Länder. There is therefore some degree of variation of formal structures between units at the lower levels of government, for instance, Hesse does not have local government associations of villages but in the 1960s amalgamated

177

The German public service

Table 7.3

Administrative units of Länder, 2000

Berlin Bremen Hamburg Baden-Württemberg Bayern Hessen Niedersachsen Nordrhein-Westfalen Rheinland-Pfalz Saarland Schleswig-Holstein Total–West

2000 1960

Brandenburg MecklenburgVorpommern Sachsen Sachsen-Anhalt Thüringen

Regional authorities

Cities

Counties

Townships

General purpose associations

– – – 4 7 3 4 5 3 – –

1 2 1 9 25 5 9 23 12 – 4

– – – 35 71 21 38 31 24 6 11

1 2 1 1 111 2 031 421 1 020 373 2 294 52 1 128

– – – 272 319 – 142 – 163 – 119

26 33

91 137

237 425

8 434 24 371

1 015 –

–

4

14

1 475

152

– 3 3 –

6 7 3 6

12 22 21 17

994 537 1 286 1 013

118 119 188 95

5 305 7 565a

672 –

Total–East

2000 1989

6 15

26 n.a.a

86 227b

Total

2000

32

117

323

13 735

1 687

Notes: a. Cities and counties not differentiated in official GDR-statistics. b. 189 rural and 38 urban counties. Source:

Statistisches Jahrbuch Deutscher Gemeinden 2000, Berlin: Deutscher Städtetag.

small local units with the towns. Furthermore, some Länder do not have regional authorities below the Land ministerial level (Table 7.3). German federalism is implementation federalism (Vollzugsföderalismus) with almost all federal laws being implemented through the Länder and ultimately through local governments. Personnel are allocated in different measure to the levels of government in the federation. The federal government, starting as the biggest employer in 1960 (43 per cent), had fallen into third place by 2000 (13.2 per cent). With 50.5 per cent, the Länder were, in 2000, the largest public sector

178

The state at work, 1

employer, followed by local governments with 36.1 per cent (see totals in Table 7.5). Thus the federal centre is lean. Two structural changes contributed to the relative loss of federal employment: the territorial expansion connected with a broader base of subgovernmental personnel and federal government privatization in 1994/95 of post offices, telecom and railways (roughly 1 million people). The federal level ministerial bureaucracy consists of the chancellor’s office and 14 ministries employing in 2001 (together with the president’s office and the court of accounts) 18 121 public servants in Bonn and, since 1999, Berlin. Of these 5078 are higher civil servants. While the chancellor’s office is relatively small (527 personnel, 149 higher civil servants), the defence and finance ministries as well as the foreign office are the largest departments in terms of ministerial staff, not in terms of total personnel (Table 7.4). The size of the field staff of the federal ministries or their personnel in agencies is implied in the difference between total federal government personnel and ministerial personnel in Table 7.4. Although the number of non-ministerial personnel is 15 times greater than ministerial staff, Table 7.4 illustrates that most federal government departments are hardly engaged in policy implementation, the exceptions being defence, foreign affairs, customs (under finance ministry purview) and the federal border police (interior). In the other departments, non-ministerial personnel work in some 50 quasi-autonomous governmental and 250 quasiautonomous non-governmental, semi-state organizations as well as in the five federal courts of justice; these federal agencies, corporations and courts are traditionally located throughout the country. As a rule, the Länder are in control of cultural affairs (universities and schools, 78.3 per cent of total education employees in 2000), the police (66 per cent), the legal services including prisons (97.8 per cent) and general administration (42.7 per cent, Table 7.5). Local governments employ heavily in health (82.7 per cent) and social services (65 per cent), in housing (86.9 per cent, a small absolute number though) and in general administration. In fact, local governments employ the largest share (45.2 per cent) of the entire general administration personnel. Staffing differences between the western and eastern Länder decreased considerably after 1991. By 2000, the new Land government personnel had been reduced by 94 000 staff (minus 18.6 per cent) and local governments had shrunk by 337 000 people (minus 50.9 per cent). During that decade downsizing was considerably less dramatic in the western Länder and local governments (minus 9.9 and 11.8 per cent respectively). Nevertheless, the population ratios show that service density is still considerably higher in eastern Germany than in the western part of the country, but less so at the newly created state level than at the local government level. Here a number

179

Source:

Note:

562 998 – – 296 708

– 6 887 56 327 4 496 48 966 8 422 4 325 2 495 26 232 129 880* 2 573 4 545

Federal Budget, 2001 (positions).

* Without conscripts.

Chancellery Foreign affairs Interior Justice Finance Economics Food/agriculture Labour Traffic/housing Defence Environment Health Family/female affairs Development aid Research Federal President Accounting office Total

Department

Total federal gov. personnel

Federal government personnel, 2001

527 2 079 1 268 633 2 148 1 567 887 957 1 576 2 503 742 492 420 543 908 162 709 18 121

Total ministerial personnel

Table 7.4 Civil servants 297 1 405 909 411 1 503 957 660 583 945 1 519 436 315 282 350 561 87 602 11 822

Percentage of total 56.4 67.6 71.7 64.9 70.0 61.1 74.4 60.9 60.0 60.7 58.8 64.0 67.1 64.5 61.8 53.7 84.9 65.2

Higher civil servants 149 477 331 189 637 444 293 287 518 493 256 164 121 187 311 35 186 5 078

Percentage civil servants 50.2 34.0 36.4 46.0 42.4 46.4 44.4 49.2 54.8 32.5 58.7 52.1 42.9 53.4 55.4 40.2 30.9 43.0

Section heads 53 104 97 60 123 132 78 95 118 137 67 55 41 55 76 15 76 1 382

Subdivision heads 16 23 14 14 30 22 12 16 24 21 13 10 8 8 15 2 10 258

Division heads 6 12 10 6 11 8 7 8 7 7 6 4 4 3 7 3 1 110

1 2 2 1 3 2 1 2 2 2 1 1 1 1 1 1 1 25

State secretaries

180

370 800 – 233 500 184 000 93 000 361 800 111 200 173 000 91 400 20 500 16 600 93 500 201 600 489 900 359 500 16 000 14 000 2 818 300 141 100 2 959 400

1960a 480 300 – 409 000 312 300 145 100 809 200 140 600 354 000 165 300 22 900 19 600 101 000 133 000 246 600 439 200 39 800 – 3 817 900 230 500 4 048 400

1990a

Total

644 400 6 000 320 500 437 900 194 000 1 554 700 196 400 414 800 167 800 31 200 22 900 98 900 75 400 73 600 – 69 800 112 600 4 420 900 488 000 4 908 900

2000 14.9 – 100.0 8.9 2.7 1.3 1.4 0.3 – 2.6 7.1 22.4 3.3 100.0 100.0

32.5

43.0

1990

18.9 – 100.0 9.6 1.3 1.4 0.9 – – 2.0 17.5 24.4 2.3 100.0 100.0

1960

13.2

12.1 99.2 100.0 9.7 2.2 1.0 2.0 0.5 – 1.3 8.8 26.7 – 100.0 –

2000

Percentage federal

33.7

42.5 – – 64.1 98.7 85.6 50.1 29.5 31.1 91.2 69.9 34.4 37.3 – –

1960

40.2

45.2 – – 69.2 97.3 84.7 35.3 19.3 17.3 89.1 64.3 37.0 44.4 – –

1990

50.5

42.7 0.8 – 66.0 97.8 78.3 33.0 16.9 13.1 92.9 48.2 43.7 41.6 – –

2000

Percentage Länder

Source:

Derlien et al. (2005), p. 25ff.

Notes: a. 19601990 full-time personnel only, 2000 including part-time employment. b. Before 1991 included in general administration. c. Railway personnel are the remainder, after privatization, often civil servants seconded to the new privatized corporation. d. Author’s calculation.

General administration Foreign serviceb Civil and military defence Police, public security Legal services Education/research Social service Health service Housing Agriculture Energy/water Transportation Public enterprises Railwaysc Post and telecom. Local special purpose ass. Un-accountedd Total ‘core public service’ Semi-state Total public service

Level of government

Table 7.5 Public employment by policy area and by level of government, 1960–2000

23.3

38.6 – – 26.3 – 13.0 49.0 70.5 68.9 6.8 12.6 41.2 60.4 – –

1960

27.3

40.0 – – 22.0 – 14.0 63.2 80.4 82.7 8.3 28.6 40.6 52.3 – –

1990

36.1

45.2 – – 24.3 – 20.8 65.0 82.7 86.9 5.8 43.0 29.4 58.4 – –

2000

Percentage local

The German public service

181

of facilities that are privately operated in the West are still under public tutelage, often inherited from the industrial conglomerates (housing, kindergartens). The federal government, although always the biggest spender, carries the absolutely and relatively smallest burden of personnel costs of the three levels of government (some 11 per cent). Owing to privatization as well as the tremendous expenditure of transfers to eastern Germany, even the relative share of personnel expenditure in the federal budget has gone down since 1990. As the Länder are responsible for administration in the most important policy areas, not only are their personnel the most voluminous in numbers but personnel expenditure is the single most important item in the Länder budgets. The statistical category ‘personnel expenditure’, however, includes pensions for retired civil servants. The east–west Länder difference in personnel costs is due largely to the fact that the retired GDR state functionaries derive their pension from the general (semi-state) insurance fund and not from the current Land budget; furthermore, the number of newly appointed civil servants who have already retired is (naturally) very small in the new Länder; thus the share of personnel expenditures in their budget is only 26 per cent whereas in the West it amounts to roughly 45 per cent.

SOCIAL DIFFERENTIATION OF THE PUBLIC SERVICE Empirically, there is no ‘typical’ public servant nor is there something like an esprit de corps giving the public service social cohesion. At best, the public service stands united when attacked politically by outside groups. On the other hand, the public service is well perceived as special because of the differences from the rest of the labour market, in particular with respect to job security. Otherwise, public perception is not sophisticated enough to judge the public service along formal lines of differentiation: whether someone is a federal government civil servant or a local government employee is not reflected in opinion polls, nor is whether one works in a state revenue office, with the police or in a local hospital covered under social generalizations and the stereotype of servant of the state. The expert, aware of the heterogeneity of the public service, will start analysing cleavages according to legal distinctions: status groups, career groups and employer. Genuinely social categories for coming to grips with social differentiation are invited when social representativeness, in particular the public service elite and the role of women in public employment, are under consideration.

182

The state at work, 1

The Dual Structure of the Public Service As mentioned before, there are two classes of public servants: first those with the traditional public law status of civil servants, who however are not limited to the administrative elite as in Britain nor can be regarded a (grand) corps as in France; civil service status extends to policemen and teachers, too. Second, there is the historically more recent group of those working under private law contracts, (manual) labourers and (white-collar) employees. Article 33, Section 5 of the 1949 federal constitution states that the public service should be regulated according to its ‘traditional principles’, which implies the principles valid in 1919 and even in 1873. These traditional principles making up the allegedly privileged civil service status are: ●

●

●

●

●

●

● ●

Guarantee of the institution of the civil service as such; unilateral regulation of status and salary by law as well as appointment by authoritative act and not through a contract; some restrictions on basic rights (for example, the right to go on strike). Protection and support by ‘the state’, which entails among other things that civil servants’ pensions and health care are paid out of taxes (as opposed to public employees and labourers, who contribute to the general social insurance system). Obligation of the civil servant to devote his or her full work capacity to ‘the state’ without engaging in additional professional activities or going part-time. Salaries are not regarded as payment for work done but as a contribution to allow for a living standard adequate to the status of the official. Duty of political loyalty to the constitution and to observe party political neutrality when in office; furthermore, political self-restraint in public. Deviations from civil service duties are sanctioned according to a disciplinary code and, in severe cases, in separate court procedures. Private conflict with the law can invoke disciplinary reactions. Legal protection against the state as employer by access to administrative courts (and not the normal social and labour affairs courts) corresponds to public law regulation of the terms of employment. Co-determination/representation of interest (hearing rights) in the agency (since 1959) as far as organizational and personnel affairs are concerned. Professional training; achievement principle in recruitment (according to education certificates) and promotion. Career system (as opposed to job staffing for employees) and lifelong tenure; automatic increases of the base salary in certain intervals as

The German public service

183

a loyalty premium, independent of promotion and periodic salary adjustments. Civil servants and their association tend to defend this special status, although it is difficult to detect the rationale behind this attitude as public employees enjoy similar job security (after 15 years) and equivalent monetary rewards (although differently calculated because of their obligation to contribute to social security funds). Furthermore, many of the specific duties not mentioned here and even some of the principles can be found in the treaties that public service unions negotiate for employees. In fact, the civil service serves the unions as a model; on the other hand, codetermination through councils in offices and agencies in matters concerning the working environment and staffing was included in the assembly of traditional principles after the public service unions had successfully negotiated for this sort of participation. Whereas positions that are defined as entailing authoritative capacity in the law and order branches (e.g. police and revenue service) must be staffed with career civil servants (Article 33, Section 4, Basic Law), the other groups are found predominantly in social and health services, in research and in clerical functions (employees) or do physical work, for example in local waste disposal. For them job staffing is typical. However, employees often work on jobs issuing authoritative acts and so should be carried out by civil servants. On the other hand, in the state-operated education system, teachers and professors enjoy civil service status although delivering personal social services. This dualism of the German public service is not undisputed. In 1973 an unsuccessful attempt was made to harmonize public service law and to create a uniform public service, be it by extending public law status to all groups, or by changing the constitution and abandoning the special status of the civil service. However, the two-third majority necessary for constitutional amendment was not available, thus the differentiation persisted. Today critics worry mainly about the pension load conferred on future budgets and seek to limit civil service status to ‘core’ task areas. In 2000 there were, among the total of 4.4 million public servants in core public administration, 1.8 million civil servants of federal, Länder and local governments (41.1 per cent), 2 million public employees (44.5 per cent), and 636,600 labourers (14.4 per cent) (Table 7.6). In 1950, labourers (39.3 per cent) had constituted the relatively largest public service group, followed by civil servants (35.8 per cent) and employees (24.9 per cent). By 1990, before unification, the number of civil servants had more than doubled and brought their portion up to 44.1 per cent of the public service, while employees had gained 33.8 per cent and overtaken the labourers

184

The state at work, 1

Table 7.6 Public employment by status group, 1950–2000 (core public administration) Total public employment

Civil service

Employees

Labourers

(N  100%)

(N)

(%)

(N)

(%)

(N)

(%)

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991

4 420 900 4 512 100 4 619 300 4 713 300 4 813 900 4 921 100 5 666 400 6 114 800 6 305 000 6 412 700

1 818 800 1 831 600 1 845 000 1 859 700 1 853 200 1 848 200 2 141 600 2 134 300 2 112 500 2 069 000

41.1 40.6 39.9 39.5 38.5 37.6 37.8 34.9 33.5 32.3

1 965 500 2 014 500 2 075 600 2 122 000 2 176 800 2 235 700 2 402 300 2 595 200 2 701 600 2 738 400

44.5 44.6 44.9 45.0 45.2 45.4 42.4 42.4 42.9 42.7

636 600 666 000 698 700 731 600 783 900 837 200 1 122 500 1 385 300 1 490 900 1 605 300

14.4 14.8 15.1 15.5 16.3 17.0 19.8 22.7 23.6 25.0

1990 1985 1980 1975 1970 1965 1960 1955 1950

4 656 300 4 598 800 4 430 200 4 201 500 3 692 600 3 396 900 3 004 100 2 510 100 2 199 900

2 054 900 2 062 600 1 972 200 1 843 400 1 642 200 1 506 400 1 321 400 1 052 200 788 400

44.1 44.8 44.5 43.9 44.5 44.4 44.0 41.9 35.8

1 575 300 1 475 400 1 401 900 1 278 100 1 041 100 871 700 698 000 563 500 547 900

33.8 32.1 31.6 30.4 28.2 25.7 23.2 22.5 24.9

1 026 100 1 060 800 1 056 100 1 080 000 1 009 300 1 018 800 984 700 894 400 863 600

22.0 23.1 23.8 25.7 27.3 30.0 32.8 35.6 39.3

Percentage change 91–2000 31.1 1990–91 37.7 1980–90 5.1 1970–80 20.0 1960–70 22.9 1950–60 36.6 1950–00 101.0

12.1 0.7 4.2 20.1 24.3 67.6 103.7

28.2 73.8 12.4 34.7 49.2 27.4 258.7

60.3 56.5 2.8 4.6 2.5 14.0 –26.3

Source: Derlien et al. (2005, p. 42).

(22.0 per cent). The number of labourers had stagnated whereas civil servants had mushroomed primarily as a consequence of the expansion of the education and police sectors. Unification with East Germany was bound to bring about frictions when the eastern functionaries were to be transferred to this status group system. Most of them remained in public employee status, thus increasing the

The German public service

185

absolute number and the share of that status group by 9 per cent to 42.7 per cent in a single year and thus lowering the relative civil service share from 44.1 per cent in 1990 to 32.3 per cent in 1991. Privatization of public enterprises and contracting out by all levels of government led to a reduction of more than 50 per cent of the absolute number of labourers since 1991; with 14.4 per cent they comprised an historical low in 2000. In 2000, civil servants dominated the personnel structure of federal and Land governments (education, police, general administration) with 67 and 55 per cent respectively. Local government personnel consisted, first of all, of employees (63 per cent) and, secondly, of labourers (26 per cent, see Table 7.7). The dominance of civil servants in traditional, partly authoritative, state sectors contrasts with the social service and health sectors where employees work overwhelmingly at the Land and local levels. The civil service is more dominant at the federal government level in 2000 than during the five decades before; it jumped up after the privatization of rail and mail in 1994/95 and the consequent loss of labourers. The decrease in Land civil servants since 1990 is due to keeping teachers in eastern Germany in employee status, therefore also enlarging that status group. By 2000, the Land civil service share was back to normal (55 per cent as in 1980).

FOUR CAREER GROUPS Professional training, the merit principle and the career system were already reflected in Max Weber’s concept of bureaucracy. There are four categories of public servants corresponding to the hierarchy of educational institutions in Germany and certificates obtained there. First, the university trained higher civil service make up 22 per cent of the civil service (2000), whereas employees with equivalent education comprise only 13 per cent of that status group. The group below (with A-level certificates, the German Abitur that qualifies for university admission) amounts to 46 per cent; it increased considerably since 1994. The next career group corresponds to the British A level or a US high-school diploma; until 1994 it used to be the largest civil service career group (45 per cent) but shrunk by 2000 (33 per cent) as the elevated service expanded. Owing to ‘structural improvements’ civil servants, in particular those in the police force, were collectively moved up into a higher career group with a better salary. The ‘simple service’ comprises those who have completed basic schooling after nine years’ attendance; this group, owing to upgrading in the past, has almost disappeared. Two-thirds of the employees are in jobs equivalent to the middle and simple civil service – an indication of the clerical and personal social service jobs staffed by employees. Taken together, 48 per cent civil servants and

186

of which:

Länder of which:

Local

Source:

Derlien et al. (2005), p. 43.

15 15 17 15 15 16 26 32 31 35 28 30 31 37 51

2273.3 2313.7 2363.1 2401.9 2429.9 2453.5 2482.0 2510.8 2531.4 2571.9 1934.8 1823.4 1334.3 1003.8 739.9

55 54 57 52 51 50 48 46 44 42 55 55 53 52 44

39 39 36 40 41 42 43 43 44 47 34 34 33 30 34

6 7 7 7 8 8 9 11 11 11 10 11 14 18 21

1572.0 1609.7 1648.5 1683.3 1739.3 1801.6 1873.3 1946.7 2073.5 2051.3 1358.1 1401.6 1106.5 733.5 574.0

11 11 11 11 10 10 10 9 8 8 12 25 27 16 18

63 63 62 62 61 60 60 60 59 59 55 44 38 38 37

26 26 27 27 29 30 30 31 33 33 33 31 35 46 45

* Core public administration only; add roughly 500 000 semi-state personnel for 2000 to match with totals in the other tables.

18 18 18 18 18 18 16 20 21 18 12 11 12 9 9

Note:

67 67 65 67 67 66 58 48 48 46 60 59 57 54 40

575.6 588.7 607.7 628.0 644.7 666.1 1311.1 1657.3 1700.1 1789.4 1362.9 1410.0 1410.5 1266.7 885.9

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1980 1970 1960 1950

4420.9 4512.1 4619.3 4713.3 4813.9 4921.1 5666.4 6114.8 6305.0 6412.6 4655.8 4635.0 3851.3 3004.0 2199.8

Total Civil Total Civil Total Civil Grand (100%) servants Employees Labourers (100%) servants Employees Labourers (100%) servants Employees Labourers total (%) (%) (%) (%) (%) (%) (%) (%) (%)

of which:

Federal

Table 7.7 Public employment by level of government and status group (000s), 1950–2000*

The German public service

187

employees work in middle and lower ranks, 35 per cent in elevated ranks (often called the backbone of the German public service) and 17 per cent in higher civil service jobs. The university training in law of the higher civil service and internal law-focused administrative college training of the career group below produce the classical generalist civil servant with common professional values and skills. All leading functions in the general administrative branch including the ministries are, as a rule, staffed with civil servants. Table 7.4, registering federal departmental staff according to career groups, might serve as an illustration: 65.2 per cent of all ministerial positions were staffed with civil servants, of whom 43 per cent were higher civil servants (5078 out of 18 121 ministerial staff). It might be questioned whether the public service is still held together by a strong esprit de corps as it allegedly was in pre-democratic times. Nor is it any longer an estate (Berufsstand) as it was perceived in Max Weber’s days. A peculiarly subjective understanding of roles might though be found in top positions of the service; these are responsible for personnel policy and have encounters with the union-dominated councils in their agencies. ‘Class consciousness’ as documented in trade union membership varies along status group lines (Ellwein 1980). It might rather be their function than their legal status that engenders a certain esprit de corps among the higher civil servants. The Administrative Elite After the Second World War the so-called monopoly of jurists was reestablished. However in the early 1970s, influenced by the Fulton report in the UK, the notion began spreading that leading civil servants were not adequately prepared for management functions implied in their job. This also led to the acceptance of economics and social science degrees as general entry qualifications equivalent to a law degree – at least for the federal higher civil service and most of the Länder. Since 1966 the share of economists in the federal administrative elite steadily increased up to roughly 15 per cent in 2000. Nevertheless, jurists are still the backbone of the higher civil service. Law is also the subject most often studied by executive politicians (Derlien 1990). Although in Germany there is no centre for educating the administrative elite as the ENA in France or Oxbridge in Britain, law training in the universities, regulated by federal law, provides for an abstract elite (Dahrendorf), for the curriculum for judges, state attorneys, barristers/solicitors as well as for the higher civil service is the same throughout the entire university system, thus providing a high degree of uniformity of training and (it may be assumed) a common basic role understanding of all law professions and

188

The state at work, 1

the higher officials in all branches of government and at all levels of the administrative system. Another aspect of the administrative elite frequently discussed in the western world is their politicization (Derlien 1996b). Despite the performance principle in recruitment and promotion, informal political streamlining of the ministerial bureaucracy in the two top ranks, best visible after changes in government (1969, 1982, 1998), has become normal in Germany. The most modest form of gearing top career civil servants to the political requirements of the day is to reshuffle them and place those regarded as in disgrace into less politically sensitive positions. If reshuffling is not possible, new positions are occasionally established and filled with trustees in order to circumvent or control mistrusted office-holders. Another widespread means of political control through personnel policy consists of selective promotion of political trustees, for value consensus between minister and top administrators reduces the need to communicate normative political decision premises. Thus, patronage is not limited to the American ‘spoils system’ but is possible even in a closed career system like the German one. Increasing party-politicization has been evident since 1970. Among the federal bureaucratic elite, party membership spiraled from 28 per cent in 1970 to 36.7 (1972), 51.7 (1981), 57.3 per cent (1987) and 60 per cent (1995) (see Mayntz and Derlien 1989, p. 397ff.; Derlien 2003). So long as party affiliation is treated as a criterion additional to professional performance during an administrative career and not as an exclusive criterion of staffing, the historically inherited system of matching professional expertise and political loyalty is not getting out of balance. Another, even institutionalized possibility is the German particularity of temporary retirement of so-called ‘political civil servants’. Political civil servants are required to share the policy principles of the government of the day and can be temporarily retired any time on any grounds without justification. Cases accumulate after government changes. In all these instances, of the 25 state secretaries roughly every second one and of the 110 division heads every third one (see last two columns in Table 7.4) have been temporarily retired (Derlien 1988; 2003). The other side of the coin, however, is that incoming governments carry on working with the majority of this administrative elite and the ranks below, subdivision and section heads. Furthermore, most of the vacancies are staffed with candidates from within the ministries or experienced officials from Länder governments. At most 10 per cent of the new appointments accepted by the civil service commission can be classified as outsiders. This reminds us that ‘political civil servants’ are not defined by the way they are recruited but by the modalities of being put into temporary retirement (although hardly ever being reactivated).

The German public service

189

SOCIAL AND GENDER REPRESENTATIVENESS Article 3, Section 2 of the federal constitution prohibits any racial, religious, social, political or gender discrimination in general, and Article 33, Sections 2 and 3 restate the case for access to public office. The only legally admitted selection criteria are feasibility (for example, physical fitness), qualification (as documented in educational certificates) and performance (in job-related entry examinations or in periodic performance appraisals). Furthermore the merit principle is counted among the traditional civil service principles. Thus, any lack of representativeness of the public service for the entire population is supposed to be the result of lack of qualifications or of self-recruitment processes that make public service or certain areas of it less attractive for parts of the population. None of the potentially discriminating criteria except gender (since 1986) is recorded in official statistics. The issue of race in the public service is legally and factually non-existent and, for historical and constitutional reasons, must not be surveyed. Party membership must not even be asked. The following generalizations are nevertheless possible: ●

● ●

Because the career system is geared to school certificates, the social background of the public service in its entirety is representative of the population. Nevertheless, owing to self-recruitment and filters in the education system, the middle class is over-represented in the higher civil service. There is a fair regional representation thanks to the territorial principle embodied in the macro-structure of the executive branch and the practice of distributing federal agencies throughout the country in response to the imperative of Article 36, Section 2 Basic Law to adequately represent ‘all German tribes’ in the federal administration.

Due to the accession of the female-dominated eastern German public service (60 per cent women), public service employment became evenly distributed between the sexes in the 1990s. Female employment (Heinemann 2002) increased between 1960 (23 per cent) and 1990 (42 per cent) but reached parity only in 1996 (Table 7.1). Female employment in 2000 is most widespread in local government ( 58.6 per cent) but rare in the federal administration (33 per cent), while the Länder service is composed of 51 per cent female personnel. Women are over-represented among the part-time employees, amounting to almost 90 per cent of total part-time public employment. Furthermore, women dominate the employee group (68.2 per cent) and make up roughly one-third of the civil servants (38.1 per cent) and labourer groups (36.1 per cent). This statistical pattern is linked to

190

Table 7.8

The state at work, 1

Female employment by policy area, 1986–2000 1986a Total

General administration 558 196 Foreign serviceb – Civil and military defence 429 606 Police, public security 318 464 Legal services 155 413 Education/research 1 219 148 Social service 165 360 Health service 346 398 Housing 175 470 Agriculture 28 406 Energy/water 23 757 Transportation 106 215 Public enterprises 145 220 Railways 288 086 Post and telecom. 529 217 Local special-purpose organizations 48 752 Semi-state 251 383 Unaccountedc 84 909 Total public service

4 874 000

1991

2000

Women (%)

Total

Women (%)

Total

Women (%)

42.0 –

725 760 5 728

49.8 40.5

644 363 6 044

52.8 45.6

11.7

462 721

13.9

320 465

15.8

17.4 42.7 54.2 64.4 80.3 15.5 29.1 28.6 9.4 17.7 6.2 34.4

421 704 181 189 1 740 807 295 834 682 639 216 629 37 472 33 310 125 357 182 681 473 827 663 499

21.5 46.1 61.2 73.6 69.6 20.5 38.8 31.3 14.0 21.7 18.1 42.4

437 850 193 974 1 554 676 196 458 414 837 167 803 31 191 22 935 98 903 75 415 73 619 –

26.4 51.1 63.1 68.2 67.4 26.9 41.4 41.4 16.6 23.2 9.3 –

52.2 55.4

55 518 325 135 107 955

53.9 61.5

69 788 487 988 112 574

50.0 61.1

41.0

6 737 765

46.8

4 908 883

50.8

Notes: a. Data for female employment by policy area not contained in official statistics before 1986. b. Prior to 1991 included in general administration. c. Own calculation. Source: Derlien et al. (2005), p. 62.

policy areas of employment (Table 7.8). Since 1986 (and before) women were over-represented in social services (68.2 per cent in 2000), the health services (67.4 per cent), education (63.1 per cent) and the semi-state social insurance system including labour administration (61.1 per cent). In general administration and the legal services their share, owing to the incorporation of the female-dominated eastern German public service, climbed

The German public service

191

to 52.8 per cent and 51.1 per cent respectively in 2000. On the other hand, women are underrepresented in the authoritative state sectors like the military (15.8 per cent) and the police (26.4 per cent) where civil service status and full-time employment is or was imperative. In the second highest career group though, the female share climbed to 52.1 per cent (2000) from 21.7 per cent in 1960, and in the career groups below they had already reached 50 per cent by 1990 (Derlien et al. 2005). Although the gender proportions are balanced in the public sector overall, the rank distribution of females in government employment is not as even. Despite equal opportunity measures since the 1980s and female commissioners in all offices, women hold relatively fewer positions at the top of the civil service. Starting off from 12.1 per cent in 1960 women reached the 20 per cent threshold of the higher civil service by the end of the 1970s and gradually broadened their presence to 34.7 per cent in 2000. Since women started to enter the system of higher education in larger numbers only in the late 1960s, their share of top positions was also gradually growing, with a time lag of some 20 years, the normal duration of this ‘march through the institutions’.

PUBLIC SERVICE MANAGEMENT The NPM tool kit, reflecting the mostly unitary nature of Commonwealth states, includes the proposal to decentralize personnel management from central to line departments and from ministries to agencies. As is the case in all federal systems, in Germany, personnel management was always decentralized on to the three tiers of the federal structure; each Land and local government is autonomous in recruiting and promoting personnel. Furthermore, government departments at the federal and Land levels are autonomous, too; what NPM is suggesting has been the rule in Germany for a long time. However, all levels and departments are operating within a common legal framework of federal civil service legislation, federal salary law and centrally negotiated work and wage conditions for employees and labourers. In other words: while micro-management is totally decentralized, there is a national framework regulating inter alia duties and rights, civil service entry requirements and payment. The ministries in charge of drafting, negotiating and adjusting this framework are, at federal and Land levels, the ministries of the interior, and local governments act through their associations in ministerial and parliamentary hearings as well as in the annual national public service employer negotiations with the unions. Of the public service, 89.6 per cent are members of labour unions, first of all the diminishing group of blue-collar workers and the employees, but

192

The state at work, 1

only 36 per cent civil servants. Following the fusion of major public service unions in 2000 so that they now represent 3 million public servants (8.6 per cent of whom are civil servants), civil servants are mainly represented by the German Association of Civil Servants (Deutscher Beamtenbund, 1.2 million members in 37 special branches) with 72 per cent of their members holding civil service status, the police union (69 per cent civil servants) and the teachers’ union (46 per cent civil servants). Although civil servants must not strike, civil service associations have hearing rights when the civil service code is amended or salary increases are to be legislated (Keller 1993). Typically, these hearings take place after the public service unions have negotiated with employer representatives of federal, Land and local governments and bargained improvements ‘in the shadow of the strike’. There have been only two large public service strikes since the Second World War, in 1974 and in 1992. In 2000 a strike was narrowly averted.

CONCLUSION Four trends in the German public service warrant emphasis. No skeleton state Unless fiscal conditions deteriorate further than under the current economic stagnation and concomitant high social subsidy payments, the overall size of the public service will only slightly diminish below the year 2000 level; further reduction of service intensity can at present only be expected in the new Länder. After the privatization of the major public enterprises, both at the federal and Länder levels, and the re-allocation of public properties within the eastern Länder there is only a narrow margin for further privatization. Also, the military capacity will be maintained with about 270 000 compulsory and professional servicemen. The only policy areas in which public employment will grow might be the already dominant fields of general education and university training and research as well as public health care including provisions for the elderly. Third, the law and order capacities will not decrease, given the present population size, urbanization and open borders of the Federal Republic. Uniformity of the service? The status group system and the institution of civil service may be loosened or even given up and be replaced by a uniform public service code, depending on the political process and public opinion. Most likely though the Länder, in anticipating future pension burdens, will try to transfer as much public employment as possible to the employee category. The

The German public service

193

labourer category might, in the course of further marginal privatization, disappear altogether. Another erosion of differentiation might apply to the career group system; at least more transition between the elevated civil service and the higher civil service is imaginable while the lowest career group may completely disappear and become merged with the last middle group. Temporarization Unorthodox employment (Derlien 1999) is likely to gain in importance; both part-time and temporal employment might increase, in particular under conditions of a uniform public service code. Part-time employment will, for a considerable period of time, remain closely related to female employment, although male part-time employment is imaginable, in particular in less qualified jobs (as in the UK). Feminization With an increasing share of total employment, women will move into higher positions in larger proportions owing to their growing participation in the higher education system. This will occur in the traditional female domains of education, social services and health care, but possibly also in the judiciary where flexible working hours are the rule. Civil service status will no longer bar this development after the doors were pushed open in 1997 for part-time civil service employment, even at the entry level. Some of these changes are intended politically; downsizing public employment and temporarization are undoubtedly planned responses of public employers to fiscal problems and, as far as part-time employment is concerned, to unemployment of young teachers. Broader and more sweeping changes occurred in eastern Germany as a consequence of joining the Federal Republic, when the differentiated western public service system was imposed on the formerly unitary system. In this case it has become customary to speak of transformation. After German unification in 1990, there were though, transformative processes of the welfare state underway in the western part of the country too. Privatization and contracting out appear to have fundamentally affected the traditional self-production of services by the state. The increasing employment of women, whether partly unintended or resulting from a push-and-pull process, in turn might have encouraged part-time employment. Despite these autonomous trends, intended partial changes and systemic transformations, the public service has preserved many of its traditional structures, most notably the dual system of employment with the persisting status of the civil service, its principles and primary employment areas.

194

The state at work, 1

NOTE 1. Employment in semi-state agencies and corporations is not included in some of the tables below, rather data refer mostly to core public administration (unmittelbare Staatsverwaltung).

REFERENCES Derlien, Hans-Ulrich (1987), ‘State and bureaucracy in Prussia and Germany’, in Metin Heper (ed.), The State and Public Bureaucracies. A Comparative Perspective, New York: Greenwood, pp. 89–105. Derlien, Hans-Ulrich (1988), ‘Repercussions of government change on the career civil service in West Germany: the cases of 1969 and 1982’, Governance, 1, 50–78. Derlien, Hans-Ulrich (1990), ‘Continuity and change in the West German federal executive elite 1949–1984’, European Journal of Political Research, 18, 349–72. Derlien, Hans-Ulrich (1991), ‘Historical legacy and recent developments of the German higher civil service’, International Review of Administrative Sciences, 57, 385–401. Derlien, Hans-Ulrich (1993), ‘German unification and bureaucratic transformation’, International Political Science Review, 14, 319–34. Derlien, Hans-Ulrich (1996a), ‘Germany: the intelligence of bureaucracy in a decentralized polity’, in Johan P. Olsen and B. Guy Peters (eds), Lessons from Experience. Experiential Learning in Administrative Reforms in Eight Democracies, Oslo: Scandinavian University Press, pp. 146–79. Derlien, Hans-Ulrich (1996b), ‘The politicization of bureaucracies in historical and comparative perspective’, in B. Guy Peters and Bert A. Rockman (eds), Agenda for Excellence 2. Administering the State, Chatham, NJ: Chatham House, pp. 149–62. Derlien, Hans-Ulrich (1999), ‘Unorthodox employment in the German public service’, International Review of Administrative Sciences, 65, 13–23. Derlien, Hans-Ulrich (2000), ‘Actor constellation, opportunity structure and concept feasibility in German and British public sector reforms’, in Hellmut Wollmann and Eckhard Schröter (eds), Comparing Public Sector Reform in Britain and Germany. Key Traditions and Trends of Modernization, Aldershot: Ashgate, pp. 150–170. Derlien, Hans-Ulrich (2003), ‘Mandarins or managers? The bureaucratic elite in Bonn, 1970 to 1987 and beyond’, Governance, 16 (3), 401–28. Derlien, Hans-Ulrich and B. Guy Peters (1998), Who Works for Government and What Do They Do? Common Trends and Structural Differences of Public Employment in the United States and Germany, Verwaltungswissenschaftliche Beiträge, 32, Bamberg: University of Bamberg. Derlien, Hans-Ulrich, Stefan Frank, Silke Heinemann and Stefan Lock (2005), The German Public Service Structure and Statistics, Bamberger Verwaltungswissenschaftliche Beiträge, 35, Bamberg: University of Bamberg. Ellwein, Thomas (1980), Gewerkschaften und öffentlicher Dienst, Opladen: Westdeutscher Verlag. Heinemann, Silke (2002), Teilzeitarbeit im öffentlichen Dienst, Frankfurt: Lang. Keller, Berndt (1993), Arbeitspolitik des öffentlichen Sektors, Baden-Baden:Nomos.

The German public service

195

Mayntz, Renate and Hans-Ulrich Derlien (1989), ‘Party patronage and politicization of the West German administrative elite 1970–1987 – Towards Hybridization?’, Governance, 2, 384–404. Wagner, Adolph (1911), ‘Staat in nationalökonomischer Hinsicht’, in J. Conrad, L. Elster, W. Lexis and E. Loening (eds), Handwörterbuch der Staatswissenschaften, vol. 7 (3rd edn), pp. 727–39.

8.

Working for the government in Spain: from authoritarian centralism to democratic political devolution Carlos R. Alba and Carmen Navarro

Since the 1960s, Spain has undergone dramatic political and economic change, commencing with the modernizing strategies of the new technocrat advisers in Franco’s government, followed by the successful political transition to democracy in the late 1970s and the extension of the welfare state under the socialist government in the 1980s and, more recently, the policies of vigorous privatization at the end of the 1990s. This chapter traces the concomitant changes in public employment, focusing in particular on the extensive political decentralization that has accompanied the process of change. Any analysis of public employment in Spain must make reference to Spanish ‘public administrations’ (plural). The machinery of government is organized into a complex administrative network comprising: ● ● ● ● ●

central administration, frequently referred to as ‘state administration’; peripheral administration, referring to the central administration presence in the several territories; institutional administration, comprising the public companies and autonomous bodies linked to specific ministries; regional administration, denoting the 17 new political actors in democratic Spain; and local administration (Ayuntamientos and Diputaciones).

The newly-created regional level of government has emerged during more than 20 years of a process of transition and democratic consolidation. The country is now divided into 17 Autonomous Communities (ACs) and two Autonomous Cities (Ceuta and Melilla) in Northern Africa, all with their own political and bureaucratic structures. At the same time, the municipalities are undergoing a democratic revitalization. These new 196

Working for the government in Spain

197

political arrangements enhance participation and decentralization and are a primary factor in explaining the growth in public employment (Table 8.1). Another important factor has been the impact of the development of the welfare state, which was not fully implemented until the beginning of the 1980s. New programmes in unemployment benefits, health coverage and universal education were developed in a country which at the time lacked a public service with the size and effectiveness required to meet these goals. Unpredicted imbalances were always present as were maladjustments between governmental tasks and the administrative structure. In historical terms, the complex network of administrations, born mainly in the last part of the 19th century, was developed under the influence of the Francoist authoritarian regime (a system that lasted almost 40 years) on a centralist and undemocratic model in which the central government administration played a hegemonic role. Public administration suffered from a continuous barrage of unsuccessful modernizing strategies that, in the end, have prevented the consolidation of a public service equivalent to that in other Western European countries. The establishment of a modern professional civil service has been hampered further by the long-lasting political instability, the constant conflict among political factions and, overall, the dominance of the spoils system. Favouritism and patronage have been dominant features in the recruitment and organization of the civil service. From the very beginning, certain groups of civil servants developed mechanisms to avoid the effects of employment instability and built professional organizations (cuerpos following the French pattern of the corps administratives) with their own regulations, each corps controlling its own recruitment and organizational procedures. Despite these features inherited from the past and still present to some extent, the structure of public employment in Spain today continues to undergo dramatic change, accelerated by trends in the growth of the number of public employees, integration into the European Union, the territorial redistribution of government employees, increasing levels of specialization, the decline of the public economic sector, the increasing number of non civil-servant public employees, the feminization of public employment and increasing ‘quangoism’.

PUBLIC EMPLOYMENT DATA DURING FRANCOISM: A HISTORICAL OVERVIEW One of the unsolved problems of Spanish public administration is the absence of basic information which would allow the government to

198

The state at work, 1

adequately analyze the public sector in employment terms. The fragmented and partial information available is sometimes the result of not maintaining homogeneous criteria for gathering statistical data. In other cases, the problem is attributed to the lack of continuity within the administration and to government secrecy. This lack of knowledge might be understandable were it about public employment in the ‘parastate’ sector, but in fact there was a lack of information even about career civil servants in the central administration. Since 1930, the National Institute of Statistics had published a census of the labour force population with a chapter on public administration that confused data on civil and military personnel with data on public employees at the several levels of government. The main source for public employment data was then the annual budget, but the information did not cover local public employment or other ‘parastate’ organizations. Even with respect to the central administration the information was incomplete because the positions budgeted were never equivalent to the real number of public employees. There were cases of civil servants holding a variety of positions at the same time. Also, labourers working for the government under private law contracts were never included. During a wave of administrative reform in the early 1960s, there was considerable discussion about the lack of data concerning the actual number of public employees. Several studies were conducted, mainly by administrative law professors, attempting to estimate the number of public employees (Baena 1984). The first study appeared in 1964 and included a critical appraisal of data sources for public employment (Beltran 1964). In 1966, the Ministry of Finance began publication of the so-called Red Book of the Budget, which included a ‘genuine estimate’ of the number of public employees at the central level. In the same year, the High Commission of Personnel published an ‘alphabetical index’ including around 200 000 employees (approximately half of the total public employees at that time) but in such a fragmented and irregular form that it defied statistical analysis. When the same office published a study with data on civil personnel working for the central government in 1973 there were approximately 460 000 employees, almost 80 per cent of whom were career civil servants. In 1975, the Ministry of the Presidency attempted a complete analysis of the personnel working in the public sector with the following results: by 1975 Spain had a total public employment of approximately 1.5 million employees, half of whom were professional tenure civil servants (750 000), around 600 000 were labourers and the rest (130 000) were employees with special administrative contract systems. According to the territorial analysis, almost 39 per cent were civil servants or labourers working in the central government (564 780); 19 per cent of the central government employees

Working for the government in Spain

199

belonged to the military administration; 14 per cent were working in local authorities; 17 per cent were employed in autonomous bodies or in field staff in the periphery; 28 per cent were employees in the commercial and industrial public sector; and around 35 000 persons were employed in the Franquist authoritarian legislature and trade union organizations. Twenty years later almost another million had been added to the number of public employees. It is only since the late 1980s that there has been more accurate information on the number and main characteristics of public employment in Spain. This is due mainly to the creation of the Register Office of Personnel, whose Bulletin is one of the main databases for this study (Ministerio de Administraciones Públicas 1990–2000).

PUBLIC EMPLOYMENT AND PERSONNEL MANAGEMENT Until the first quarter of the 19th century, entry to the Spanish bureaucracy was mainly hereditary, pure patronage or simply the traditional ‘sale of offices’ of the Old Regime. In 1852, the reform of Bravo Murillo established for the first time, technical training as a requirement for entering the administration. The mechanism for testing the qualifications of personnel were the ‘oposiciones’ (public competitive exams). In fact, the ‘Administrative Corps’ (following the French tradition) and specifically, the ‘Elite Corps’, had a monopoly on the recruitment process. Through the years a mixture of the spoils system and corporative control were the regular entry mechanisms, sometimes leading to political appointments and, in other cases, to clientela or parentela networks. The first serious attempt to modernize the public employment system was the Basic Law of 1964, but in the end the corps continued to control the whole process, through either concursos (merit system) or oposiciones (competitive exams based on several rote-memory exercises and tests) or a mixture of both. The Law for the Reform of the Public Service (1984) developed a constitutional framework for public administration and introduced an important change from a personal category system to a job system of public employment. It required that the government provide a classification of public positions explicitly identified, defined and valued. At present (2001), the structure of the public service in Spain covers the central administration of the whole country, the regional (AC administration), the local (including provinces, cities and towns) and the institutional (semi-state and partly state-owned companies). Personnel management is organized separately at each of these different levels. Even within each level,

200

The state at work, 1

diverse bureaucratic cultures and practices impinge upon the managerial specifications of each departmental organization. Even with a common legal framework regulating civil service conditions, the salary systems, working conditions and union bargaining systems are specific to each governmental level. The management of central government administration is the responsibility of two departments: the Ministry of Public Administration for all matters related to human resources management and training, including the administrative organization of the civil service, together with its position as the central personnel agency, and the Ministry of Economy and Finance, which prepares the State Budget and has the final word in all matters of public expenditure, including the salaries of civil servants. As a general framework, recruitment policies require that all public service vacancies be advertised. To qualify for a position, candidates must succeed in a public competition concours system, followed by specific training for the job. Once recruited, he or she is integrated into a complex structure based on two broad categories: ‘groups’ and ‘corps’. The groups are defined according to the minimal academic qualification required to enter. There are five groups: group A comprising top managers (full university degree required), group B for intermediate level managers (middle and technical university degree required), group C for employees and clerks (high school education diploma), and groups D and E for minor clerks and workers (primary education certificate). The formation of corps is one of the singularities of the Spanish system, shared with the French case. The number and status of the several administrative corps are the key factor in explaining the actual workings of the Spanish system, including access to civil service positions, economic rewards, potential administrative careers, prestige and even the possibility of being part of the political elite. The nature of administrative corps, historically and even today, has two sides. In one respect, the corps are useful structures for organizing training, recruitment and establishment of administrative careers. In fact, a corps is also a group of civil servants who have been selected according to a specific procedure, have similar professional backgrounds and, in principle, are supposed to fulfil common tasks. At the same time, all these elements constitute the base for the corps becoming powerful pressure groups, with the advantage of acting within the political system and with an inherited tendency to monopolize specific sectors of the administrative organization to the benefit of their members. They display inner group loyalty and become socialized according to the moral code of their own corps so that they can distinguish between ‘their territory’ and a foreign one. The corps are organized according to the level of government (central, regional or local) and also into general and special

Working for the government in Spain

201

corps, depending upon the tasks to be met (general corps are in charge of administrative management; special corps work as engineers, state lawyers, judges, diplomats, finance auditors, notaries and so on). Some of the special corps form the traditional recruitment ground for the administrative and political elite of the country. With each corps representing – or claiming to represent – a specialty, the country has experienced as many types of concours as there are administrative corps. The result is a segmentation of employees, with personnel policies often representing the interests of the particular corps in question. On top of all that, civil service status is also determined according to a scale of 30 levels depending on the position occupied. The 30 levels are grouped, with each grouping representing the threshold for admission to each specific group of civil servants. The public employee in Spain, according to his or her juridical statute, is classified as a permanent civil servant (having succeeded in a competitive exam, entered a corps and filled an advertised vacancy), a temporary one (hired by the administration for a limited period), a permanent labourer or a temporary one (these last two cases are employed under private law conditions). The narrow path for entering into the administration through the corps gates has produced an increase in the number of labourers throughout the public workforce. In central government ministries labourers represent almost 20 per cent. A similar figure is also the case at the regional level, although the ACs vary a great deal depending on the specific region and its level of ‘funcionarizacion’, that is, depending on the percentage of permanent civil servants employed. For instance, the Navarra public administration is made up almost totally of permanent civil servants. At the other extreme, the Madrid civil service comprises 73 per cent labourers. The ACs are heterogeneous in the relative weight of each administrative group within their respective public services. The career system is not fully developed in Spain. Either because job descriptions are not always clearly enough defined or because discretionary appointments can lead to the recruitment of civil servants and labourers to positions unrelated to the formal requirements, examples can easily be found of civil servants appointed to elite positions without former experience or alternatively, highly qualified civil servants relegated to secondary positions.

TOTAL PUBLIC EMPLOYMENT IN SPAIN Public sector employment is at three main levels: state, regional and municipal. State level consists primarily of the central administration,

202

Table 8.1

The state at work, 1

Public employment overview, 1965–2000 Total government employment (000s)

Percentage of population

Percentage of labour force

2208 2116 1993 2013 1920 1801 1738 1781 1760 1783 1704 1480 1213 971 806 726

5.6 5.3 5.1 5.1 4.9 4.4 4.3 4.5 4.5 4.6 4.9 3.8 3.6 2.9 2.1 2.3

13.1 12.9 12.2 12.5 12.0 11.5 11.2 11.6 11.6 11.8 11.3 10.8 9.3 7.4 6.4 6.0

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1975 1970 1965 Percentage change 1995–00 1985–95 1975–85 1965–75 1965–00

22.6 21.7 52.4 33.7 204.1

Source: Encuesta de Población Activa, (1964 to 2000) and Official Census Data. The total government figure does not include employees in the public economic sector. Labour force is defined are those in labour plus those seeking labour.

autonomous bodies (they number more than 200 including ports, water authorities, the Institute of National Industry, state TV and so on), public companies and civil and military bodies. At all three levels (state, AC and local authorities) the personnel comprise career civil servants, temporary employees and labourers. In the year 2000, total government employment stood at 2.2 million (Table 8.1). In the period 1965–2000 total employment increased by 204 per cent, with the highest growth rate during the years 1975–85 (52.4 per cent). Those ten years were critical in terms of the democratic transition and consolidation, the mobilization of the expectations of the citizenry regarding new services and the broadening of the framework of governmental tasks. As Table 8.1 also shows, the increase in total government employment

Working for the government in Spain

203

between 1995–2000 represents a percentage of 22.6 in five years, which is more than the total growth in the previous ten years. This growth was mainly in new public employment developed at the regional and local governmental levels (Table 8.2). In a period of dramatic political devolution and reallocation of tasks between the three tiers of government, the new regional governments took on new tasks and new infusions of money but rejected personnel from the centre, meanwhile opening the door to their own public employees. In the decade 1985–95, public employment at the central level increased by just 2.3 per cent whereas at regional level it increased by 104 per cent and at local level by 41 per cent. During the last five years (1995–2000) central government figures decreased by 16 per cent. During the same period, employment at the regional level grew by 54.3 per cent and local government by 28 per cent. As these figures show, decentralization has not entailed a redistribution of personnel among levels of government but rather a net increase in new public employment at the subnational levels. In 1965 total government employment represented a ratio of 23 public servants per 1000 inhabitants; in the year 2000 it reached a figure of 56 public servants per 1000 inhabitants. While in 1965, 6 per cent of the labour force were public employees, by 2000 the figure had grown to 13 per cent. This doubling of the public sector component of the labour force does not imply that Spain has a large public sector in comparison with other countries. Among the OECD countries, in terms of the number of public employees per 1000 of inhabitants, although Spain doubled its ratio in the period 1970–90, it continues to be one of the lowest among industrialized nations. Similarly to other European countries, Spain has experimented with privatization policies over the last ten years and as can be seen from Table 8.2, the figure for employment in public companies has decreased from 392 900 in 1984 to 216 500 in 2000, representing a decrease of almost 45 per cent. At the time of writing the privatization policies are still in progress and further reductions can be anticipated.

EMPLOYMENT BY LEVEL OF GOVERNMENT Historically the decentralization process in Spain began with the First Republic (1873–76), a federal republic built on endemic political cantonalism. It was followed by the Second Republic (1931–36) period in which three regions were granted and exercised autonomous status (Catalonia, the Basque Country and Galicia). The third period was the Political Transition, the drafting of the new Constitution of 1978 and the democratic consolidation. The

204 28.0 41.0

482.2 466.7 451.4 426.2 402.9 377.0 393.5 395.6 401.2 416.7 389.1 304.6 267.3

Local

17.0 51.5

384.8 339.4 343.9 375.8 357.6 328.9 291.6 311.5 321.0 313.3 307.2 241.6 217.0

Health

34.9 15.4

216.5 211.6 208.9 248.3 324.1 332.5 310.5 338.6 382.5 394.3 417.4 395.9 392.9

Public companies

24.1 91.1

6.7 3.0 3.5 5.7 2.7 5.4 4.4 5.9 2.1 9.0 4.2 40.9 60.4

Other

11.1 21.9

2370.2 2230.7 2201.4 2261.1 2244.4 2133.8 2048.5 2107.7 2142.7 2177.0 2121.6 1875.6 1749.8

Totala

19.6 32.7

2153.7 2019.1 1992.5 2012.8 1920.3 1801.3 1738.0 1769.1 1760.2 1782.5 1704.2 1479.7 1357.0

Totalb public companies

Source: Encuesta de Población Activa, INE, the data refer to the last quarter of each year. The Active Population Survey (EPA) conducted since 1964, is a continuous quarterly survey targetting households. Its main objective is the obtainment of data on labour force and its categories (employed, unemployed and people outside the labour market).

Notes: a. This category refers to a large concept of the public sector, including employees in public companies. b. This one refers to a narrow definition of the public sector. It does not include employees in public companies.

54.3 104.1

Percentage change 1995–00 16.0 1984–95 2.3

Region

801.4 666.7 629.6 594.0 569.7 519.5 510.8 503.1 464.3 480.7 448.5 321.2 254.5

Central

Public employment by level of government (000s), 1984–2000

478.6 543.2 564.1 611.0 587.3 570.4 537.4 553.0 571.7 562.7 555.2 571.2 557.6

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1984

Table 8.2

Working for the government in Spain

205

transitional period attempted to address three main questions: (1) the transition to democracy; (2) a new structure for the state, defined as a non-unitary central state; and (3) modernization of the administrative structure to cope with the new state model. The initial 1978 framework can be characterized by uncertainties and the lack of a clear model among political leaders. In fact, the founding fathers rejected the federal model. Historical experience contributed to mistrust of it. During the constitution-making process, before the approval of the Constitution of 1978, three territories (those which enjoyed autonomous status in the 1930s) were given pre-autonomous status. The Constitution does not establish a set model of territorial organization, but it opens the door for it. This door could lead to an uneven territorial distribution of power or to a homogeneous one. The so-called ‘availability principle’ recognizes the right of autonomy as a constitutional right, accepts the ‘difference principle’, that is, acknowledges the singularities of certain regions, and establishes limits that deny the possibility of granting sovereignty to any of those territories. Following the approval of the Constitution, the main steps were: (1) the first autonomous statutes (1979); (2) the promulgation of the LOFCA (Organic Law establishing the financial mechanisms of the new political entities) in 1980; and (3) the first Autonomous Agreements between the two major parties in 1981, leading to the drawing up of the territorial map and the acceleration of the political devolution process. Between 1983–85, massive competence transfers took place and were completed 11 years later in 1992 with the second Autonomous Agreements, which opened the door to transfers in areas such as education, health and labour and allowed the so-called ‘slow path’ AC (Comunidades de via lenta, including all ACs except Andalucia, Catalonia, Basque Country and Galicia) to enlarge their powers. In public employment terms, the political dynamics resulted in a model that deviates considerably from the one the founding fathers had in mind. In Article 2 of the Constitution they had written of ‘nationalities and regions’ with the notion of political devolution for some territories and simply decentralization for others. The actors in this devolution process were expecting a final map in which for each ten public employees, six would be working at regional level, two at local level and the other two at central level. The outcome to date is quite different, as can be seen from Table 8.3. According to the figures, central government, even with the dramatic decrease in the last ten years, still represents 34.1 per cent of total employment, while regional government, even though it experienced a growth of nearly 70 per cent, represents only 40.9 per cent of total employment and the local level represents 20.6 per cent. Even when the figure corresponding to university employment (4.3 per cent of the total) is added to

206

41.0

Percentage change 1990–2000 1.7

34.1 41.7 42.3 42.2 43.5 45.0 46.2 53.8 55.8 57.0 58.9

Percentage of total

68.3

904.0 690.4 677.1 671.5 637.2 620.5 625.7 600.9 593.5 565.5 537.0

AACC

40.9 32.6 32.3 31.9 30.8 30.3 31.4 26.6 27.3 25.7 24.7

Percentage of total

28.1

455.3 450.7 441.0 460.0 450.4 425.2 371.5 367.0 366.4 374.7 355.5

Local

20.6 21.3 21.1 21.8 21.7 20.8 18.6 16.3 16.9 17.1 16.4

Percentage of total

29.3

95.0 92.9 89.9 86.2 83.8 80.8 77.4 73.5 – – –

Universitiesa

4.3 4.4 4.3 4.1 4.0 3.9 3.3 3.3

Percentage of total

Notes: a. From 1993, year of the generalized transfer of the university education to the CCAA the Boletin disaggregates the University personnel. b. Decline compared to previous year due to obscurity of official statistics.

753.8 882.0 886.0 889.2 899.8 921.1 918.0 1215.2 1213.2 1256.7 1277.8

Central

2208.1 2116.0 2094.0b 2107.0 2071.2 2047.5 1992.6 2256.7 2173.1 2196.8 2170.3

2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990

Total governm. employment

Table 8.3 Public employment by level of government in Spain, 1990–2000 (000s)

Working for the government in Spain

207

the regional public employment figures, the total percentage for this level does not reach the anticipated 60 per cent. The process of transfers from the central level to the regions is however not yet completed and an increase in numbers at the regional level can be anticipated. Before outlining the main characteristics of public employment by level of government, it is necessary to reiterate the difficulties in presenting clear and coherent public employment statistics. Table 8.2 is based upon data from the Survey on the Labour Force (Encuesta de Población Activa) undertaken by the Statistics National Institute (INE). Table 8.3 is drawn from data from the Ministry of Public Administration (Central Registry of Public Administration Employment). These two sources differ not only in their definitions of ‘public administration’ but also classify the several groups of public employees differently, resulting in some inconsistencies in the data. For identical periods the figures for public employment show a difference of about 7 per cent depending on the source. In 1984 the Encuesta de Población Activa registered 41.1 per cent of public employment at the centre, 16.0 per cent in the health care system, 18.7 per cent at the new regional level (AC) and 19.7 per cent in the municipalities. Sixteen years later, in 2000, the number of public employees at the centre decreased to 22.2 per cent, the health care employees increased slightly to almost 18 per cent, the regional autonomous government rose to 31.2 per cent and at the local level the percentage was 22.4 per cent. Employees in the public companies dropped from 28.9 per cent in 1984 to 10 per cent in 2000. Privatization and new management techniques probably explain this decrease. In general terms, the central administration component of total public sector employment decreased by almost 8 per cent while regional level employment increased following the dynamic of the devolution process. If we took into account that block transfers from the centre to the periphery commenced in 1982 and that those transfers implied complete ‘packages’ (competences, budget and also personnel), the logical result would be a decrease in the number of public employees at the centre of the same magnitude as that of the corresponding increase at the regional level. In fact, the ACs increased their numbers of public employees above the transferred number because each of the 17 regional governments created their own personnel systems, their own public employment strategies and their own recruiting mechanisms. The final picture reflects duplication of effort and expense, an overlap of administrative structures and retention of personnel at the centre. The new law for the organization and working mechanisms of state administration (LOFAGE), at implementation stage in 2003, attempts to solve these problems and other conflicts between the centre and periphery administrative structures. Despite the administrative reform

208

The state at work, 1

effort and specific policies addressing the overlapping structures and the socalled ‘single administration’, a clear territorial definition of tasks is still lacking. Statistics for the individual ACs highlight the differences in the composition of their respective public workforces (Table 8.4). On the one hand, they show without exception a decrease over the last ten years in the number of central government employees located in each region. On the other hand, the territorial representation of central level field staff varies between regions. The more consolidated and stronger ACs show a lower percentage of field staff from the centre (see Catalonia, the Basque Country, Galicia, Navarra, Valencia), whereas the other regions still reflect a major presence of centre field staff employees in their territories (for instance, Murcia, Asturias, Extremadura, both Castillas, Cantabria and Aragon). When celebrating the 20th anniversary of the first autonomous agreements, the following aspects should be noted: ●

●

●

●

●

The peripheral nationalisms (Basques, Catalans, Galicians and even Andalusians) have played a strategic role in the building of the model for devolution. Demands for autonomy have been uneven, as also has been the process of the transfer of competences. The model however is not yet completed. We can expect in the future either a homogenizing scenario or alternatively an intensification of the political differences between regions. The early stages of the devolution process saw severe conflicts between the centre and the periphery, specifically concerning the distribution of powers and the drawing of a demarcation line between central and regional tasks. Most of these conflicts had to be resolved before the Constitutional Court, although there has since been a remarkable cooling down in favour of political negotiations. The outcome so far, in organizational terms, is extraordinarily complex, leading to a type of asymmetrical federalism which makes the Spanish case different from the typical federal model in which each of the partners to the federation enjoys the same level of competences. The political parties, at both the central and regional levels, are additional strategic actors in the process. The two main nationwide parties have a role as consensus builders and crafters of the politicoadministrative architecture; the regional parties, particularly the Basques and the Catalans, draw on their monopoly over the regional institutions and their potential for coalition in the central government; and

209

56.5 12.0 45.8 13.7 33.7 44.9 58.5 10.3 40.3 56.0 53.7 100.0

12.2 47.9 8.5 57.9 10.9 9.6 47.6 50.9

2000

13.5 56.3 16.5 61.8 8.4 20.1 13.5 56.5 13.9 – – –

59.2 13.0 80.2 8.7 55.2 39.0 13.4 13.7

1991

16.5 54.8 16.8 60.4 36.3 27.0 12.9 61.0 18.8 13.6 16.7 –

56.2 19.9 73.3 11.5 54.8 58.2 18.0 13.6

2000

AACC

20.0 26.0 23.0 17.2 30.1 17.7 21.7 29.3 15.2 – – –

22.3 19.0 5.7 22.0 26.5 29.2 15.2 23.6

1991

Local

21.0 29.7 34.0 20.7 26.6 22.7 23.0 25.0 35.0 29.4 27.9 –

26.6 24.4 14.3 21.9 28.8 27.3 26.5 33.1

2000

1991

6.0 3.4 3.3 5.1 3.3 5.4 5.6 3.6 5.9 0.9 1.6 –

4.9 7.7 3.8 8.4 5.3 4.9 7.7 2.3

2000

Universities

114 569 177 075 49 176 90 442 26 578 11 222 46 288 75 890 39 383 – – –

285 109 53 894 17 867 298 977 131 023 66 479 21 516 69 045

1991

Total

128 517 197 485 64 642 114 480 29 944 9 128 48 870 81 065 40 089 5 347 5 395 6 925

334 907 46 624 24 583 253 345 166 389 87 788 18 797 89 799

2000

Source: Boletin Estadístico del Registro Central de Personal. Public economic sector, police, army, justice and autonomous public bodies not included.

66.5 17.7 60.5 20.9 61.5 62.1 64.8 14.1 70.8 – – –

18.4 68.0 14.2 69.3 18.3 31.8 71.4 62.6

1991

Centre

Territorial distribution of public employees, 1991–2000 (%)

Andalucía Aragón Navarra Madrid Valencia Canarias Cantabria Castilla la Mancha Castilla-León Cataluña Extremadura Galicia I. Baleares La Rioja P. Asturias Pais Vasco R. Murcia Ceuta Melilla Abroad

Regions

Table 8.4

210 ●

The state at work, 1

The challenges for the near future will be to do with financial relations, the political devolution process from the autonomies to local government, the role of the Senate as a potential federal chamber and the participation of the ACs in EU decision-making processes.

BUREAUCRATIC ORGANIZATION AND PERSONNEL AT SUB-NATIONAL LEVEL At present (2001), the process of devolution is almost completed, the new structures have been consolidated, the 17 new regional governments enjoy a full range of competences and powers (legislative, executive, judicial). They have approved more than 2000 new laws, they control around 35 per cent of total public expenditure and they employ almost 46 per cent of the total public workforce. The perception of this new reality by the Spanish public is consistently very positive, but the outcome of the process in organizational and personnel terms shows an inert and imitative model. The new political regional leaders decided to avoid the uncertainty and risk of innovation and instrumentally copied the structure of the central state administration with all its main features, even the well-known failures. One feature of the regional administrative structures for instance is the inflation of head units (‘jefaturas’) in comparison with the regular public employees in each region. According to data from Fundacion Encuentro, in Andalusia there are three public employees per head unit, ten in Catalonia, 11 in Galicia and 16 in Madrid, among other examples. The over-abundance at the top levels of the administrative structure is closely related to political strategies and interest group pressures and not linked at all to the rationality of the dimension and size of the administrative structure. Nevertheless, the number of regional public employees (excluding the top managers and those working in the policy areas of education and health) is relatively proportional to the size, the organizational complexity and the level of competences of each regional government. The three ACs with the highest public employment rates are Andalusia, Catalonia and Valencia, followed by Galicia, the Basque Country, the Canary Islands and Madrid. The ratio between population and public employees varies between 31 and 47 inhabitants per public employee in these regions. Other factors explaining the differences found among the ACs relate to differences in their economic situations, personnel management models and degree of public/private mix in the delivery of services. The composition of public employment at the regional level comprises a mix of career civil servants and labourers. The ACs vary in terms of the proportion of public employees under labour contract versus those under

Working for the government in Spain

211

administrative law and civil service statute. Females and the young are more likely to be labourers and temporary employees. Probably due to the lack of a model, the fragmented and uneven devolutionary process and the absence of clear direction, the administrative organization of the ACs does not embody the principles of organizational rationality. There are frequent instances of employees with different qualifications doing the same job or public employees overqualified for the positions they hold. What is clear is that the policy of decentralization and the block transfers have stimulated growth in public employment in the new regional political arenas. Additional explanatory factors for the growth in the number of public employees include the introduction of the welfare state, especially the education and health services, and concerns about service quality, together with a policy of public service recruitment as an answer to the relative scarcity of jobs in the private sector.

LOCAL GOVERNMENT EMPLOYMENT Spain has a heterogeneous and highly fragmented local government system. There are over 8000 municipalities, almost 75 per cent of which have fewer than 2000 inhabitants, 50 provinces and around 800 ‘mancomunidades’, that is, special-purpose associations of several municipalities established in order to avoid problems of scale and focused on shared interests. Under the new democratic arrangements, local governments were challenged by new and stronger demands and were also legally obligated to deliver new and more services. The dramatic growth in local-level public employment is a clear indicator of the new framework. There was an increment of 69 per cent in the years 1984–2000 when the number of employees expanded from 267 300 to 482 200 (Table 8.2). Local-level public employees are organized around three main categories: the career civil servants belonging to state corps, representing around 1 per cent of the total; local civil servants, recruited in each municipality and making up almost 50 per cent of the total; and labourers who comprise the remaining 49 per cent. The state career civil servants are the leaders and the elite of the administrative structure. Secretary, fiscal controller and accountant officer are the three main state corps groups at this level. Mainly the figures reflect the phenomenon of ‘labouralization’ of public employment at the local level. Labourers and temporary employees make up a large percentage of public employees at the local level. It is an outcome of the lack of dynamism of the administrative corps and their control of the recruitment of newcomers. Labourers enter the administration ‘through the back

212

The state at work, 1

door’ and after several years become either permanent civil servants (‘funcionarización’) or permanent labourers. The main disadvantage for the latter group is that they have no career path into an administrative career, remaining at the same level (excluding promotion) all their life. Today almost 50 per cent of all labourers work at the local level, 26 per cent at the centre and 24 per cent at the regional level. Although the devolution process has produced vast changes at the regional level of government, this has not extended with the same intensity to the local level. During the 1990s, local authorities developed through their national association FEMP (Federacion Espanola de Municipios y Provincias) a well-articulated group of demands known as ‘Pacto Local’ (local covenant) which was only partially addressed by the central government in 1999. Currently they are lobbying nationally and regionally for a second devolutionary wave as a mechanism to demand ‘local autonomy’ and to avoid the new centralism embodied in the ACs.

PUBLIC EMPLOYMENT BY POLICY AREA AT THE CENTRAL GOVERNMENT LEVEL In order to understand the distribution of personnel and public expenditure across the different policy areas and its evolution, we need to consider the specifics of the welfare state building process in Spain. Between 1960 and 1975 when most EU countries were developing pension schemes, health and housing programmes, unemployment benefits and education services, government expenditures comprised only a minimal part of the Spanish economy. At the end of that period the average total public sector spending for the European Communities countries represented 40 per cent of GDP; the Spanish equivalent was only 25 per cent. In the following two decades, the country underwent an accelerated process to reach European standards, with an inevitable impact on the public employment figures. As a consequence, the distribution of personnel according to governmental tasks has changed dramatically. In this relatively short period, Spain has not only developed a welfare state but has also launched an extensive decentralization process that has transferred personnel and economic resources to the regional and local government levels, most particularly in the policy areas related to social services, education and research, and health. The available data regarding the percentage of personnel and public expenditure invested in specific policy areas allow only tentative and fragmentary analysis, limited to the central government level. There is no ready information about the distribution of staff across policy areas at the subnational levels of government. The data discussed below are drawn from

Working for the government in Spain

213

two main sources, the Ministry of Public Administration and the annual budget papers produced by the Ministry of Finance. The lack of correspondence between the figures recorded in the separate tables arises from differences in definition and methodology. The evident inconsistencies arise also from the difficulties associated with any attempt to trace personnel movements during a period of extensive reorganization. One particular difficulty is that the data relate not only to the period of dramatic change arising out of the devolution process but also include a time when legislative changes affected the classification of personnel and even the type of juridical linkage. Tables 8.5–8.7 present data on personnel numbers and expenditure categorized according to policy area and drawn from the Ministry of Public Administration (Tables 8.5 and 8.6) and the annual budget papers prepared by the Ministry of Finance (Table 8.7). Again the data are for central government only. As noted above, the apparent lack of consistency arises from different data-collection methods and attests at least in part to the problem of tracking employment data during periods of considerable change between and within portfolios. Despite the broad scope of the changes accompanying devolution, the structure of central government shows an extraordinary historical continuity. As a result of political bargaining or party agreements, frequent minor changes have occurred nevertheless in the number and denominations of departments, the number of portfolios and the policy areas covered. In 1999 there were 13 departments (Table 8.5). The Ministry of ‘Fomento’, the 19th century name for the infrastructure and transport policy area, was known as the Ministry of Public Works under the socialist government. Also, among numerous other minor reshuffles, the Ministries of Justice and Interior have been alternately merged and separated. Despite the continuity in the structure and denominations, as a result of devolution the total number of employees working in the central ministerial departments decreased by 22.2 per cent in the ten years 1990–99. As expected, the 1999 figures show that the highest percentage of public employment was in the education area. The decrease since 1990 reflects the devolution of university and non-university services to regional governments. The second largest department was Public Works (including Postal Service and Communications) with almost 19 per cent. Another salient ministry was Labour Affairs and Social Security with 13.4 per cent. Agriculture experienced a notable decrease, brought about by the transfer of previous state powers to the regional level. Although Table 8.6 uses a different policy area classification, it shows the same evolution. In 1994 the policy areas showing the highest employment levels were health and social services (15.2 per cent), education

214

Table 8.5

The state at work, 1

Personnel in central government ministries, 1990–99 1990 N

1995 %

N

Presidency 3 247 0.63 3 868 Public Administration 3 060 0.60 2 579 Agriculture, Fisheries and Food 16 341 3.17 11 009 Foreign Affairs 5 349 1.04 5 550 8 376 1.63 7 268 Defencea Social Affairs 8 785 1.71 11 498 Trade and c 2 857 Tourism Economy and Finance 47 327 9.19 47 131 Education and Science 220 552 42.81 166 374 Industry and Energy 3 296 0.64 3 517 Justice and 27 918 5.42 31 930 Interiord Public Works, Transport and Environment 23 017f 4.47 75 857 Culture 4 857 0.94 4 857 Labour Affairs Social Security 52 282 10.15 58 544 Total 515 131 100 441 164

1997 %

N

1999 %

N

%

0.87

3 759

0.87

3 676

0.90

0.58

2 490

0.58

9 102

2.30

2.04 1.28 1.64 2.62

5 561 5 616 6 848

1.29 1.30 1.59

4 266 5 743 6 457

1.00 1.40 1.60

b

0.65 10.80

49 776 11.53 48 901 12.20

38.26 177 608 41.15 153 534 38.30 0.79 2 903 7.50 32 621e

0.67 7.55

2 672 1 956 28 530

0.60 0.50 7.10

64 621 16.10 17.40 66 272g 14.43 11 319 2.80 h – 0.94 5 760 11.11 12.85 58 837 13.86 53 879 13.40 100 431 563 100 400 731 100

Notes: a. Military personnel and labourers in the Defence department not included. b. This Ministry disappeared in the Popular Party government structure. c. The Ministry in this year was Transport, Tourism and Communication with a total of 78 842 employees. d. Police and security personnel and employees in the Justice administration not included. e. From this year on Justice and Interior are again two different Departments with 1876 and 30 745 public employees respectively. f. In this year transport, communication and environment was the competence of another ministry with a number of public employees of 78 842 and almost 50 000 of those working in the postal service. g. In 1996 a new Ministry of the Environment is created for the first time in Spain. h. From this year comes under the Ministry of Education and Culture. Source: See Table 8.6. Apparent inconsistencies arise from differences in definition and methodologies over the time period.

215

Working for the government in Spain

Table 8.6

Central government employment by policy area, 1994–2000 1994

Education Social security Prisons Posts and telegraphs National patrimony Fiscal state agency Other employees at the central ministries Military civilian employees Social and health services Police Military Justice Other public bodies Total

1996

1999

2000

N

%

N

%

N

%

N

%

131.6 46.1 14.2

14.4 5.0 1.5

40.9 44.3 15.4

15.3 4.8 1.7

136.0 36.3 15.9

15.3 4.1 1.8

8.4* 34.8 16.1

1.1 4.6 2.1

52.1

5.7

54.1

5.9

55.5

6.2

53.6

7.1

1.7

0.2

1.5

0.2

1.5

0.2

1.5

0.2

26.3

2.8

27.6

3.0

27.4

3.1

27.6

3.6

163.0

17.8

154.4

16.8

144.5

16.3

122.6

16.3

36.7

4.0

37.2

4.0

35.0

3.9

33.2

4.4

139.1 120.2 71.4 40.2

15.2 13.2 7.8 4.4

132.3 125.1 82.2 40.1

14.4 13.6 9.0 4.3

135.7 121.9 96.7 23.8

15.2 13.7 10.9 2.7

136.1 120.9 115.8 25.1

18.5 16.0 15.4 3.3

73.0

8.0

65.9

7.1

58.0

6.5

57.9

7.7

915.6

921.0

888.2

753.6

Note: * Once the devolution process has been concluded this figure only will refer to public employees in Ceuta, Melilla, personnel abroad and teachers of religion. Source: Boletín Estadístico del Registro Central de Personal, Ministry of Public Administration.

(14.4 per cent) and police (13.2 per cent). Six years later two main phenomena emerge. On the one hand, the devolution of tasks to the regional levels has diminished the percentage of public employment concentrated at the central level. The regional governments have now become the big expenders of money and consumers of personnel. For example in 1998, education and culture represented 21.6 per cent of central government personnel expenditure; in 2001 the figure was only 2.6 per cent (Table 8.7). At the same time other policy areas such as those related to law and order functions have been reinforced. As in other countries, a high concentration of personnel is

216

Table 8.7

The state at work, 1

Personnel expenditure by government tasks, 1998–2001 Budget 1998 Million pesetas

Parliament High courts and advisory bodies Welfare and social services Foreign affairs Legal expenses Defence Economy and finance Interior and police Infrastructure and housing Education and culture Industry and energy Agriculture Environment Health Other ministries Total

%

Budget 2001 Million pesetas

%

Percentage change

11 600

0.4

13 431

0.5

15.8

9 910 837 075 30 977 107 949 515 935 59 844 548 536 38 026 662 632 4 486 10 697 15 355 5 351 183 969

0.3 27.1 1.0 3.5 16.8 2.0 17.9 1.2 21.6 0.1 0.3 0.5 0.2 6.0

12 941 939 858 35 758 134 933 588 632 63 441 579 332 34 688 71 647 9 007 10 636 16 380 5 238 218 223

0.5 33.9 1.3 4.9 21.3 2.3 20.9 1.3 2.6 0.3 0.4 0.6 0.2 7.9

30.6 12.3 15.4 25.0 14.1 6.0 5.6 8.8 89.1 100.7 0.6 6.7 2.1 18.6

3 068 178

100.0

2 769 215

100.0

9.7

Source: Ministry of Finance, Year Budget. Apparent inconsistencies arise from differences in definitions and methodologies over the time period.

not necessarily related to specific policy area expenditures. The traditional policy areas such as justice, defence, foreign policy and police, for instance, account for only about 7 per cent of total public expenditure. Welfare and social services, although much less intense in terms of the number of personnel, is nevertheless the largest item of public expenditure.

THE ADMINISTRATIVE ELITE At a general level, the administrative elite can be identified as those civil servants with a university degree. More specifically the administrative elite is confined to those who hold the top managerial posts within the organization. In the ministerial departments the highest positions are organized, in ranking order, as follows: State Secretary (the first in rank below the Minister), General Director (responsible mainly for the ministry policy areas) and Deputy General Director. The last mentioned is the only one

Working for the government in Spain

217

who must be formally a civil servant, so required by law. In addition, each ministry has an under-secretary, responsible mainly for personnel and resources management within the ministry. At present (2001) the total number of top officials in the central governmental structure is 819 in number, comprising 605 deputy general directors, 176 general directors, 17 under secretaries and 21 state secretaries. The practice for more than a century has been for this elite group to be drawn from only ten of the more than 200 civil servant corps, including, in particular, technical civil administrators, diplomats, university professors, state lawyers, engineers and finance auditors. They represent a universe of around 3000 people who function as the recruiting ground of the elite. There is no specific institution for training an administrative elite similar to the French ENA. From an analysis of social backgrounds, the characteristic common to the elite is a university law degree or more recently the new economics and management degrees. Traditionally civil servants have held a prominent place in the composition of the Spanish administrative and political elite. Between 1975 and 1982 civil servants occupied more that 62 per cent of all political-administrative positions. The democratic reforms have produced new forms of politicization of the administrative elite, based on a mix of administrative corps membership and political party affiliation. At central government level, and even more so at the regional level, the administrative structure itself has been politicized by the increase in the number of posts available for discretionary promotions and political appointments. Historically, Spain has experienced the instabilities derived from spoils system practices. The original intention behind the establishment of the administrative corps was to maintain administrative stability in face of frequent political change. Despite formal acceptance of specified civil service conditions, a type of spoils system continues to be practised, with appointments based upon considerations (political or quasi political) unrelated to administrative career criteria. More recently further politicization of the administrative elite has occurred with the development of the ‘political adviser’ category within each administrative department or ministry, including, most particularly, the regional level.

PUBLIC EMPLOYMENT BY GENDER In Spain, women have been latecomers to the labour market. One of the major changes has been the feminization of public employment, especially in the health, education, social services and welfare state areas.

218

The state at work, 1

Table 8.8 Gender and public employment at the central government level, 1989–98

1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 Source:

Male %

Female %

51.1 51.3 49.1 49.4 49.9 50.4 59.6 59.0 60.6 67.1

48.9 48.7 50.9 50.6 50.1 49.6 40.4 41.0 39.4 32.9

Boletín Estadístico del Registro Central de Personal, 1989–99.

In 1989, women represented only one-third of total central level employment. By 1998 almost half of the public employment positions were occupied by females (Table 8.8). The available data on public employment at the regional level show, again, considerable variation across the ACs. In most cases there is a fairly even gender balance, with a 2 to 7 per cent predominance of males. The extremes are represented by, at one end of the scale, the male dominated ACs of both Castillas, the Basque Country and Baleares where the difference is about 15 per cent and, at the other extreme, the female-dominated ACs of Valencia and the Canary Islands where women represent 55 per cent of total public employment. Closer analysis of the data reveals a concentration of women in the areas of education, health and social services, at both the central and regional levels. This feminization has however not been evenly distributed across administrative career levels. Women continue to be poorly represented in top positions (Table 8.9). Their participation in government employment increases lower down the administrative hierarchy.

CONCLUSION The problems noted above concerning the lack of available data clearly make a proper historical analysis difficult. The fragmentation of the

219

Working for the government in Spain

Table 8.9 Public employees by gender and administrative career level at the central government level* Female

Male

Total

N

%

N

%

N

30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 Other

364 344 1 453 960 3 900 202 5 505 345 4 527 548 2 522 122 9 633 1 107 8 269 6 067 15 354 2 106 15 816 6 883 231 – 3

20.7 19.2 27.7 33.5 40.5 22.3 46.6 35.2 44.0 36.3 44.7 10.2 56.1 21.5 63.0 34.8 64.4 71.6 58.1 23.3 18.8 0 2.9

1 394 1 443 3 791 1 905 5 738 705 6 297 636 5 750 962 3 117 1 079 7 544 4 045 4 856 11 377 8 470 834 11 394 22 672 1 000 6 102

79.3 80.8 72.3 66.5 59.5 77.7 53.4 64.8 56.0 63.7 55.3 89.8 43.9 78.5 37.0 65.2 35.6 28.7 41.9 76.7 81.2 100.0 97.1

758 1 787 5 244 2 865 9 638 907 11 802 981 10 277 1 510 5 639 1 201 17 177 5 152 13 125 17 444 23 824 2 940 27 210 29 555 1 231 6 105

Total

86 261

45.1

105 117

54.9

191 378

Note: * Administrative level is related to the educational degree. University education is required for levels 20 to 30; middle and technical university degree is required for levels 16 to 26; high school diploma for levels 11 to 22; and primary school for the lower levels. By gender, females represent 31.8% of the university degree civil servants, 43.3% of the middle and technical degrees; 39.4% of the high school level and 52.1% of the civil servants with just a elementary school level. Source:

Boletín Estadístico del Registro Central de Personal, 2000.

information and the way it has been gathered make the task complex. The arrival of democracy in Spain and the modernizing policies of the several governments, together with the requirements attached to EU membership, are leading towards a better quality of statistical information, but the fragmented structure of the institutions involved has

220

The state at work, 1

reproduced the traditional problems of data inconsistency. Even more, the transfer of competences to regional governments has added further complications to statistical data collection and hampered easy access to consistent information on public employment by level of government and policy area. The inauguration of democracy in Spain produced a rapid increase in expectations and demands for new and better services and a strengthening of the welfare state. In terms of public employment, it resulted in large increases, mainly in the areas of education, health and social security, sanitation, public works and revenue services. Democracy in Spain is also closely related to the far-reaching process of political decentralization. This factor, together with full membership in the EU, is contributing to the modernization and homogenization of the administration and its personnel to conform to European standards. From the information available, one clear development has been the large increase in the size of the public sector. In the years 1970–99 public expenditure as a percentage of GDP increased from 23 to almost 50 per cent. Spain moved from having a small public sector with heavy public regulation of the market to a spectacular growth in public administration together with a progressive deregulation of economic activity within the private sector. In the process, the percentage of public spending by level of government has changed dramatically, moving from 92 per cent at the centre in 1970 to around 50 per cent today, with the regional and local government levels responsible for 30 and 15 per cent respectively. To a certain extent, public sector personnel policies remain centralized, for instance, in respect to salary and working conditions and the role played by the Ministries of Public Administration and Finance on one side and the trade unions on the other. Personnel management however is also in the hands of each ministry, each of the ACs and each local authority. Both the lack of information and the weak and inefficient coordination mechanisms deter a serious approach to the main challenges of public employment management in modern Spain. Questions such as: the resizing of the central administration, the recruitment of managers and ‘cadres’, the drawing of clear boundaries between politicians and civil servants, the change of the administrative culture of public employees and, among other matters, the training in new techniques related to intergovernmental and inter-administrative coordination, should be addressed urgently. For this purpose, there needs to be a thorough evaluation of the structure of public employment at all levels of government, including in the several government networks, and a clear analysis of what they do.

Working for the government in Spain

221

SELECT BIBLIOGRAPHY Baena, Mariano (1984), Estructura de la Función Pública y Burocracia en España, Oñati: Instituto Vasco de Administración Pública. Beltran Villalva, M. (1964), ‘Datos para el estudio de los funcionarios públicos en España’, Documentacion Administrativa, 83, 9ff. Fundacion Encuentro (1996), Informe Espana, Madrid: Fundacion Encuentro. INE (Instituto Nacional de Estadistica) (1964–2000), Encuesta de Poblacion Activa, Madrid: Instituto Nacional de Estadistica. Ministerio de Administraciones Públicas (MAP) (1989–2000), Boletín Estadístico del Registro Central de Personal, Madrid: MAP. Ministerio de Economia y Hacienda (1998), Annual Budget, Madrid: Ministerio de Economia y Hacienda. Ministerio de Economia y Hacienda (2001), Annual Budget, Madrid: Ministerio de Economia y Hacienda. OECD (1994), Sources statistiques sur l’emploi dans le secteur public, Paris: OECD.

9.

The French paradox: a huge but fragmented public service Luc Rouban

The weight and growth of public employment in France have always been highly debated questions. Controversies have been raised since the time of the monarchy as well as during the first years of the Republic and have appeared sporadically since then on the question of whether the state apparatus is too heavy. During the Revolution years, Saint-Just, in a famous speech, argued that major problems of the new government could be resolved if it could get rid of an army of ‘twenty thousand bureaucratic idiots’ (Rouban 1998a). Later, the development of liberal ideas between 1880 and 1900 gave way to systematic criticisms denouncing the ‘invasion of civil servants’. In fact, the question of government employment has been always related more broadly to the social structures and the role of the state. The economic dimension of the problem has been connected not only with efficiency or management questions but also with the fact that the state, as a major job provider, could be the best friend of the new middle class. Since the 1970s, all opinion polls show clearly two major findings: the first one is that French opinion does not consider that there are too many civil servants; the other one is that up to 75 per cent of the interviewed persons would encourage their children to become civil servants (Rouban 1998b). No controversy occurred about the level of government employment during the welfare state golden years, that is between 1950 and 1975. A more systematic debate, raising the question of the global cost of the civil service, has been occurring since the early 1980s. Two questions present themselves. The first is related to the budget deficit level as constrained by the Maastricht criteria on economic convergence between the member states in order to launch the single currency. The other is related to the demographic trends jeopardizing the structure of pension plans, as the ratio between active and retired civil servants is likely to decline from 3 to 1, in 1998, to 2 to 1 in 2010, thus producing a huge deficit in the state budget. The French model of the civil service is clearly at a turning point and most observers think that a permanent expansion of the public service 222

The French public service paradox

223

is unaffordable. Economic arguments are generally associated with political ones through a systematic contrasting of civil servants, who are said to enjoy unacceptable ‘privileges’, and private-sector wage earners, who are forced to comply with market law. But this kind of connection between the two arguments has been observable since the early 19th century. The budgetary and political weight of the public service are at the heart of any debate on the administrative reform in France. On this point, it seems that there is no European or, more largely, western convergence (Kickert 1997). The data collected hereafter come from official statistical publications (Direction Générale de la Fonction Publique, annual; INSEE 1992). The real point is to ascertain the boundaries of government employment in France. Strictly speaking, government employment encompasses the three civil services (state, local and health civil service) including statutory and non-statutory personnel. This group could be enlarged to include the civil servants working in state or local quangos or agencies (administrative, scientific or cultural établissements publics such as the Office National des Forêts or the Centre National de la Recherche Scientifique). Government employment is generally regarded as excluding those personnel working in nationalized banks, insurance or industrial companies, who are not civil servants even if most of them enjoy a specific public status. This definition is not just a legal one. It is also based upon political criteria as most companies nationalized in 1945 or in 1981 have been partially or totally privatized during the 1990s. Including or excluding these personnels may change the budget cost significally or the economic impact of the ‘public employment’ (it is one of the criticisms that one may address to the OECD statistic package). These blurred frontiers between the ‘public’ and the ‘private’ sector, as well as the existence of a twilight zone between the two, are, of course, at the centre of many controversies as to whether one has to include in the calculus the personnel working under private law contracts in public agencies (this is the case, for instance, of the social security agencies) or the civil servants working in private companies as they may enjoy from their legal status the opportunity to leave the active state service and to come back later. The identification of the public sphere is uneasy even for the civil service ministry which has created in 2001 an agency dedicated to the overseeing of the public service in order to have a clear view not only on the data but also on the statistical frame used (Observatoire de l’emploi public 2001). I have indicated with each table the basic data used, as official statistics do not always rely on the same kind of information. Another point to be raised is that data prior to 1980 are rather difficult to obtain. Systematic statistical overviews are not always available before this year and thus do not fit in homogeneous statistical data packages.

224

The state at work, 1

TOTAL PUBLIC EMPLOYMENT IN FRANCE Between the end of the Second World War and the 1990s, government employment in France grew from about 2 million people to 5 million people. This is a 150 per cent increase, comparable to the one which occurred in the United States at the same time. As in other western countries, the rate of growth is not linear. While the rate was steady between 1950 and the end of the 1960s, an important increase occurred after 1969: between 1969 and 1989, 1.9 million public employees were added. A careful study of these 20 years reveals that there are two peaks in the increase, 1969 and 1981, followed with periods of stagnation or decline after 1974, with the first oil shock and the start of the economic crisis, and after 1983, when the Socialist government of Laurent Fabius decided to change the global economic orientation of the government. Here, one has to be rather careful with data, as actual employment does not reflect necessarily the public payrolls which lost 40 000 positions between 1983 and 1989. As a matter of fact, the 1970s decade added 30 per cent to the number of government employees while the next decade added only 10 per cent. After 1990 the rapid expansion stopped. This is due, of course, to economic considerations as well as to structural changes. For instance, a new defence policy had been initiated, implying reduction-in-force and professionalization of armed forces. A statistical reduction of about 500 000 employees occurred in 1991 when the postal service and France-Télécom were transformed into quangos (since 1997, France-Télécom has been a private company with a state participation but former employees have kept their civil service status). The first lesson that could be drawn from these global data is that there is no mechanical connection between the development or regression of the welfare state and the size of government employment. A huge proportion of French wage-earners still worked for the public sector at the end of the 1990s even though most of the former public companies and banks have been privatized and while market values have spread. The second point is that the various governments since 1983 have paid careful attention to the budget weight of the civil service which accounts for 40 per cent of the state budget. Since 1982, annual increases have been negotiated with the labour unions on a global calculus basis (‘masse salariale’, i.e. total civil service expenditures) which includes the financial impact of promotions and seniority as well as the salary pay increase itself. This point has been greatly disputed by the unions who preferred to negotiate only on the basis of net salary increases. Nevertheless, this economic dimension is always balanced by political considerations as the governments, from the Left or the Right, have faced systematic demands for more

The French public service paradox

225

public jobs coming not only from labour unions but also from political leaders (especially at the local level) and from French society as a whole. Civil service policy is a very sensitive political question in a country where the state has had an historical role in the pacification of civil society. This tension between economic constraints and political obligation has produced a permanent effort to create new categories of public jobs, blurring legal provisions or statistical data. For instance, in the 1990s, a new kind of five-year contractual jobs appeared, based on a financial arrangement between the state, private business and associations in order to hire young unemployed people. At the end of the 1990s they were approximatively 130 000 in number, but they do not appear in official statistics on a systematic basis and nobody knows whether these jobs will all be integrated into the public service regular structure when the contracts end. This is the French paradox: at first view, in the historical landscape, the public service is a rationalistic machinery, dedicated to public law; but from a socio-political point of view, it is a complex set of layers, each of them having its own rationale and obeying its own law. Public Employment Relative to Population and Labour Force As a percentage of the labour force, government employment is fairly comparable with that of Germany or the United States. Of course, a serious difference appears when data of total public employment are used. The difference lies in the fact that in the 1980s and the 1990s there are still many public companies which have been privatized in successive steps with the Right governments of 1986–88 and 1993–95 but also under the Socialist government of Lionel Jospin after 1997. The latter case was a real departure from the strategy adopted by the previous Socialist governments of the post-1981 era, which had used the public companies and hence public employment as a tool in order to fight unemployment and control major industrial or financial policies. The proportion of public employment in the whole labour force indicates that the public sector provides more than a fifth of the jobs. As a whole, public employment as shown in Table 9.1 includes government employment, traditional public companies such as the SNCF for the railways or EDF for electricity supply (which had been nationalized before or just after the Second World War) and companies nationalized between 1981 and 1983. Of course one may wonder whether comparisons including civil servants, working in ministries with purely administrative tasks, and public employees, working in business activities as do their private counterparts, are really wise. They do not share the same careers, nor the same political and cultural values (Rouban 1998c), even if they are likely to share a common standpoint as soon as the public sector is

226

The state at work, 1

Table 9.1

1998 1995 1991 1988 1985 1983 1980 1969 1962 1948

Public employment overview 1948–98

Total govt. employmenta (000s)

Total public employmentb (000s)

Percentage of govt. employment in labour force

Percentage of public employment in labour force

5236 5099 4991 4857 4827 4745 4442 3046 2500 1950

6482 6604 6754 6212 6683 6597 5530 4016 3608 3102

20.4 20.3 19.8 20.0 20.0 19.9 18.7 14.1 12.4 9.8

25.3 26.3 26.8 25.6 27.7 27.7 23.3 18.6 18.2 15.6

Notes: a. State civil service (including military personnel), local civil service and health civil service. These data exclude agents working in public agencies but include the postal service and the telecommunication service (France-Télécom). b. Including the railways, electricity and urban transport public companies, as well as public agencies whose agents are not regarded as participating in the state civil service and nationalized companies in the industrial or financial sector. Source: INSEE 1994 and 2000.

jeopardized by governmental reforms. Another question could be raised about the very notion of ‘public’ as there is a huge difference between a state civil servant enjoying a specific legal status and a wage-earner working as any other private business employee in a newly nationalized company which becomes ‘public’ just because the state controls a majority of shares. As a matter of fact, the proportion of public employment within the labour force in France at the end of the 1990s was one of the most important among the western countries. As a result, public employees may be regarded as a real political force as no major reform aiming at changing the internal equilibrium of the civil service may be launched without their consent and, sometimes, their support. This political and economic power was experienced by the Juppé government in 1995 when it failed to reform the public sector retirement system and was confronted with the most important street demonstrations since 1947. Again, in 1999, strikes in the education sector, whose unions rejected a professional reform plan, have pressed Socialist Prime Minister Jospin to fire the minister. Even after the Right landslide victory in 2002, the newly appointed government has demonstrated a special care to avoid any frontal conflict with the civil service unions.

227

The French public service paradox

Part-time Employment As in Germany or in the United States (Derlien, Peters, this volume), there is a growing share of part-time employment. Part-time employment for the state civil service increased rapidly during the early 1980s. In the period 1982–98 it increased from 2.8 per cent to 9.2 per cent (Table 9.2). Three major factors can explain this trend. The first one, as in Germany, is due to the fact that the governments have used systematically part-time employment as a strategy to offer more jobs in the civil service especially to the unemployed. The second one is related to the fact that there is a growing proportion of women within the ranks of the civil service who are looking to balance their professional and family duties. Laws and regulations have been enacted during the 1980s in order to ease professional life and allow public employees to work on a 50 per cent or 80 per cent time basis. Thus, more than 80 per cent of those employees benefiting from part-time work are women. The third factor is related to the productivity increase due to computerization of routine tasks, especially in social ministries. So, any serious discussion about the growing weight of government employment must take into account the difference between actual positions Table 9.2

Part-time employment 1982–98 Statea

1998 1995 1994 1993 1992 1991 1990 1989 1985 1982

Localb

Total (000s)

Part-time (%)

Total (000s)

Part-time (%)

1670 1626 1594 1589 1821c 1979 1970 1963 1937 1811

9.2 8.3 8.2 8.0 6.9 7.5 7.3 7.1 6.3 2.8

1507 1462 1428 1406 1383 1365 1326 1237 1185 1103

30.0 30.6 – 30.3 – – – 28.8 – 26.4

Notes: a. Civilian personnel and statutory civil servants only. b. These data do not include specific jobs dedicated to graduate unemployed young people who have been hired on the basis of a contractual arrangement within local government services. In 1998, their number was about 150 000. c. Since 1991, this figure does not include civil servants working for France-Telecom and the Postal Service which have been transformed into quangos. Source: Direction Générale de la Fonction Publique, annual.

228

The state at work, 1

and budget positions, as one budget position can be used to pay two halftime actual positions. Since 1983 public administrations are asked regularly to ‘freeze’ a part of their personnel allowances as soon as there is a vacancy (that is when the position is no longer occupied). The proportion was of one ‘frozen’ vacancy for three between 1983 and 1986, one for two between 1986 and 1987 during the conservative government of Jacques Chirac, one for four after 1988 and again one for two since 1992. As a whole, between 1991 and 1998, the number of budget positions within the state civil service (statutory, contractuals, state workers and military) has proved to be very steady with an increase of 0.09 per cent. A freeze of the global number of budget positions was decided by the government of Lionel Jospin in 1997. After many public demonstrations and protestations coming from the labour unions (especially in the education sector), it was decided to add 17 000 more budget positions in 2000. In 2002 unions carefully scrutinized the rather symbolic decision of the newly appointed education minister to cut 1700 budget positions within the ranks of the clerical staff, as a consequence of budget savings decided by the Right government. These periodic freezes encourage recourse to part-time employment.

EMPLOYMENT BY LEVEL OF GOVERNMENT France is a unitary state with a long-time tradition of state intervention in society. It is not surprising that the state civil service is bigger than the local civil service. State civil service is approximatively 60 per cent of total government employment. One has to keep in mind that there are three civil services: state, local and health, sharing many common features in terms of recruitment (competitive exams), legal and social rights. Even if each of them is managed through specific regulations (for instance, the local civil service does not fit in the corps system as it does within the state civil service but had instead ‘professional careers’ that determine positions and duties), they have been integrated within the same legal framework with the statute of 13 July 1983. One of the most disputed questions about the civil service is whether this legal integration must be pushed forward or, to the contrary, abolished, in order to give more freedom to local governments when confronted with personnel management. Another point to be raised is related to the deep change fostered by the decentralization process initiated in 1982. The 1982 reform established local management freedom as a legal principle. It was a step toward the acknowledgement of local governments in France. But the reform has not changed the classical local structure: regions, départements and municipalities are still the three basic elements of local public administration with

The French public service paradox

229

rather different duties and policy areas. Local civil service in all three levels of local government is managed through the same legal framework that establishes a balance between a free personnel management for the local politicians who have been given the right to choose and promote their senior civil servants and the need to obey national regulations that draw a complex legal framework requiring competitive exams for a majority of bureaucratic positions and defining social and professional rights in accordance with national laws. State Employment Since 1986, state employment has entered a period of stagnation or slight increase but not of decline. The rate of growth has been severely diminished. Data show that the Socialist era has not really modified long-run tendencies. Between 1981 and 1989 the annual growth rate of the state civil service is 0.7 per cent as opposed to an annual increase of 3 per cent between 1969 and 1980. Moreover, since the 1990s, this growth is concentrated in only a few ministries: Education, Interior, Justice and Culture, which have been priority policy areas for one decade whatever the political colour of the government. As Table 9.3 shows, state employment has been shrinking regularly since 1969, from 67.9 per cent of total government employment to 55 per cent. These data demonstrate that there is no mechanical law, as is usually contended, for an ever-growing state. The state part has been reduced as most public policies are now implemented through complex networks involving the state civil service and the local civil service. There is clearly a new balance between the various kinds of government employment. Employment within the health civil service (95 per cent of which are employed in hospitals) has doubled between 1969 and 1980. Since then its growth has been severely constrained (18.8 per cent between 1980 and 1994) and is concentrated almost exclusively on purely medical personnel (that is nurses or technical personnel, but the health civil service does not include medical doctors). This change can be explained by the new public hospital policy requiring the admission of more people for a shorter stay in order to reduce management costs. This trend has been the starting point of a new wave of social conflicts since 2001 when new regulations reduced the working week to 35 hours. What the Socialist government considered social progress for a majority of workers turns out to be a trap for many civil servants, especially within the health sector. The working time reduction was supposed to foster employment within the business sector while the official policy within the public sector was to slow or stop new recruitments. As a result, the medical sector was confronted with a shortage of skilled nurses

230

The state at work, 1

Table 9.3

Public employment by level of government, 1969–98

Total govt. State Percentage Health Percentage Local Percentage employment civil of civil of civil of (000s) service* total service total service total (000s) (000s) (000s) 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1969

5236 5163 5137 5114 5069 5052 5016 4991 4924 4827 4442 3046

2873 2832 2827 2812 2806 2819 2814 2815 2796 2855 2719 2068

54.8 54.8 55.0 54.9 55.3 55.8 56.1 56.4 56.7 59.1 61.2 67.9

857 849 847 840 834 826 818 810 801 787 702 360

16.3 16.4 16.4 16.4 16.4 16.3 16.3 16.2 16.2 16.3 15.8 11.8

1507 1482 1462 1462 1428 1406 1383 1365 1326 1185 1021 618

28.8 28.7 28.4 28.5 28.1 27.8 27.5 27.3 26.9 24.6 23.0 20.3

Note: * Including defence ministry (military and civilians), postal service and FranceTélécom but excluding national public agencies and public-owned companies. Source: Direction Générale de la Fonction Publique, annual and INSEE 1992.

or medical doctors while user demand for more and better health care was climbing especially within the emergency services. The pressure of work rapidly proved to be unbearable for the medical teams who did not want to compromise health care quality. In 2003 this question was still at the heart of social conflicts within the public health sector. Local Employment The 1982 decentralization process did not give the same rights and duties to the various local governments. There has been no clear-cut devolution of authority on some policy areas. Policy areas are actually shared between the various levels of government even if each level is partially specialized. Regions are dedicated to economic co-ordination and support. Of course they play a new role in connection with the EU authorities as they manage most of the grant system but their policy range is rather narrow. The decentralization process aimed at fostering departments dedicated to social and welfare policies, and municipalities with reinforced responsibilities for housing policies and public utilities such as sewerage, garbage collection, school management and so on.

231

The French public service paradox

Table 9.4

Local government employment, 1969–98 (000s)

Total local Municipalities Départements Regions Special and Local government generalquangos employmenta purpose associations 1998 1996 1993 1989 1985 1980 1969

1658.0 1521.0 1447.0 1237.3 1185.0 1021.0 618.1

1131.0 1109.9 1068.9 916.6 862.8 754.9 490.8

168.5 160.2 156.5 141.5 155.0 133.7 77.6

9.7 8.6 6.7 4.6 2.9 0.3 0.0b

170.7 134.6 120.6 83.7 86.1 66.0 13.7

172.0 108.4 94.4 90.9 77.8 66.1 26.0

Notes: a. Data in Table 9.4 differ slightly from those shown in Table 9.3 because the statistical basis (drawn here from the INSEE series) is not exactly the same. For instance, in Table 9.4, the local government civil service includes state agents at the disposal of local authorities. Other differences may also be accounted for by temporary agents with private law contracts may or may not be included in the official statistics. b. Does not apply as regions are not local autonomous entities before 1972. Source: INSEE 1992 and Ministère de la Fonction Publique and Réforme de l’Etat 2001.

Thus the pattern of growth across time is not similar. Of course the period of the most rapid expansion was the 1960s. But, between 1969 and 1980, state civil service employment increased by 31.4 per cent while the local civil service increased by 65.2 per cent. After 1982, the growth rate of the local civil service accelerated further when the new decentralization laws were implemented. Between 1982 and 1985, local employment increased by 16 per cent while state employment increased by 1.4 per cent. After 1985, there was still growth in local employment but the rate was not the same (see Table 9.4). Local governments discovered rapidly that more personnel means more local tax and decided to improve their management rather than to hire more civil servants. Local elected leaders had not been used to local autonomy and realized suddenly that decentralization has created a new pressure for local accountability. The most numerous type of local authorities (the concept of local government is somewhat inappropriate in the French case) are the municipalities (38 000), then the départements (100) and the regions (22). As in Germany, most municipalities are involved in general-purpose or specialpurpose associations in order to manage common duties (garbage collection, water distribution and so on). These associations are regulated on the basis of well-defined legal formulas (syndicat, communauté de communes

232

The state at work, 1

and so on). They generally raise additional local taxes collected by state fiscal services at the local level. Between 1982 and 1985 there was a comparable increase in employment in both municipalities and départements: 14.3 per cent and 15.9 per cent. The huge growth of employment in the regions (190 per cent during the same period) must be related to the very small absolute number of civil servants (the civil service of the regions is almost exclusively composed of managers or specialists). One may also point out the systematic increase of employment in local quangos as the decentralization process has fed a fragmentation of local administration structures: 9.4 per cent between 1982 and 1985 but 16.8 per cent between 1985 and 1989, while departmental employment was declining and that of municipalities growing at a lower rate (6.2 per cent). Between 1989 and 1998 public employment in local quangos continued to grow at a huge rate: 89 per cent as compared to 19 per cent in départements and 23.4 per cent in municipalities. The widespread use, and sometimes misuse, of these private bodies receiving public money is one of the distinctive features of the decentralization process for recent years as they allow local politicians to offer jobs or new services without too much oversight. These data disclose qualitative changes as local authorities had inherited from the centralization period a population of blue-collar workers, ageing and under-qualified. The decentralization process has called for a renewal of this population and a major effort in the middle of the 1980s was dedicated to new recruitment and training policies in order to get skilled young professionals able to negotiate with their state civil service counterparts or their private sector partners and to undertake serious management of the growing budgets of local authorities.

PUBLIC EMPLOYMENT BY GEOGRAPHICAL AREA There are differences in the level of government employment among the various regions in France. Corsica has benefited from the higher concentration of civil servants as they are 13 000 in 1998 for only 256 000 inhabitants, in a rural area. This may demonstrate that the civil service is regarded as a major source for employment or used to heal the social problems where private business is relatively rare or where the regional economy has suffered from industrial abandon as was the case in North-East regions (Champagne-Ardenne, Lorraine) when the traditional coal plants and mills were closed. As shown on Table 9.5, the differences between regions were relatively stable over the 20 years from 1978 to 1998. Almost 40 per cent of state employment is concentrated within three regions: Ile-de-France,

233

The French public service paradox

Table 9.5 State government employment, 1978–98: ratios by region (including statutory civil servants and contractual public agents) Percentage of total

Per 1000 inhabitants in 1998

1978

1988

1998

Alsace Aquitaine Auvergne Bourgogne Bretagne Centre Champagne-Ardenne Corse Franche-Comté Languedoc-Roussillon Limousin Lorraine Midi-Pyrénées Nord-Pas-de-Calais Basse-Normandie Haute-Normandie Pays de la Loire Picardie Poitou-Charente Provence-Alpes-Côte d’Azur Rhône-Alpes Ile-de-France Overseas territories Foreign service

2.7 4.6 2.4 3.0 4.0 3.8 2.4 0.5 1.9 3.3 1.5 4.2 4.4 6.1 2.2 2.7 4.1 2.7 2.6 7.1 8.6 25.2 2.4 n.a.

2.8 4.6 2.4 2.9 4.2 3.8 2.4 0.5 1.9 3.5 1.5 4.1 4.4 6.2 2.2 2.8 4.3 2.8 2.6 7.0 8.7 23.8 2.5 1.2

2.6 4.6 2.2 2.5 5.0 3.8 2.5 0.6 1.9 3.4 1.2 4.4 4.3 5.5 2.1 2.5 3.9 2.6 2.7 8.0 7.9 20.8 3.4 1.8

34.3 36.0 37.5 36.0 39.0 35.6 40.8 50.8 38.0 32.6 38.3 41.5 36.9 30.7 34.5 31.6 26.7 31.8 36.0 39.4 30.8 43.3 – –

Total

100

100

100

37.4

Note: Total per cent exceed 100 due to rounding. Source: Direction Générale de la Fonction Publique, annual.

Rhône-Alpes, Provence-Alpes-Côte d’Azur. The explanation is easy to find as these three regions include the main concentrations of the French population and offer a high density of urban structures. Between 1978 and 1998 the regional civil service population has been somewhat modified with the share of the region Ile-de-France reduced from 25 per cent to 21 per cent, as a result of the decentralization process. One of the most difficult problems for the managers of the state civil service is geographical mobility. As a rule of thumb, seniority is regarded as

234

The state at work, 1

giving a right to the civil servants to go to attractive regions, such as the Côte d’Azur, and to escape the professional stress and difficulties of urban areas. This trend is particularly observable in such sectors as education or police, where the civil servants are confronted directly with users who are not always kind and obedient. As a result, the most trained and able civil servants are localized in areas where there are few professional problems while the young and inexperienced civil servants are appointed in the industrial North or in the Parisian suburbs. Between 30 and 40 per cent of the teachers above 50 are working in the enjoyable Côte d’Azur or Midi-Pyrénées regions.

PUBLIC EMPLOYMENT BY MINISTRY A very striking characteristic of state employment in France is that more than 90 per cent of ministries’ employees work in field offices throughout the country. Contrary to the classical image of a very centralized bureaucratic system, the executive structure is not concentrated in the capital. One can find in Paris the ‘headquarters’ of the ministries (cabinets or ministers’ staffs, division heads, central staffs) and, of course, the core of state power, that is the Prime Minister’s office and its agencies as well as the President’s cabinet which could be compared accurately with the White House Office in the United States. This fragmentation may be observed also in the personnel management system. Personnel Management The French civil service has been governed since 1946 by a general statute whose basic principles were confirmed and developed in the 13 July 1983 law. Other laws were added in 1984 and 1986 for the local civil service and the health civil service in order to adapt the legal framework to these two specific groups of civil servants. As a whole, the basic statutory principles are quite similar to those one may find in other western countries (equal access to the civil service – a principle set for the first time in 1789 – careers governed by the merit system and so on) but the social rights are particularly developed as the French civil servants may strike (with the exception of the military, judges and the police) or protect their rights in labour unions. This legal framework is put under the regulation of the Direction Générale de la Fonction Publique, attached to the Prime Minister’s office, and the scrutiny of the administrative law courts which constitute a complete juridictional hierarchy with the Conseil d’État at the top. If administrative law is largely developed, the legal aspect does not tell the whole story of the way the personnel is really managed.

The French public service paradox

235

The other major aspect of civil service management in France lies in the corps system where careers are managed through specific legal but also social rules associated with each civil service corps. The balance between seniority and merit may vary considerably from one corps to another. The real power of the ministerial managers is then determined by precise rules guiding the career path (for instance, the upper ministerial positions under the department head level are reserved to the corps of administrateurs civils in due proportion of 70 per cent) or by implicit rules (for instance, it is common knowledge that the Treasury or the Budget department within the Finance ministry must be managed by a civil servant coming from the corps of Inspecteur des Finances). Management practice is highly dependent of the way each corps is managed. If the grands corps have been able to manage carefully the careers of their members other corps have not. The corps structure is an obstacle for horizontal mobility but legal provisions allow one to work in another ministry or public body while keeping the ties with one’s corps. As a consequence, ministerial autonomy in personnel management is relative as there is no Ressortprinzip as in Germany. Personnel management is not entirely under ministerial hierarchy control because managers cannot choose their assistants coming from national competitive exams and schools and because they have also to handle demands for geographical mobility coming from civil servants who want to come back to their native region or to join their family. On the other hand, there is always room to manœuvre as ministerial departments or public agencies may hire contractual agents as long as they have the financial means to do so. Personnel management has gained in fluidity in the local governments as the local civil service is not organized along the lines of corps and appointments at the upper levels are decided by local politicians. Many technical reforms have been implemented since the 1990s in order to ease personnel management at the state level, especially through the introduction of ‘job descriptions’ or the multiplication of ‘functional positions’ into the career structure. The third aspect of the civil service management is the weight of labour unions in deciding professional questions as their representatives are members of official councils and committees where promotions are debated through negotiations with management. This point has often been criticized in the name of modern public management as labour unions are likely to favour seniority over competition and performance. But the number of strikes in the public sector is so high (Mouriaux 2000) that any reasonable (that is successful) manager prefers to compromise with the unions rather than to confront them in endless conflicts. On the whole, personnel policy is a balance between budget, social, political and corporative considerations with a priority given to the last three aspects of the question.

236

The state at work, 1

State Ministerial Structure In order to understand the data shown hereafter, it is necessary to keep in mind that there is not a well defined number of ministries and state secretaries in France. Their number varies approximately between 20 and 30 and depends heavily upon the organization as well as the policy priorities of each government. A given policy area, such as research and universities, may be managed within one major ministry or within many ministries or by many state secretaries due to political considerations about the necessity to integrate the public research bureaucracy within a leading sector (such as the industry or the education ministry) or, on the contrary, to have a small specialized structure with high visibility and co-ordination powers (Rouban 1995). The same could be said about environmental, transportation or tourism policy. Needless to say that this changing structure makes any statistical follow-up very difficult. I have thus chosen to present the data according to big classical policy areas which have always been managed by a ministry between the 1970s and 2000. It must be added that, as in Germany, many state functions are also performed by administrative agencies under ministerial control. Since the 1980s there are also independent regulatory agencies such as those which rule the stock exchange, public broadcasting or computerization of official data. Personnel Including Field Staff As shown in Table 9.6, over half of state civilian employment is located in the education ministry and nearly 60 per cent in the education policy area if the universities/research ministry or state secretary is added. The other big ministry over time is the finance ministry with the treasury field networks. The other ministries, in 1998, were all under 10 per cent, the most important being the interior ministry and the infrastructure ministry. What is the trend between 1980 and 1998? The first point to stand out from the figures is the decline of the defence sector. The employment number has dropped as a consequence of the systematic reduction-in-force and reorganization of the defence policy. The Veterans Secretary of State is being phased out as there are fewer and fewer veterans to take care of. Another point is the relative decline of foreign affairs functions. Since the early 1990s, the foreign affairs bureaucracy has been under scrutiny for its cost and its alleged inefficiency. A systematic reorganization of its locations abroad has taken place in coordination with the Finance ministry and many services have been merged within the ministry central location since 1993. A third point is the strengthening of the industry and the universities ministries. Although both of them have benefited from ministerial

237

3 074 1 017 037 – 6 820 494 954 199 782 146 908 96 294 47 935 29 925 27 165 18 081 16 346 11 633 11 540 4 897 6 125 1 330 353 141 809

1985

Source:

Direction Générale de la Fonction Publique, annual.

1990 3 089 1 073 307 – – 474 958 196 923 155 299 102 093 52 950 29 735 26 063 15 485 14 701 6 213 10 909 3 794 8 897 1 789 – 128 702

Notes: a. Including statutory civil servants and contractual public agents. b. Civilian employment only including state workers and civil servants.

2 627 851 096 70 611 28 047 440 623 186 630 135 306 95 636 38 048 27 145 26 826 24 778 14 345 11 164 7 815 5 011 4 577 1 307 319 142 068

1980

State civilian employment by ministry, 1980–98a

Prime Minister’s Office Education Universities Sports Postal service Finance Interior Infrastructures and environment Justice Agriculture Labour, human services Foreign affairs Transport Cooperation Culture Veterans Industry Overseas territories Tourism Defenceb

Table 9.6

3 160 1 039 138 143 568 10 504 – 204 472 183 981 116 664 65 996 38 074 30 086 14 076 – 3 511 14 744 2 347 – 2 639 – 91 066

1998 0.2 52.9 7.3 0.5 – 10.4 9.4 5.9 3.3 1.9 1.5 0.7 – 0.2 0.7 0.1 – 0.1 – 4.6

Percentage of total in 1998 20.3 22.1 103.3 62.5 – 9.6 36.0 22.0 73.4 40.3 12.1 43.2 – 68.5 88.6 53.2 – 102.0 – 35.9

Percentage change 1980–98

238

The state at work, 1

reorganization, this trend indicates that these policy areas have been regarded as priorities by recent governments. The Socialist government decided in 1997 to merge the industry ministry and the finance ministry, a significative political and administrative change that indicated that there was no more industrial policy. Two other sectors have benefited from new recruitments: culture and justice. The case of justice could be explained by the growing number of judicial activities since the 1980s and the fact that magistrates have asked, with the support of public opinion, for the means to manage legal procedures more quickly. The culture sector has always been a priority for Socialist governments but this alone cannot explain a 54 per cent increase. Its evolution has to be connected with a growing concern for the national patrimony and its preservation. A fourth point is that, contrary to expectations, the social affairs ministry has not benefited from the huge priority given to social questions by the various Socialist governments. This can be explained by the growing number of agencies, national and local, as well as by the systematic use of the associations, which have been given official administrative clearance and budget support, in order to get close to the population demands. Core Ministerial Personnel In order to appreciate the hierarchy within French ministries, it is necessary to understand that there is a real (that is social and financial) difference between the ‘managers’ and the higher civil servants. The civil service is globally divided into three main categories : the ‘A’ category, which includes the managers and all the civil servants who have been recruited through competitive exams after a four-year degree course at university; the ‘B’ category where one can find the intermediate level managers; and the ‘C’ category, which is the category of the employees and clerks. To belong to the ‘A’ category does not indicate automatically that you are (and you are regarded as) a higher civil servant. Higher civil servants have also to be appointed to higher positions (division head, subdivision heads and so on) or to belong to the administrative or technical grands corps. The corps structure indicates the ministerial hierarchy. For instance, most upper management posts in the industry ministry are occupied by managers coming from the corps des ingéniears des Mines, which is a technical grand corps created in 1793. The upper levels of the finance ministry are offered to the members of the Inspection générale des Finances, which is an administrative grand corps. Entrance into a grand corps depends upon your rank after having been selected through the Ecole nationale d’administration for the administrative corps and, generally, through the Ecole Polytechnique for the technical grands corps. But things are not always so simple. Members

The French public service paradox

239

of the same corps may be differentiated along the lines of the ministry where they have been appointed. For instance, everybody knows that managers in the finance ministry, coming from the administrateurs civils corps, are better paid, have better career prospects and social consideration than those of their colleagues who are also members of the administrateurs civils corps but have been appointed to the social affairs ministry. On a whole, the best way to evaluate the proportion of higher civil servants is to use a functional criteria and to select those who occupy the upper management positions. There are approximatively 5000 upper positions within the ministries. The proportion of ‘A’ category civil servants within French ministries may vary between 8 per cent (interior) and 72 per cent (cooperation). This proportion depends upon the nature of the administrative work implying a variable proportion of expertise and is related to the fact that some ministries have a staff administration structure (it is clearly the case of the cooperation ministry) while others are organized as line administrations which have to handle the actual implementation of public policies. The number of division heads (directeurs d’administration centrale), who are political appointees, indicates the organizational importance of each ministry and the number of its functions. As indicated on Table 9.7, the proportion of women in upper positions is very low. On the average, women accounted only for 17.6 per cent of division heads in 1999 and for 20.6 per cent of the other upper management positions in the ministries. The proportion of women does not climb much in the most prestigious sectors such as the finance or the foreign affairs ministries or in the traditional maledominated sectors such as the interior ministry which manages the police. Nevertheless, these data show a real increase, as the average proportion of women in these positions was about 5 per cent in the early 1990s.

PUBLIC EMPLOYMENT BY POLICY AREA In order to understand the distribution of total public employment in major policy areas in France, it is necessary to keep in mind some basic features of public administration. Although the core of each policy area is comparable with that of Germany or the United States, there are some problems of data collection. The first one is related to the fact that, as in Germany, civil servants have been trained generally to become generalists. The second complication is due to the various legal provisions which allow a public servant to be paid by one ministry while serving in another one, so there is no exact connection between corps, ministries and policy areas. A third point is that some services, like fire protection, are local services while others, school teaching for instance, always involve state civil servants even

240

The state at work, 1

Table 9.7 Positions in ministries in 1999 (civilians only, statutory and contractual agents)

Prime Minister’s Office Education Universities Sports Finance Interior Infrastructures and environment Transport Justice Agriculture Labour, human services Foreign affairs Cooperation Culture Veterans Overseas territories Defence Notes: a. Included in the figure of b. Included in the figure of c. Included in the figure of d. Included in the figure of e. Included in the figure of f. Included in the figure of

Percentage of A category civil servants

Division heads

Percentage of women as division heads

35.3 61.5 70.3 51.4 20.8 7.6

12 14 –a –b 24 28

16.7 35.7 – – 12.5 7.1

15.2 –c 19.9 45.5 24.3 24.5 72.3 35.2 9.8 28.9 8.6

17 – 9 9 28 19 –d 6 –e –f 6

11.8 – 33.3 33.3 25.0 10.5 – 33.3 – – 0.0

the education ministry. the education ministry. the infrastructure ministry. the foreign affairs ministry. the defence ministry. the interior ministry.

Source: Direction Générale de la Fonction Publique, annual.

if they work all over the country. The fourth point is that statistics for local services are not available in terms of policy areas as the local civil service is organized along the lines of ‘careers’ (administrative careers, technical careers, cultural careers and so on). In order to allow comparisons with the Derlien–Peters research, Table 9.8 includes the employment in postal and telecommunication services, railways and professional military as subtotals of public employment. In 1990, education and research took the major part of the civilian public employment. This figure includes all educational and scientific activities, that is personnel working in the ministry or for the various state

241

The French public service paradox

Table 9.8

Public employment by policy area in 1990 and 1997 1990 N

%

1997 Ratioa

N

%

Ratioa

Education and research Health and recreationb General administrationc Financial administration Police, fire Social services Housingd Defence, civilians Legal services Transporte Other servicesf

1 168 796 830 000 452 067 196 923 202 601 240 063 102 093 128 702 52 959 54 513 627 794

23.2 48.7 16.5 68.6 9.0 126.0 3.9 289.0 4.0 281.0 4.8 237.0 2.0 558.0 2.5 443.0 1.0 1076.0 1.0 1055.0 12.5 90.7

1 242 000 849 000 544 272 207 262 229 498 277 915 118 460 95 674 64 783 40 000 688 134

23.5 46.6 16.1 68.3 10.3 106.0 3.9 280.0 4.3 253.0 5.3 209.0 2.2 491.0 1.8 610.0 1.2 906.0 0.7 1450.0 13.0 84.3

Sub-total

4 056 511

80.4

14.0

4 356 998

82.7

13.3

474 959 209 400 299 569

9.4 4.1 5.9

120.0 272.0 190.0

439 700 176 000 295 486

8.3 3.3 5.6

132.0 329.0 196.0

5 040 439

100.0

11.3

5 268 184

100.0

11.0

Post, telecommunications Railways Professional military Total

Notes: a. Ratio inhabitants/civil servants. b. Including only the health civil service. c. Including: foreign affairs, culture, cooperation, overseas territories, agriculture, veterans, Prime Minister’s Office, national administrative and business-like public agencies and administrative staff working in local authorities. d. Infrastructure ministry. This figure excludes local agencies. e. In 1997, transport was merged with the infrastructure ministry and the data calculate only the population of the Parisian urban transport system (RATP). f. Estimate based upon the population of public agents working in the sports ministry and in technical, cultural and sport careers of the local civil service. Source: Personal calculations on the basis of date from the Director Générale de la Fonction Publique and INSEE.

secretaries, in universities and colleges, in scientific agencies such as the CNRS (the most important agency covering each field of research), the INSERM (biology), the CEA (nuclear research), the CNES (space research) or, for instance, the IFREMER (oceanography). It includes also administrative and technical staff working in the ministry or the various agencies. As in Germany or the United States, the second largest policy area is health. This figure includes only the health civil service (medical and

242

The state at work, 1

administrative staff) but does not include social activities linked with health care which are accounted for in the social services. The third largest category is the ‘other services’ as this figure includes all the local personnel working in technical or cultural professions. Agents working within technical careers account for about 50 per cent of the whole local civil service. The fourth largest category is the general administrative personnel. I have included in this figure agents working in ‘administrative’ ministries and having to manage regulations as well as the major part of administrative agencies. Social services are the fifth largest category with about 5 per cent of the total as in Germany. This rank is rather surprising as the various governments have put the emphasis on social policies. But this figure does not include social security personnel who are not regarded as public agents as they work under private law contracts (even if they benefit from outstanding professional guarantees). Police and fire services (the fire services are only local) are sixth in rank. The figure includes the administrative staff working in the interior ministry. But the population ratio (in 1990 the French population was 57 million) show that there are relatively fewer policemen and firemen than in Germany or in the United States (1 for 281 inhabitants). The seventh category is financial administration with about 4 per cent of the subtotal. This relative importance is due to the fact that the finance ministry network covers all the country and work for both the state and for local authorities (tax collection). The defence sector is at the eighth rank but the real figure for civilians working for the defence sector is actually more important as some public agencies (such as the CEA which is accounted for in the research sector) devote most of their activity to military research. As a whole, one may observe that the overall employee/inhabitants ratio is lower in France than in Germany or in the United States (11 against 12) but this would certainly change if one had included the figures of the public companies. This ratio highlights the fact that there is a high administrative coverage of each policy area more than a concentration of public employment on some functions. Another point to be raised is that the administrative density is higher in social policy areas: in 1990, one employee for 69 inhabitants in the health sector and one employee for 237 in the social and para-medical sector. As shown on Table 9.8, there is no change in the policy area hierarchy between 1990 and 1997. Proportions remain constant and one may just observe the slight effects of the priority given to some sectors (legal services or social services) as well as the growing share taken in the overall data by the general administration or the ‘other services’. Another visible trend is

The French public service paradox

243

due to the effects of the privatization process of the postal and telecommunication sector or to the process of reduction-in-force in the military and railways sectors.

SOCIAL DIFFERENTIATION IN THE PUBLIC SERVICE French public employment is heavily dependent upon the administrative law system. Administrative courts have always played a crucial role in determining the boundaries of the civil service as compared to public employment as a whole or in drawing the frontiers between public employees and other employees used by public services but obeying common law rules and judicial power. As in Germany, personal characteristics are considered only for gender, as race or ethnicity are not (and do not have to be) given in official statistical surveys. Of course, equal opportunity for women is a common legal and political issue. Public Employment by Legal Status In France, at the state as well as at the local level, public employment is divided into two main legal categories: civil servants, enjoying the protection of the 1983 law, and contractual agents, who are public agents appointed on the basis of public law contracts. Two others categories exist: state workers, who are not statutory civil servants but enjoy specific legal provisions, and can be found especially in defence industrial plants, and contractual agents appointed on the basis of private law contracts, generally on a part-time basis. The first of these two categories is rapidly disappearing with the privatization and contracting-out process: they were 57 533 in 1997 as opposed to 112 000 in 1980. As shown in Table 9.9, there is a growing proportion of statutory civil servants within the state employment. The decreasing proportion of contractual agents between the 1980s and the 1990s is largely due to the effect of sporadic attempts to transfer contractual agents to statutory legal status. This trend has been counterbalanced by the fact that most ministries need contractual agents in order to fill specialist positions or to manage gaps in their human resources policies (this is especially the case of the national education ministry). This explains why the proportion of non-statutory agents remained steady in the 1990s. Moreover, the integration of former contractual agents within the statutory framework is not an easy task, because it generally requires adjustment of salaries and change to the career equilibrium which already prevails within the statutory corps. This is why labour unions both ask for

244

Table 9.9

The state at work, 1

State employment by status groups (civilian only), 1947–98a Statutory civil servants

1998 1996 1994 1992b 1990 1980 1969 1962 1947

Contractual agents

N

%

N

1 670 356 1 661 941 1 626 776 1 585 213 2 001 205 1 791 421 1 175 366 937 509 534 044

85.7 86.1 85.2 84.1 84.7 79.9 70.6 70.4 50.4

208 082 198 553 210 025 219 985 275 959 337 333 375 996 266 313 356 143

Percentage change 1947–69 120 1969–80 52 1980–98 6.8

State workers %

10.6 10.2 11.0 11.7 11.7 15.0 22.6 20.0 33.6

5.5 10.3 38

N 69 187 69 695 70 712 79 391 83 204 111 778 112 436 126 971 169 465

% 3.5 3.6 3.7 4.2 3.5 5.0 6.7 9.5 16.0

33 0.5 38

Notes: a. Full and part-time but excluding public agencies. b. The decrease in the population of state statutory civil servants between 1992 and 1990 is due to a change in the statistical framework as the postal service as well as France Telecom are excluded since then. Source: Direction Générale de la Fonction Publique, annual.

more integration of contractual agents and career improvements for the statutory civil servants who are also their members. It is necessary to underline that the legal situation of contractual agents is not always very clear as they obey public law principles without having the same protection as their statutory counterparts (they do not enjoy job security) and without benefiting from the judicial rules used for private sector wage-earners. Of course, the growing proportion of statutory civil servants is a major constraint the various governments have to face when they consider any reform of the public sector structures. Public Employment by Gender As in many other western countries, French governments have always tried to use the civil service as a social laboratory for developing new labour practices and improving working conditions. Historically the civil service has been a pioneer in social advancement, initiating innovative health

245

The French public service paradox

Table 9.10 Public employment by gender – proportion of women (%), 1969–98

1998 1993 1989 1987 1983 1969

State civil service (civilian)

Local civil service

Health civil service

56.7 54.7 51.4 50.9 50.2 45.6

59.9 59.5 56.5 55.6 53.8 42.9

74.0 73.0 72.4 72.0 71.8 69.3

Source: Direction Générale de la Fonction Publique, annual and INSEE 1992.

protection systems or retirement plans that have been used some years later in the private sector. Another major point is related to the fact that the whole public sector has been regarded as a way for social promotion of middle classes. Gender is one of the most important aspects of the public service social representativeness. In 1998 a majority (57 per cent) of the civilian state employees were women and public employment, as a whole, is largely female with comparable proportions to those observed in Germany. As shown on Table 9.10, state female employment has increased steadily from 43 per cent at the end of the 1960s to 56 per cent at the end of the 1990s. The difference with Germany lies in the fact that women are not under-represented in full-time employment. The proportion of full-time female employment in state ministries was 84.7 per cent in 1998. At the local level the proportion of fulltime female employment was still 82 per cent for statutory agents and 88 per cent for contractual agents, as the proportion of full-time male employment was higher, at 98 per cent for both statutory and contractual agents. As it has been already outlined, the relative position of women within the higher ranks of the civil service is not very good. Most top civil service positions are occupied by men. The comparison between legal categories is somewhat misleading. Of course, in 1998, women are 54.3 per cent in the A category, 67 per cent in the B category, and 53 per cent in C category. They also constitute 57.4 per cent of the contractual agents. But these figure have to be modified as they include national education teachers, who belong to the A category. The national education data excluded, women are only 33.8 per cent in the A category. Within the A category, it is clear that women accede to the highest positions in small proportions. For instance, women constitute more than 59 per cent of secondary school teachers, but only 40 per cent of line managers and 19 per cent of top level

246

The state at work, 1

managers. This is a very similar proportion to the one found in Germany. Anyway, the situation of women has considerably improved since the 1980s. For instance, the proportion of women occupying top level management positions was 11 per cent in 1992 and 7.5 per cent in 1982. Fresh attention has been paid since 1997 to the career and the access of women not only to administrative top positions but also to political responsibilities (Sineau 2001). Table 9.11 Total state budget expenditures and detailed state civil service expenditures, 1970–98 (salaries and budget cost of retirement pensions) (FF millions)

1970 1975 1980 1985 1990 1991 1992 1993 1994 1995 1996 1997 1998

Total state budget expenditures

State civil service salaries

State civil service retirement pensions

157 285 296 094 579 602 1 058 790 1 281 924 1 335 610 1 425 155 1 502 845 1 552 549 1 596 764 1 642 041 1 655 436 1 661 506

43 836 86 493 172 380 274 120 339 738 354 452 372 143 393 394 409 914 424 049 436 853 446 851 461 486

12 706 25 380 52 040 96 300 123 184 129 588 136 831 144 874 150 033 158 052 164 920 170 757 174 314

Source: Direction Générale de la Fonction Publique, annual and INSEE 2000.

CONCLUSION The evolution of the public sector shows that there are many common macro-trends between France and Germany, whatever the constitutional and structural differences. Both countries have experienced a huge increase in overall public employment since the 1960s, a growing proportion of women within the ranks of the civil service, an expansion of part-time jobs, a high priority given to education. But there are also marked differences, especially in the will to stop the growth of public employment. The data show clearly that French governments have been quite able to change the distribution of public employment among the various ministries when they

The French public service paradox

247

want to do so. This is particularly clear in the defence sector. But there is also a broad consensus between the 1980s and the early 2000s to safeguard public employment to maintain social peace. France is perhaps the only western country having rejected the new public management model. Whatever their political colour, all the governments have resisted pressures asking for state retrenchment. The growth of public employment has only slowed down since 1993 and the overall data are very impressive at the end of the century. There is still a high administrative coverage of every aspect of social life and one may notice that public opinion polls do not show any kind of popular exasperation. The difference between France, Germany or the United States does not lie in the constitutional framework but in the relationship that prevails between society and the various bureaucracies which feeds and protects public employment. The decentralization process in France has led to reinforcement of the local civil service and the search for a better way of life for the elderly (and demographic considerations such as the evolution of the average lifespan) have required more jobs in the health civil service. Only public enterprises have experienced drastic changes with the privatization process and the EU deregulation policies. The changes that have occurred in their management have not always been translated into employment terms. In order to avoid tough conflicts with labour unions, most managers have chosen to reduce employment at the rate older agents retired. This is at the heart of the French paradox: the proportion of the labour force employed in the public service is very high but the public service is so fragmented that there is always room for local or sectoral evolution. Of course, things began to change in the 1990s. More attention was paid to the quality of public service delivery and the budget burden of 5 million civil servants became a major concern in a time of low economic growth, even for Left governments. At the end of the 1990s, the cost of the retired civil servants alone represented about 10 per cent of the total national budget as demonstrated in Table 9.11. Demographic projections showed that governments could confront a serious budget crisis when almost half the state civil service would be retiring between 2000 and 2012. Moreover, the budget expenditures for the entire civil service, national and local, grew at the rate of 3.6 per cent per year. Nevertheless, real changes occurred only after 2003 with the Raffarin government. The public service retirement system, which gave advantages to civil servants when compared with business retirement plans (the civil servants could keep 75 per cent of their last salary on the basis of a 37.5 years working life) was completely reformed on the private sector model and civil servants were forced, after some sporadic conflicts, to accept the new retirement plan designed to save the budget equilibrium.

248

The state at work, 1

REFERENCES Direction Générale de la Fonction Publique (annual 1982–1999), La fonction publique, Paris: La Documentation française. INSEE (Institut National de la Statistique et des Etudes Economiques) (1992), Annales statistiques de la fonction publique 1945–1969–1989, Paris: INSEE. INSEE (Institut National de la Statistique et des Etudes Economiques) (1994), Marché du travail, séries longues, INSEE Résultats 305-306, Paris: INSEE. INSEE (Institut National de la Statistique et des Etudes Economiques) (2000), Tableaux de l’économie française, Paris: INSEE. Kickert, Walter (ed.) (1997), Public Management and Administrative Reform in Western Europe, Cheltenham, UK and Lyme, USA: Edward Elgar. ˙ tat (2001), Observatoire de Ministère de la fonction publique et de la réforme de l’E l’emploi public, annual report, Paris: La Documentation française. Mouriaux, René (2000), ‘La conflictualité dans les services publics’, in Luc Rouban (ed.), Le Service public en devenir, Paris: L’Harmattan, pp. 157–76. Rouban, Luc (1995), ‘Public administration at the crossroads : the end of the French specificity?’, in Jon Pierre (ed.), Bureaucracy in the Modern State, Aldershot, UK and Brookfield, US: Edward Elgar, pp. 39–64. Rouban, Luc (1998a), The French Civil Service, Paris: La Documentation française. Rouban, Luc (1998b), ‘The view of public opinion’, in F. Gallouédec-Genuys (ed.), About French Public Administration, Paris: La Documentation française, pp. 183–91. Rouban, Luc (1998c), ‘La politisation des fonctionnaires en France: obstacle ou nécessité?’, Revue Française d’Administration Publique, 86, 167–82. Sineau, Mariette (2001), Profession: femme politique, Paris: Presses de Sciences Po.

10.

The political allocation of incessant growth in the Danish public service Lotte Bøgh Andersen, Jørgen Grønnegaard Christensen and Thomas Pallesen

INTRODUCTION A few keywords summarize the development of the Danish public sector in the post-war period. First of all, the growth has been immense. In comparison with other Western countries, Denmark was a low spender until the first part of the 1960s. The position was radically changed 30 years later. Now the Danish public sector claims more than half of the Gross National Product, and in this respect ranks among the top spenders in the OECD area. Second, the Danish public sector is distinguished by its allocation of resources between transfers and public consumption. In a comparative perspective – also in comparison with other public sector high spenders such as the Netherlands – public consumption takes a high share of the public budget. At least in the Danish context, the relatively strong emphasis on public consumption has entailed a vast increase in public sector employment. The third major feature of the Danish public sector is the allocation of public tasks across government levels. The growth in public spending and public employment is concentrated at the local government level (counties and municipalities) while the central state level has stagnated. In part this development is explained by a major reform of the public sector in the 1970s by which a number of public sector areas were transferred from central to local government authority. Nevertheless, the main explanation for the increasing role of local government is that growth has taken place in areas where traditionally local governments were the main public provider, for example education, social and health care services. Along with the development of these policy areas, the Danish public sector has beome increasingly a female workplace. By the turn of the century, the typical Danish public employee was a woman providing welfare service in a local government organization. 249

250

The state at work, 1

The main purpose of this chapter is to describe the development of the Danish public sector and public sector employment in greater detail. The presentation is based on a variety of available official data. A longitudinal study of Danish public employment has to be based on these data, although they do not allow a strict historical comparison. The data problems are described in further detail in the methodological appendix.

INSTITUTIONAL STRUCTURE AND PUBLIC EMPLOYMENT From one perspective, the institutional structure of the Danish public sector is marked by extreme stability. In modern times, there have always been three levels of government: a central government (state) level and two local government ones (14 regional counties and 275 municipalities). The two metropolitan municipalities, Frederiksberg and Copenhagen, are exceptional in having the status of both county and municipality. Municipalities and counties have long been features of the Danish public sector. The autonomy of local governments was written into the first democratic constitution in 1849. However, as stated in the constitution, local government autonomy was and still is circumscribed by central government regulations: ‘Local government autonomy is regulated by law’. Autonomy regulated by law may seem to be a contradiction in terms but it reflects that basically Denmark is a unitary state. Central government may delegate public tasks to local government or it can withdraw delegation by law, depending on the political will of the parliament. This constitutional safeguard of local government autonomy is lacking in meaning as it only prevents the total abolition of local governments. However, local government autonomy has been safeguarded politically. Municipalities and counties are dominated by the Liberal Party and the Social Democratic Party which always – but most often not at the same time – have been part of the governing coalition at the national level. Although they disagree on a number of political issues, these two influential parties in Danish politics have such strong stakes in local government that they have effectively guaranteed local government autonomy. In effect, the Danish public sector has for long been fairly decentralized. In the 1970s the local government position in the public sector was further consolidated by a comprehensive reform. As a consequence, a large number of municipalities and counties were amalgamated and a number of important tasks were transferred from central to new, larger (in Danish administrative jargon ‘sustainable’) local and regional governments.

The Danish public service

251

Most public sector organizations and their employees can be grouped at one or other of the following three levels: central government, counties or municipalities. Still even a simple descriptive analysis is confronted with major difficulties. Apart from the organizations clearly belonging to one of the three levels of government, an ill-defined group of organizations has also always existed. In legal terms, they are often defined as private institutions that are not subjected to the governmental hierarchy. They perform heavily regulated tasks that for all practical purposes are carried out on behalf of central government, the counties or the municipalities. There are several organizations of this type. In Danish public sector statistics they are defined as part of the public sector: ● ●

●

If at least half of the institution’s income is due to public subsidies. If central and/or local governments own more than half of the shares in an enterprise, or are otherwise the dominant owner. This is and has been typical for public utilities, especially those run by either individual municipalities or as the legal and organizational framework for a joint venture among several municipalities. The latter type of enterprises are very similar to the special-purpose associations found at US and German local government levels (Derlien and Peters 1998). Although in recent times they have expanded into new areas, for example tax administration, their use is restricted mainly to utilities and infrastructure. The Danish National Church (the Lutheran church to which about 86 per cent of the population belongs) together with the local government associations and their subsidiaries (chartered accountants, data processing, publishing) are also included.

It is not easy to create a valid and precise picture of the size of the Danish public sector and data do not allow a detailed longitudinal analysis of the development of public sector employment (see methodological appendix). Nevertheless, strong and long-term expansion is an undisputed fact. Even for the post-1978 period for which comparable data of relatively high quality are available, strong expansion in public sector employment is the most conspicuous feature. During the following two decades, public employment grew by 30 per cent. Allowing for a loss in comparability and precision due to the shift in data registration (see Table 10.1, last column), growth increases to about 50 per cent when considering the period since the early 1970s. Since 1970 data covering full-time employment by both government and public enterprises exist, they are also available for formally private institutions that receive most of their income in the form of government subsidies.

252

Source:

1970 1971 1975 1978 1980 1985 1990 1995 2000

1949 1967

275 288 328 384 459 496 510 547

35

Counties and municipalities

See methodological appendix.

210 190 192 189 216 203 189 147

104

Central government

407 485 478 520 573 676 699 699 694

139 368

Total government

n.a. 73 88 104 111 118 113 120 n.a.

n.a. n.a.

Public enterprise/subsidized institutions

}

}

558 624 684 794 811 820 n.a.

Total public employment

Table 10.1 Public employment by level of government, 1949–2000 (000s, full-time equivalent)

F.T.E. – based on pension fund contributions

No. of employees No. of employees – survey data F.T.E. – survey data F.T.E. – based on payroll

Data source/type

253

The Danish public service

Table 10.2

The structure of government employment, 1975–2000

Total number of employees (000s) Central government Counties and municipalities Public enterprise/ subsidized institutions Total As percentage of total labour force As percentage of population

1975

1980

1990

1995

2000

183 363

214 562

201 620

182 599

186 658

89

112

128

130

102

635

888

950

911

946

26 13

34 17

34 18

33 17

33 18

Source: Table based on payroll statistics (personalestatistik).

Table 10.1 shows also the decentralized nature of the Danish public sector and the accelerated local government growth after the comprehensive local government reform in 1970. Public sector growth has given the public sector an important overall position in the labour market. A third of the total labour force in 2000 were public employees. This compares to about a quarter of the labour force in the early 1970s. Notably, the counties and the municipalities became virtual centres of governmental growth. As a result, their share of total public sector employment has increased from less than 60 per cent in 1975 to close to 70 per cent in 2000 (see Table 10.2). This compares to the decreasing share of central government and a relatively stable share of public enterprises and subsidized private institutions. However, behind this stability, as shown below, remarkable changes have taken place.

EMPLOYMENT BY POLICY AREA Table 10.3 displays the distribution of public employment by major policy areas and level of government, including the subsidized private sector and public corporations. The table demonstrates the major bulk of public sector employment in three policy areas: social and health care services, education, and hospital services. Nearly two-thirds of total public sector employment is within these areas. The table also illustrates the decentralized nature of Danish public sector employment where local government employment is three times that of central government. The strong position of local governments in the three

254 71 44 657

80 29 185

Total number employed

102

4 16

– – –

82 –

Supported sector/public corporations

Table based on payroll statistics (registerbaseret arbejdsstyrkestatistik).

421 117 304

2 – –

Source:

– 121

– 74

Local government

Post, telephone, TV, transport, utilities Education and research Hospital, social and health care services Hospital and health Social General administration, armed forces and police Other public services

Central government

944

155 89

423 117 304

82 196

Total

Table 10.3 Public employment by policy area and level of government, 2000 (000s)

100.0

16.4 9.4

44.8 12.4 32.4

8.7 20.7

Percentage of total

6

34 60

13 45 17

65 27

Population ratio

The Danish public service

255

main policy areas, social and health care services, education and hospital services accounts for this. The subsidized private sector plays an important role in education and social and health care services. The same goes for the user-fee-financed public corporations in utilities and telecommunications. In total, the employment of the subsidized private sector and public corporations constitutes more than two-thirds of central government employment. Overall, public employment is a major feature of Danish society. The public employment ratio, that is the number of citizens divided by the number of public employed, is as low as six. In other words, there is one public employee per six inhabitants in the country. Employment by Policy Area Over Time Due to various technical problems with the data, it is impossible to establish a strictly comparable picture of public employment by policy area covering the entire post-war period. As elaborated in the methodological appendix, a number of changes took place in the official statistics in the mid-1970s. Nevertheless, it is possible to give an account of the development in the period from 1950 to 1975, and with some correction for the changes in the statistics, also an account for the 1975–95 development. Due to changes of the official statistics in the late 1990s, it is only possible to provide an estimate of the development from 1995 to 2000. In total, the gap between the early and later accounting principles cannot be bridged entirely, notably because of changes in part-time and full-time definitions. Table 10.4 illustrates that the massive growth of public sector employment in the period from 1950 to 1975 was not evenly distributed between the five broadly defined policy areas. The cultural and social/health care employment increased dramatically in the period. In 1975, more than four times more people were employed in these two sectors than in 1950. Compared to the increase in these two areas, the growth in administration, law and order employment was more modest, but still 2.6 times higher in 1975 than in 1950. Public sector employment in infrastructure stands out as the only area with a low increase in employment. Postal, telegraph and telephone service employment increased 40 per cent, while transportation employment was reduced by 20 per cent in the period from 1960 to 1975. Both categories increased only 10 per cent in the 1960–75 period. In principle, the 1975–95 statistics based on payroll statistics report the same story of public sector employment increase as the statistics of the period from 1950 to 1975. This derives from Tables 10.5 and 10.6. The strong growth of public sector employment continues in 1975–95 in education and social and health care services especially. This is also the case with

256

The state at work, 1

Table 10.4 Major public sector employment areas, 1950–75 (full-time and part-time employees, 000s) 1950

1960

1970

1975

Ratio 1975/1950

51

66

109

133

2.6

44

67

132

182

4.1

74

102

201

303

4.1

32 –

34 32

43 26

45 25

1.4 0.8*

–

301

511

688

2.6

General administration police, armed forces Education, churches, libraries Health care, social services Postal, telegraph, telephone Tram, buses, railways Total (1960 to 1975) Note: *1960–75.

Source: The employment-investigations 1950–75.

Table 10.5 Public employment by policy area and level of government, 1975–95 (full-time equivalent, 000s) Central government

Local government

Supported sector/public corporations

1975

1995

1975

1995

1975

1995

1975

1995

54 31

51 30

30

6

19

32

102 31

89 30

25

39

74

109

4

38

103

186

20 13

1 1

52 64

197 88

29 4

33 1

101 81

230 90

12

13

48

59

60

71

18 18

22 13

20

45

9

13

18 46

22 71

Total number employed 190

169

288

504

64

117

542

790

Post, telephone, TV, transport, utilities Armed forces Education and research Social and health care services Hospital services General administration Police and legal services Other public services

Source: Table based on payroll statistics (personalestatistik).

Total

257

Source:

25 17 35

42 4 – 28

70 8 – 59

13 16 4

–

27

53 43

14 20

Total

13 6

Supported sector/Public corporations

54 29

Local government

Table based on payroll statistics (personalestatistik og registerbaseret arbejdsstyrkestatistik).

Total number employed

Postal, telephone, TV, transport, utilities Education and research Hospital, social and health care services General administration, armed forces, police and legal services Other public services

Central government

2.1 1.3

1.2

1.8 3

Percentage of total

0

6 12

0

10 3

Population ratio

Table 10.6 Change in public employment by policy area and level of government, 1995–2000 (total number in 000s – not full-time equivalent)

258

The state at work, 1

police and legal services. On the other hand, the number of military personnel, as well as employment in postal services, telecommunications, transportation and public utilities is stagnating or even declining. However, the growth in other fields of the welfare state is so strong that the ratio between the population and the public sector employed decreased from nine in 1975 to six in 1995. The strong growth of the public sector is reduced in the period from 1995 to 2000. The number of public employees still increases, especially due to continued growth in social and health care services, but the ratio of public employees to the total population was constant in the late 1990s. Restructuring of the public sector rather than public employment growth is the main feature of this period. The corporatization of the public sector continues in the postal, telecommunications, and infrastructural sector with the corporatization of the postal service in 1995 as a major example of this development. On the other hand, in education and health care services, this trend is reversed and public employees from the supported sector/public corporations are transferred to ordinary government organizations.

THE PUBLIC FEMALE PART-TIME WORKFORCE Welfare state literature sometimes distinguishes between ‘warfare’ and ‘welfare’ states (for example, Wilensky 1975), the former characterized by a relatively high military share of the (often low) total number of public employment, and the latter characterized by a high fraction of civilian public employment in social, health and educational services. The Danish public sector has never been a ‘warfare’ state in modern times, not even in the early part of the 20th century when social, health care and educational public personnel were counted in more modest numbers. Thus, even in the early post-war period, women dominated the Danish public sector. However, while the traditional and male-dominated public services (infrastructure, military and law enforcement) since have stagnated in absolute numbers, the social, health care and educational services have been growing rapidly. This change of public sector focus in terms of policy areas has further skewed the gender profile of public sector employment. Once again, it is impossible to establish a longer time-series of the gender profile of public employment. The 1960 and 1970 estimates are based on population censuses, the 1980 estimate is from the register-based employment statistics, and the 1980 and 1990 estimates originate from yet other public sector employment statistics. However, with some overlap in time between the different statistical sources it is possible to establish a

259

The Danish public service

Table 10.7

Female Male Total

Gender profile of total public sector employment, 1960–2000 (%) 1960

1970

1980

1980*

1990*

1990

1995

2000

60 40

62 38

63 37

61 39

64 36

66 34

65 35

68 32

100

100

100

100

100

100

100

100

Note: * In 1995, the definition of part-time changed from ‘less than 37 hours per week’ to ‘less than 32 hours per week’. Using the new definition, the percentages of female employees in 1980 and 1990 were 61 per cent and 64 per cent respectively; using the old definition, the figures were 63 per cent and 66 per cent respectively. The female percentage is lower under the new definition because more females than males work between 32 and 37 hours per week. Source:

Statistisk Aarbog.

fairly coherent overview of the gender profile of the public sector. As displayed in Table 10.7, by 1960 female employment had already surpassed male employment in the public sector, and the female share of total employment slightly increases in the following decades. Thus, in total, at the turn of the century more than two-thirds of the people employed are women. The rising female employment in the public sector has also had an impact on the distribution between full-time and part-time jobs. Female labour market participation is nearly at the same level as the participation rate of the male population. Nevertheless, part-time and full-time jobs are also unevenly distributed in the public sector, as is seen in Table 10.8. While three-quarters of the male employees are working full-time, only about half of the female employees are working full-time. And while the full-time share of male employment has been stable, the part-time employment of women has slightly increased. Hence, the high female rate of labour market participation is increasingly based on a part-time job, especially in the public sector. Combined with the overall female domination of public sector employment, we are approaching a situation with an even number of part-time and full-time jobs in the public sector.

PERSONNEL POLICY Traditionally the majority of public employed had civil service status. Not only in the core of the central government, the ministries, were employees hired on a civil servant basis. This procedure also applied to other groups

260

The state at work, 1

Table 10.8 Female and male part-time and full-time employment – share of total employment, 1980 and 1995 1980

1995

Full-time (%)

Part-time (%)

Total (000s)

Full-time (%)

Part-time (%)

Total (000s)

Female Male

51 78

49 22

481 308

50 78

50 22

502 278

Total workforce

62

38

789

60

40

780

Source:

Statistisk Aarbog.

of public employees, for example postmen, policemen, railway workers, schoolteachers and university professors. During the 1950s this situation began to change. The economic situation improved and the labour shortage increased. For public employers, hiring the manpower needed became increasingly difficult. The civil service pay and pension system was increasingly seen as part of a recruitment problem. Some of the attractions of the civil servant status were no longer confined to this group. Automatic inflation adjustment of salaries, which was a distinctive feature of the civil servant payroll system, became more widespread. Also in terms of length of holidays and pay during sickness, civil servants’ privileged position compared to the rest of the labour market had changed. Above all the job security of the civil servant status was no longer a strong argument in a full employment situation. In this light, the Civil Service Act was amended in 1958. However, the amendment was too insignificant to make a real difference, especially for the most attractive labour force (Betænkning 483, 1965). Instead, public employers started to hire labour on individual contracts with conditions comparable to those of employees in the private sector. These new public employees fought to obtain a collective agreement and in 1962, the first collective agreement covering central government civil servants with university degrees was signed (Bruun 2000). It became the general rule that junior staff obtained employee status with terms defined in collective agreements while personnel with managerial responsibility were employed as civil servants, but with their rights also negotiated with their unions. Later on, other changes were introduced. Even if top civil servants and other managerial staff have kept their civil servant status, there is now a clear distinction between their rights and obligations regulated by law and their salaries regulated by collective agreement. Furthermore, a number of

The Danish public service

261

public personnel groups have switched to employee status on the basis of collective agreements (for example, schoolteachers, high-school teachers and university professors). Finally, privatization in the field of infrastructure (for example, telecommunications, buses and ferries) has reduced the number of civil servants. In total, while civil service status was the dominant practice of public employment until the 1960s, nowadays 40 per cent of central government and only 15 per cent of local government employees are civil servants. The difference between civil service status and public employment in accordance with the collective agreement is in many ways quite insignificant. Employees working under collective agreements have obtained pension schemes increasingly matching those of the civil servants. For public personnel with a university degree, the pension schemes have been of similar value for a long time. The job security of civil servants – which contrary to popular wisdom is not and never has been a right to hold a given position, but rather a right to payment and pension in case of dismissal – is in reality not different from the job security enjoyed by other public employees. Dismissals of civil servants and employees working under the collective agreement are quite rare. The general personnel policy of avoiding dismissal in the public sector, a policy strongly supported by public sector unions and rarely questioned by public employers, applies to all job categories of public personnel. If there is a difference, higher civil servants are now more frequently than earlier leaving their position before the official retirement age. Civil service status implies that employees have no right to strike and that the employer renounces the right to lockout. Employees hired in accordance with the collective agreement preserve the right to take industrial action. But only in a few instances has this difference in working conditions had any impact. From time to time, at the expiration of the collective agreement, junior hospital doctors and nurses have lined up to strike only to see the parliament intervene and dictate the contents of the collective agreement by law, often even before the conflict has got off the ground. In the few situations of an actual strike, the remaining civil servants, for example consultants and chief nurses, have established emergency measures for the patients requiring most care. However, the emergency measures are sustainable only for a short while and no government has accepted longer strikes. In effect, personnel performing vital functions are not allowed to strike, regardless of their formal rights. In the late 1970s, despite their status as civil servants at that time, schoolteachers went on strike although they rephrased the incident as having a ‘professional meeting’. The striking schoolteachers were not subsequently sanctioned according to the civil service disciplinary system that includes

262

The state at work, 1

an arsenal of sanctions such as warnings, fines, displacements, downgrading and in grave situations, dismissal. Ironically the civil servant schoolteachers arrogated to themselves a right to strike, which they did not have, while junior doctors and nurses have seldom been allowed to execute their right to strike, which they do have according to their collective agreements. Despite the bilateral nature of a collective agreement, parliament has reserved the right to intervene and to dictate the contents of the agreement in times of conflict in the public and private labour markets. However, ‘dictate’ has normally meant a compromise worked out by a conciliation board. Originally, being regulated by law, parliament could, in principle, dictate the contents of the civil servants’ working conditions. However, parliament gave up this right as early as 1919 (Hansen 1980). Since then the civil servants and their organizations have been acknowledged as negotiation counterparts rather than as subordinate receivers of orders. However, it was not until 1969 when the civil service law was revised, that this right to frame the civil servants’ working conditions unilaterally was also changed in legal terms to a bilateral negotiation relationship. In a narrow perspective the 1969 revision of the Civil Service Act was motivated by the introduction of the universal old-age pension. The public sector recruitment problems however, had not diminished during the booming economy of the 1960s. The scope of the revision of the civil service law was enlarged to meet a politically felt need to increase flexibility within the pay system. Notably, a joint employer–union system for salaryreassessment of public positions was developed. Yet rewards of individual employee qualifications and work effort were beyond the scope of the 1969revision. In effect, the individual employee, whether holding the status of civil servant or employed in accordance with the collective agreement, could only increase to a minimal extent his or her own salary (Finansministeriet 1988). Based on an agreement between the public employer and public sector unions, this situation was modestly changed in the late 1980s. This agreement allowed a small-scale reward of individual effort. Less than 1 per cent of total pay has been earmarked for this purpose. The collective agreement of 1997 has taken this development a step further by the introduction of a ‘New Wage’ built on the same basic idea of rewards for personal effort and qualifications (Finansministeriet 1996, 1998c). By the millennium the ‘New Wage’ system is almost fully implemented. However, in order to accommodate public sector unions and employees, the system has been designed to prevent any public employee from losing income. Overall, the changes in the pay-system have to some extent decentralized the issue of settlement of payment to each ministry, local government organizations and subsections within these bodies. However, it is questionable to what extent the settlement of payment has been decentralized

The Danish public service

263

all the way down to the individual level. A more decentralized, flexible pay system has chiefly been an employer desideratum. The public sector unions have in all cases agreed to the changes in the pay system, although they basically dislike the idea of individual and decentralized payment. The unions’ agreement has been quite reluctant and they have taken actions accordingly. The public sector unions have agreed to the introduction of flexible payment as a second best but negotiable way of obtaining higher payment for their members. The public sector unions’ agreement has been conditional on their participating in the implementation of wage reforms. As for what is left of decentralized pay, the public sector unions have calculated that their influence at the individual public workplace is strong enough to compromise the idea of genuine individual rewards. We do not have any solid facts at the present time, but the unions’ calculations may not prove to be incorrect. In general, public employees and their unions have a strong say in the management of the Danish public sector. The employee and union influence is in many ways formalized. Implementation of wages is only one of several examples. Co-determination boards with equal representation of employer and employees are other prominent examples. All management decisions regarding economic disposition, personnel policy and so on are discussed and negotiated in the co-determination board. The public manager can formally override the conclusions of the co-determination board discussions, but most often he does not. Consent of the employees is a strong management card in the Danish public sector. It fosters harmony in the day-to-day working relationship. Furthermore, it may be a crucial card to hold if the public organization is in the full glare of publicity or otherwise in the full scrutiny of politics. The autonomy of Danish public sector organizations in personnel policy is circumscribed by public sector union influence. Personnel policy, of course, is also circumscribed by general legislation (for example the Civil Service Act) and the general agreement concluded by the public sector unions and the Ministry of Finance on behalf of the entire central government. Local government organizations have a parallel system of centralized negotiations with public sector unions. This means that the very same collective agreement is applied in every municipality and county. Central government organizations are also circumscribed by internal public sector economic management guidelines issued by the Ministry of Finance (again local governments have parallel systems, although in this respect less centralized than the collective agreement system). Of course the general guidelines have implications for personnel policy. Central assessment of the need for new employees and quantitative restrictions on the number of employees was in force in the 1960s and 1970s. In the 1980s,

264

The state at work, 1

these restrictions were eased. As part of a reform of the central government budgetary system in the middle of the 1980s, central control of the number of subordinate personnel was abolished. Instead a system of total frames for full-time employee equivalents and total pay was installed. Within these frames and restrictions the ministries were allowed to compound their own individual personnel structure and decide their own personnel policy (Finansministeriet 1996).

PUBLIC EMPLOYMENT AND PUBLIC SECTOR ANALYSIS The above discussion reveals serious problems with the quality of data on Danish public sector employment. Especially for data covering the early period – 1949 to 1971–77 – the data are deficient in at least two important ways. First, they are not collected in a way that allows their inclusion in longitudinal analysis. Second, they are infected by several inconsistencies arousing doubt about their validity. However, for the later part of the period, starting in the middle of the 1970s, comparable and valid data seem available on Danish public sector employment. Despite the flaws in the data, a number of strong and noteworthy developments are observable. First of all, the growth in public sector employment has been immense. Second, this growth has been concentrated in local government organizations and the policy areas where local governments are in charge, for example social services, health care services and education. Before the strong public sector growth even took off in the 1960s, the Danish public sector was dominated by female employees. The growth in the local government welfare areas has accentuated the female domination of the public sector even further. However, an increasing proportion of the public sector employed women are working on a part-time basis. In a way, public sector personnel policy is centralized. Pay and working conditions for an increasing number of public employees are settled by general collective agreements concluded with the powerful public sector unions. The pay and working conditions of the employees covered by collective agreements are fairly similar to those of the decreasing number of public employees employed as civil servants. Only a few benefits and disadvantages connected with the civil servant status still remain. In another way, public sector personnel policy is quite decentralized. Within the restrictions of the legislation, collective agreements and internal public economic guidelines, each public organization, for instance each ministry within the central government, frames their own personnel policy.

The Danish public service

265

METHODOLOGICAL APPENDIX ON DATA SOURCES AND PROBLEMS The methods used for collecting statistical data have changed considerably over the years and the reliability of the methods used is highly disputable. The specifications in the early part of the period were highly deficient. They did not allow a specification of public sector employment in municipalities and counties, and they did not report any figures for public enterprises or for institutions with government subsidies as an important part of their income. Over time, the methods have improved considerably. Since 1978, information on public sector employment is based on the fairly reliable data of employers’ payment of premiums to the ATP (an obligatory pension fund covering the entire labour market). These payments imply that the fund runs a virtual employment register. For more detailed analysis of central government employment, the 1985 budgetary reform brought about important improvements in data quality. This allows for a precise calculation of the number of full-time equivalents and detailed information on central government employment in different categories (for instance grades, types of organization) as well as information on their salaries (Finansministeriet 1998a, 1998b). Third, as a consequence of the ongoing reforms of the public sector salary systems, data on the actual variation in salaries for certain categories of government employees have been reported since 1979. The Danish categorization of policy areas is in some cases more fine grained and in other cases less fine grained than the German and American categorization (Derlien and Peters 1998, section 4). For example in Denmark ‘police’ and ‘legal service’ are not reported separately. On the other hand, hospital service is separated from other (primary) health care services, but primary health care is included in the category of ‘social and health care services’. In addition, postal, telecom and public transport services are reported in one category. The Danish categorization of ‘administration’ has changed over time. Currently, general administration includes central administration, planning and administrative employees in specific policy areas. However, administrative personnel working in the public institutions are included in the specific policy area. Hence, administration is only to some extent separated from the policy area in which it occurs. public sector employment statistics based on payroll statistics were initiated in 1975. The payroll statistics contain information about public employment in central and local government, policy areas and part/fulltime employment. Still, the policy area statistics lack information on

266

The state at work, 1

employment by ministry for central government employees. Furthermore, in a central–local government comparative perspective, the statistics suffer from the problem that full-time and part-time employment is differently defined at central and local government levels. The sample-based estimate of public sector employment continued until 1979. A comparison of the two types of statistics reveals that when applying the sample-based methods, the assessment of public sector employment is somewhat higher than when the pay-register method is applied. What is worse, the two methods do not produce the same size of differences across policy areas. Therefore, despite the overlap between the two types of statistics, it is not possible to assess total public sector employment (number of employed) by policy area over the timespan from 1950 to 1995. Finally, employment in the private but publicly subsidized sector was not assessed prior to the 1970s. In effect, the conditions for evaluating public sector employment by major policy areas are quite different before and after 1975. This, however, does not imply that the post-1975 statistics are without problems. One major shift has taken place in the handling of part-time employees. From 1975 to 1979 the actual number of working hours converted parttime employment to full-time equivalents. Later on each part-time employee was converted to a half full-time employee. It is possible to control for the changes in the assessment of the full-time equivalent. A major problem is the change of categorization at different points in time. This is especially problematic in the fields of social services, health care and hospital services. Primary health care was included in the category of ‘social and health care services’ from 1975 to 1979. From 1980 to 1988, health care services and hospital services were placed in the same column. In 1988, health care services returned to the common category of ‘social and health care services’. Categorization of housing, environmental protection, roads and urban renewal was changed several times in the 1970s and at the beginning of the 1980s. All these changes influence the reliability of the category ‘other public services’. Therefore, it is impossible to present a consistent pattern of public sector employment in the major policy areas for the entire period from 1975 to 1995. However, as the categorization is changed and then changed back again, the 1975 and 1995 employment data are fairly comparable provided that the change of the definition of administrative personnel is handled as well. About 30 000 public employees were technically moved from the policy areas to general administration in 1994. The major part of these employees was formerly included in the ‘social and health care’ category. In order to obtain comparable 1975 and 1995 figures, this change has been corrected in the 1975 figures as well.

The Danish public service

267

After 1995, the definition of ‘part-time’ changed significantly from less than 37 hours per week to less than 32 hours per week. This coincides with the requirement of 32 hours per week in order to be ‘full-time’ unemployment insured. It does, however, reduce the overall comparability of the preand post-1995 public employment figures. As one of the consequences, the latest development in ‘Female and male part-time and full-time employment’ is not reported (Table 10.8).

REFERENCES Betænkning nr. 483, Tjenestemandskommissionen af 1965, Copenhagen: Statens Trykningskontor. Bruun, Finn (2000), ‘Fra embede til arbejde’, in Peter Bogason (ed.) Dansk forvaltningshistorie, Stat, forvaltning og samfund efter 1950, Copenhagen: Jurist- og Økonomforbundets Forlag, pp. 194–220. Derlien, Hans-Ulrich and B. Guy Peters (1998), Who Works for Government and What Do They Do?, Verwaltungswissenschaftliche Beiträge, 32, Bamberg: University of Bamberg. Finansministeriet (1988), 90’ernes aftaler og overenskomster, betænkning 1150, Copenhagen: Statens Information. Finansministeriet (1996), Budgetredegørelse, tillæg om styringsformer, Copenhagen: Schultz. Finansministeriet og Centralorgnisationernes Fællesudvalg (1996), Lønsystemer, Copenhagen: Finansministeriet. Finansministeriet (1998), Budgetredegørelse 1998. Appendix, Copenhagen: Schultz. Finansministeriet/Økonomistyrelsen (1998b), Statens personaleforbrug, Copenhagen: Finansministeriet. Finansministeriet (1998c), Lønpolitik, Copenhagen: Finansministeriet. Hansen, Svend Aage (1980), 100 år i statens tjeneste, Copenhagen: Laane- og Sparekassen for offentligt ansatte. Økonomistyrelsen (1998), Lønstatistik 1997, http://www.oes-cs.dk/publikationer/ lstatistik/1997. Wilensky, Harold L. (1975), The Welfare State and Equality, Berkeley and Los Angeles: University of California Press.

11.

The welfare state is female: trends in public sector employment in Sweden Jon Pierre

MEASURING PUBLIC EMPLOYMENT IN SWEDEN This chapter provides a fairly broad-brushed image of recent changes in public sector employment in Sweden. Public employment in Sweden is measured using several different types of data. The chapter therefore discusses developments in Swedish public employment in the context of general issues and problems associated with measuring it. These problems are discussed extensively elsewhere in this volume. Therefore, this chapter will only point at a couple of issues which seem to be particularly salient in the Swedish case. The public sector in Sweden, particularly those institutions which deliver welfare state services, developed an almost worldwide reputation during the first three decades after the Second World War. It should be noted, however, that a large part of the ‘welfare state’ were redistributive programmes, that is a progressive tax system coupled with income-related public transfers. In addition to these programmes there are several transfers which are not related to income but rather, universal. What is more, and the most relevant sector of the Swedish welfare state in the present context, the welfare state developed extensive services in areas such as medical care, care of the elderly and daycare centres for preschool-age children. These sectors are fairly labour-intensive hence it is here that we find a large part of the explanation to the previous growth in public sector employment in Sweden. By the same token, the cutbacks that have been implemented across the welfare state during the last decade has curbed public sector employment. Thus the total number of public employees in 1992 was lower than that of 1991; this was the first time in post-war Sweden that this figure decreased. Many of the general problems associated with measuring public employment are present in the case of Sweden. One problem, arguably the most complex issue, is what could be called the institutional dimension of the 268

Public sector employment in Sweden

269

problematic, that is defining exactly what is ‘public’ about an organization. While there is a consensus on what characterizes the core of the public sector (such as departments or ministries and parliaments), there exist a myriad of organizations which are located on, or very close to, the public– private border. Also, there is the perennial discussion about how to define organizations that might not necessarily be strictly public but which nonetheless operate under public auspices or which deliver public services. In the present analysis we adopt what we consider is both an efficient and valid definition of public employment, namely the one used by Statistics Sweden (SCB) in their statistical material. While this is a formal, legalistic definition of public employment which may not capture some of the more subtle aspects of the public sector, it does offer both the ‘official version’ of public employment, consistency over time and, realistically speaking, the best possible measurement of public employment. In addition, this chapter draws on two different types of data on public employment, as will be discussed in some detail below; on the one hand we use ‘hard’ data, for example, figures on the number of employees and, on the other hand we use panel data which allows us to look more closely at different types of such employment. Together these two data sets present a picture of public employment which is as valid and reliable as is humanly possible. It should be borne in mind, however, that public employment – however measured – is an inherently slow and inert indicator of changing political priorities. Longitudinal data on public employment tell us a great deal about changes in scope of government and the priorities of its commitments, but it does so with some built-in ‘lag’. Given the tenure system, public employment is not a very useful indicator of short-term ideological or partisan differences in terms of public spending on different sectors. In many countries a large percentage of public employees enjoy considerable security in their employment, something which may make employment statistics a less accurate indicator of policy priorities than, for instance, changes in public expenditure in different policy sectors. That having been said, the institutional dimension says a lot about the political priorities of the state in the longer term; the changing division of labour and responsibilities between different levels of government; and also about the overall, long-term policy orientation of the political elite. Another dimension of public employment is more functional. The functional aspect is focused on developments over time in different types of generically defined professions and jobs. It measures the relative share of people in any given category which are conducting these functions under public and other auspices. A functional approach to public employment seeks to highlight the degree of professional specificity within the public sector but also, when panel or survey data is available (see below), on

270

The state at work, 1

similarities between the public and private sector with regard to, for instance, research and development. Obviously this approach raises slightly different types of research questions than the institutional approach and, indeed, presupposes that the institutional ambiguities can be reasonably resolved. The functional approach also requires a different type of data than the institutional approach; since similar types of jobs can exist in the public as well as in the private sector, we would need survey data which map type of job as well as type of employer. Furthermore, this approach requires much more detailed empirical data than the aggregate data which is usually available. Again we can fairly easily identify a core of public employment functions. But there are a large number of organizations whose employees are extremely difficult to categorize in terms of their employment and, by the same token, a large number of jobs which are surprisingly similar in the public and private sectors. The functional approach brings out developments over time within the labour market tout court and thus helps explain patterns of public–private competition for skilled personnel in different sectors. Thus, the more similar a job in the public sector is to a corresponding job in the private sector, the lower the cost of transfer for the individual, and hence the greater the need for the public employer to provide conditions and terms which are sufficiently attractive to keep the individual under public employment. This also means that jobs in the public sector that do not have a clear equivalent in the private sector (in the Swedish case this would include some types of jobs in the medical care sector, the childcare sector, basic education, some sectors within the universities and so on) and do not have to offer the same conditions, as the employee’s cost of exit is much higher. This functional approach to public employment may help us understand differences in wage development between the public and private sectors, and also the difficulties in enhancing the status of employment in subnational government. Given the fact that some 75–80 per cent of subnational government employees are women, the functional perspective on public employment thus also brings out an important gender dimension. Obviously, which of these aspects of public employment is most appropriate depends on the specific research questions. For the most part, it appears as if both the institutional and functional dimensions need to be taken into consideration in order to produce a complete picture of public employment. The strength of the institutional dimension is that it gives a comprehensive map of changes in the public sector staff. The functional approach requires statistical material which may not be available in all national contexts but which yields an image of public employment as embedded in the larger, societal labour market with all that entails in terms of tenure, salary, and competitiveness.

271

Public sector employment in Sweden

Public Employment as a Dependent and Independent Variable One of the most obvious problems in public employment research is whether employment should be thought of as a dependent or an independent variable. If public employment is the phenomenon to be explained, then we need to specify which factors we believe account for public employment development, for instance public spending, saliency of different policy sectors, the ideological orientation of the incumbent elite, or ‘policy style’. Sceptics may argue that there is no single explanation to developments in public employment and that, as a result, such analyses frequently tend to assume the form of a provision of answers in search of questions. While there may be some validity to that critique, placing employment data in an analytical framework such as a cross-national analysis is a very meaningful project indeed even if there are no clear images of ‘independent variables’. We can also treat public employment as an explanatory variable although this research design is much less common than the previous approach. Employment data – especially when coupled with measures on public spending – are an excellent indicator on the labour intensity of any given policy sector which, in turn, help explain such things as problems with policy change or difficulties in introducing new management techniques. Along with measures on public spending, employment figures also tell us much about state development and transformation. These two clusters of variables are closely, but not perfectly related, see Figure 11.1. The emergence of the modern state could be thought of as a development from the upper left cell towards the bottom right cell; over time public services have become increasingly cost-intensive as well as labourintensive.1 The laissez-faire state – to the extent that it ever existed – was characterized by a small staff and a minimal budget. Contemporary states differ significantly with regard to the different political priorities they display and employment data serve as a good indicator on these differences. Relative labour intensity

Relative cost intensity

Figure 11.1

Low

High

Low

Judiciary

Social services

High

Social transfers

Education, Health

Cost- and labour-intensive policy sectors

272

The state at work, 1

METHODS AND MEASUREMENTS We have previously touched upon issues related to different ways of measuring public employment. The conventional approach draws on statistics reporting, in totals and collapsed into different policy sectors, changes in the number of employees. The other strategy is to let the analysis depart from data collected from a panel which is monitored over an extended time period in order to track the dynamics in the labour market. These two different approaches correspond closely but not perfectly to the previously discussed institutional and functional dimensions of public employment. ‘Hard’ Data ‘Hard’ data refers to statistical material provided by government itself on various aspects of its employees. Such data are usually very reliable, although they cannot tell us very much more about employment than the observation that a person is employed by a public institution. Thus, we cannot say very much about the specificity of the position held by the person, nor his or her education and how these two factors are related. We can also say very little about how long this person has held the public office, what he or she did prior to that occupation and if the private sector is a viable alternative source of employment. Thus, ‘hard’ data are highly accurate but cover only a few of the aspects of employment which are of interest. Panel Data An interesting and useful alternative to the data provided by government employment offices is panel data collected to trace long-term developments in the labour market. Sweden Statistics (SCB) has conducted panel studies (arbetskraftsundersökningar  AKU, or workforce studies) on labour market developments since the early 1960s. Studies were initially conducted on a quarterly basis but are now done monthly. The sample size is currently 17 000 but was previously 18 000. Each individual is included in the panel for eight months, that is one eighth of the panelists are replaced each month. Over time, Sweden Statistics has developed considerable expertise in the planning and execution of the studies. Panel data provide a much more detailed and comprehensive picture of the nature of employment compared to the traditional statistical employment material. For instance, panel studies enable us to identify similar job types across the public–private border and also between governments at different levels. We can also trace typical career paths through the public sector. That having been said, it must be remembered that these are

Public sector employment in Sweden

273

sample-based studies and although the error in the estimates is known – and is surprisingly small – they might not provide the same sense of security as hard data can offer. The conclusion here is that by using hard data and panel data side by side we generate an accurate and detailed account of long-term changes in public employment. Interpretation and Analysis Interpreting development of public employment is an exercise where the same data can lend support to radically different propositions. For instance, in Sweden the defence sector in 1976 had 52 900 employees accounting for 13.1 per cent of state employment. In 1976 Swedish defence had 38 600 employees or 16 per cent of the state’s workforce. Defence critics can argue that defence has grown from 13 to 16 per cent; an argument which can be countered by showing that defence employment has decreased by more than 14 000 employees. Differences in data interpretation caused by different political vantage points is not the end of the problems, however; we encounter similar problems as soon as we want to give some analytical meaning to the changes we observe. While almost all academic observers would agree that changes in relative sizes are the only meaningful basis for longitudinal or latitudinal comparison, practitioners probably count heads much more than they think in terms of percentages which place a particular sector in the context of the public sector as a whole. Thus, the academic understanding of public sector management to some extent suffers from the same problems as political assessments of changes in public employment. Panel data might have a slight edge over hard data in these respects since they provide a broader and more nuanced picture of developments within the labour market. But the basic problem of interpreting sector development in relationship to the larger context of the public sector tout court remains largely the same.

THE SWEDISH CASE IN CONTEXT As was mentioned earlier, the Swedish welfare state is – to the extent that the present tense is still valid – not just a matter of redistributive policies and comprehensive welfare programmes; it is also a labour-intensive political project. Childcare, care of the elderly, and social work, to name but a few sectors of the welfare state, grew rapidly between the mid-1960s and up through the mid-1980s. Since then there has been a gradual contraction of the public labour force in many of these service sectors.

274

The state at work, 1

Historically, the state has also played an important role as employer in the context of accommodating fluctuations in the economy. In recessions, the public sector has been able to offer temporary employment for laid-off labour from the private sector. This was particularly true up until 1980 or so. After that public employment at different levels of government was cut back (see Table 11.1). Overall Developments in Public Employment Looking first at the grand totals, it is interesting to note that public sector employment in 1996 is very close to the 1976 figure and that the peak – located exactly in the middle of the timespan between these two data points (1986) – shows an employment figure which is almost 25 per cent higher than the 1976/1996 levels. Thus, public employment has described a fairly dramatic development over these two decades. Turning to developments at different tiers of government, the regional level seems to correspond closest to the overall pattern of development. This level of government, in 1995, was only 4 per cent smaller in terms of staff than it was in 1976. Meanwhile central government has decreased from about a third in terms of total public sector employment to about 20 per cent. Local government displays the most dramatic development; a virtually uninterrupted growth from 40 to 60 per cent of total public employment in the mid-1990s. The Swedish development in these respects must be understood against the severe fiscal crisis that hit the Swedish state by the late 1970s and worsened into the 1980s. The crisis first became apparent at the central government level; regional and local governments were doing financially reasonably well throughout the 1980s. Central government could report a reduction in staff in 1983. But it was not until 1990 that total public employment began to decrease, and public employment on all three levels did not show a reduction until 1993 when for the first time, local governments reported a net reduction in staff. These data, in turn, highlight the extensive decentralization – some would say ‘hiving off’ – of state responsibilities to subnational governments conducted during the 1980s. Thus, the state to a considerable extent managed its own crisis by relocating responsibilities to lower levels of government. As the financial crisis hit subnational governments in the 1990s we saw, for the first time, extensive cutbacks in staff on all three levels. Employment in Different Sectors A first quick glance at the developments in different policy sector conveys an image of a high degree of stability. Most of the changes that have taken place are more the result of structural changes within the public sector than

275

Source:

Note:

1995 1994 1993 1992 1991 1990 1985 1980 1976

246.1 249.4 309.5 362.4 387.3 401.3 423.1 429.1 404.6

N

State

19.5 19.6 22.6 24.8 25.5 26.0 27.7 29.9 32.7

% 238.0 274.6 294.1 312.6 382.6 394.0 418.8 360.0 283.3

N

Regional

18.9 21.5 21.5 21.4 25.1 25.6 27.5 25.0 22.8

% 757.6 734.6 746.5 768.4 731.1 723.4 660.0 623.1 514.7

N

Local

60.1 57.5 54.5 52.5 48.1 46.9 43.3 43.3 41.5

%

Employment at different levels of government, 1976–95 (000s)

1241.7 1258.6 1350.1 1443.4 1501.0 1518.7 1501.9 1412.2 1202.6

(N)

Total

18.0 18.3 18.6 19.0 19.9 23.0 22.8 26.1 37.4

N

1.4 1.4 1.4 1.3 1.3 1.5 1.5 1.8 3.0

%

Other public employment*

AKU, Statistics Sweden. Data are ‘hard’ statistics, not panel data.

* ‘Other public employment’ is not specified in the statistical material. However, this category includes employees of the Church.

Table 11.1

1259.7 1276.9 1368.7 1462.4 1520.9 1541.7 1524.7 1438.3 1240.0

Grand total

276

Table 11.2

The state at work, 1

State employment in different sectors, 1976–96 (000s, %) 1976

Courts Prisons Police service Defence Road agency Universities Labour market Regional administrationa Local tax authorities Postal services b Telecommunicationsb Railways Manufacturingc General administration Other state employmentd Total

1980

1985

1990

N

%

N

%

N

%

N

%

6.0 6.1 23.0 52.9 13.2 32.5 9.9

1.5 1.5 5.7 13.1 3.3 8.1 2.4

7.3 7.9 25.8 52.0 12.2 38.1 12.4

1.7 1.8 6.0 12.1 2.8 8.9 2.9

7.9 8.5 26.8 49.4 9.9 41.9 15.1

1.9 2.0 6.3 11.7 2.3 9.9 3.6

6.8 8.7 27.6 48.8 8.9 40.8 11.8

1.7 2.2 7.0 12.3 2.3 10.3 3.0

13.1

3.2

13.6

3.2

13.4

3.2

6.5

1.6

8.2 58.9 42.9 40.2 8.3

2.0 14.6 10.6 9.9 2.0

9.0 62.1 43.9 39.2 6.2

2.1 14.5 10.2 9.1 1.4

9.0 64.1 45.1 39.4 5.5

2.1 15.1 10.7 9.3 1.3

9.3 66.0 42.6 26.4 4.8

2.3 16.7 10.8 6.7 1.2

64.1

15.9

71.6

16.7

61.5

14.5

62.8

15.8

25.2

6.2

27.8

6.5

25.6

6.1

24.4

6.2

404.6

429.1

423.1

401.3

Notes: a. Regional administration refers to the state bodies at this level of government (länsstyrelser). b. Postal services (Posten) and telecommunication services (Televerket, later Telia) were transferred from the state sector into primarily state-owned companies in 1994 and 1993 respectively. c. Manufacturing refers to state-owned companies within the FFV (Förenade Fabriksverken) conglomerate. FFV was converted into a private company in 1991. d. ‘Other’ includes higher education in the agricultural sector (Lantbruksuniversitetet), the Riksdag and its authorities, and the agencies for hydro-electric power (Statens Vattenfallsverk), Forestry (Domänverket) and civil aviation (Luftfartsverket). Source: AKU, Statistics Sweden. Data is ‘hard’ statistics, not panel data.

of policy reassessments and shifts. Thus, postal services and the telecommunication sector accounted for some 25 per cent of state employment throughout most of the period covered in this chapter. In the early 1990s, these sectors were converted into private companies, albeit with significant public ownership. Similar structural changes have taken place in the railroad sector where railroad construction and maintenance was split off into a largely state-owned company in the early 1990s.

277

Public sector employment in Sweden

1991

1992

1994

1995

1996

N

%

N

%

N

%

N

%

N

%

6.5 8.0 28.0 46.1 8.4 43.1 11.7

1.8 2.2 7.7 12.7 2.3 11.9 3.2

6.5 8.1 28.1 44.5 9.0 44.3 12.2

2.1 2.6 9.1 14.4 2.9 14.3 3.9

6.4 8.4 27.3 41.0 8.3 42.8 12.4

2.6 3.4 11.0 16.5 3.3 17.2 5.0

6.3 8.3 25.4 39.1 8.6 45.2 12.6

2.6 3.4 10.4 15.9 3.5 18.4 5.1

5.9 7.9 24.4 38.6 8.0 45.7 11.6

2.5 3.3 10.1 16.0 3.3 19.0 4.8

1.8

6.7

1.8

6.2

2.0

6.1

2.4

5.8

2.3

5.3

2.2

2.9 0.7 64.0 16.5 40.4 10.4 22.1 5.7 – –

3.1 59.1 36.0 19.1 –

0.9 16.3 9.9 5.3 –

3.1 51.5 0.6 17.0 –

1.0 16.7 0.2 5.5 –

3.1 – – 16.4 –

1.2 – – 6.6 –

3.1 – – 16.0 –

1.3 – – 6.5 –

3.0 – – 15.7 –

1.2 – – 6.5 –

62.5 16.1

63.3

17.5

56.1

18.2

55.0

22.1

53.5

21.7

52.5

21.8

23.2

6.4

22.0

7.1

22.0

8.8

22.0

8.9

21.8

9.1

N

%

1993

6.7 1.7 8.4 2.2 27.8 7.2 47.7 12.3 8.7 2.2 42.3 10.9 11.4 2.9 7.1

35.3 387.3

9.1

362.4

309.5

249.4

246.1

240.6

Thus, cutbacks in public employment seem to have been more a matter of relocating functions and services outside the public sector than reassessing political commitments and priorities and then implementing these changes. One might interpret the developments displayed in Table 11.2 as illustrations of the state resorting to some of its core functions, abandoning more optional political projects. Police, defence, and particularly higher education have all grown whereas service sectors such as postal service, telecommunications and railroad services have been ‘hived off’. To be sure, this return to the core of state functions has been a leitmotiv in much of the administrative reform in Sweden during the 1990s (Pierre 1993; Pierre 2007; Premfors 1999). Public Employment and Gender Eight out of every ten local and regional government employees are women (Table 11.3). A very large portion of these women hold low-pay, low-skill

278

Table 11.3

1997 1996 1995 1994 1993 1992 1991 1990 1985 1980 1976

The state at work, 1

Female employment in central and subnational government, 1976–97 National government (000s)

Women (%)

Subnational government (000s)

Women (%)

222 227 233 246 323 368 387 394 424 437 415

42 41 40 40 43 43 43 43 40 39 36

1072 1111 1135 1147 1176 1225 1254 1258 1212 1072 903

79 79 79 79 79 79 79 79 77 77 75

Source: AKU; panel data. Techniques for estimates were modified in 1986 which slightly distorts comparisons over time.

positions in the medical care sector, in childcare, or in care of the elderly. Another large group of women hold administrative positions in local and regional government. Feminists debate whether women-intensive sectors tend to become lowwage sectors or if it is rather the case that women, for a variety of reasons, are confined to look for work in less prestigious sectors. While this is a debate which the present author does not want to enter, it could be argued that wage development tends to be slower in sectors which are typically dominated by a female workforce; a large number of studies substantiate the systematic wage differences between men and women with regard to salary. It seems clear that with some 80 per cent of the workforce being female, these differences become institutionalized in the system of employment. The longer-term consequences of that may well be problems in recruiting personnel, particularly male staff. It may also affect the overall status of public employment. A key aspect of the gender issue in the context of public employment is the extent of part-time employment. Table 11.4 presents data on part-time employment by gender and levels of government. Table 11.4 shows that male part-time employment has remained remarkably stable from the late 1980s through the 1990s. Female part-time employment, however, has decreased significantly. In the late 1980s close to 40 per

Public sector employment in Sweden

279

Table 11.4 Part-time employment in central and subnational government, 1987–99 (%) Central government

1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 1988 1987

Subnational government

Male

Female

Male

Female

6 7 7 8 6 7 7 7 6 6 6 6 6

19 19 19 20 23 24 27 28 28 28 31 34 36

12 13 14 13 14 13 8 12 11 12 12 12 12

40 41 42 43 44 45 45 44 45 45 46 48 49

Notes: Part-time employment is defined as working 34 hours or less per week. Subnational government includes local and regional government. Source: AKU; adapted from panel data.

cent of female central government employees worked part-time; a decade later that figure is around 20 per cent. There has also been a clear, if not as impressive, decrease in female part-time employment in subnational government, from about 50 to 40 per cent. There are several contributing explanations to these developments. One is that women have to a growing extent embarked on career-track positions within the public sector; although the majority of senior civil servants are still men, the percentage of women at the higher levels of the bureaucracy is slowly increasing. Another explanation to the decreasing female part-time employment could be that the budgetary crisis during the 1980s and 1990s was an impetus for women to secure their employment in the public sector and taking up full employment could be a strategic move to that end.

REGIONAL AND LOCAL GOVERNMENT The subnational levels of government have historically been accorded important roles in the Swedish political system. The notion of local

280

The state at work, 1

autonomy has been heralded by all political parties and is written in the Constitution. By international comparison, the definition of local autonomy is generous. Unlike the British system, where local governments are only allowed to do what Parliament explicitly states that they may do, Swedish local government autonomy is the normative foundation and any restriction on this autonomy has to be stated by the Riksdag. Regional government enjoys the same level of autonomy. However, as we have seen earlier in this chapter, subnational government is the key provider of welfare state services. Indeed some 75–80 per cent of subnational government programmes see local authorities implementing central government policies for which they receive financial compensation through grants. Regional government is primarily about health care. This sector alone accounts for more than 80 per cent of regional government employment. If we add care of the mentally handicapped and other forms of care we find that more than 90 per cent of the regional government employees work in these sectors. Also, regional government is involved in some educational programmes and in the cultural sector (Table 11.5). Local government is also primarily concerned with the delivery of welfare state services but is less focused on one particular sector than Table 11.5 Regional government employment in different sectors, 1980–2000 (%) Health Social care care 2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1985 1980

90 90 89 89 88 84 78 80 81 83 85 85 86

3 3 3 3 2 2 1 3 2 1 1 1 1

Care for Education Other Central Total N mentally and administration (%) (000s) handicapped culture 0 0 0 0 1 5 13 10 11 11 10 9 8

3 4 5 5 5 5 5 4 3 3 2 2 2

2 2 2 2 2 2 2 2 2 1 1 2 2

2 1 1 1 2 1 1 1 1 1 1 1 1

100 100 100 100 100 99 100 100 100 100 100 100 100

260 260 256 251 247 274 306 319 340 424 435 411 353

Note: Data are based on number of employments, including full- and part-time employment. Source: National Association of Regional Authorities (Landstingsförbundet).

281

Public sector employment in Sweden

Table 11.6 Child care (%) 1997 1996 1995 1994 1993 1992 1991 1990

19 19 19 19 19 20 22 22

Local government employment in different sectors, 1990–97 Other Education Culture Technical forms (%) and staff of care leisure (%) (%) (%) 34 35 34 32 31 30 24 23

21 20 21 21 21 20 22 22

3 3 3 3 3 3 3 3

15 15 15 16 17 18 20 20

Administration N (%) (000s)

8 8 8 9 9 9 9 10

583 597 588 566 577 593 573 570

Notes: Data for 1997 are preliminary. Data are based on full-time equivalents. Source: National Association of Local Authorities (Svenska Kommunförbundet).

regional authorities (Table 11.6). Over the past couple of decades a number of service sectors have been decentralized to local government. In 1989, primary education was transferred from central to local government. Local authorities have also assumed responsibility for care of the elderly. This extensive decentralization explains why local government employment has increased during the financially troubled 1990s while central and regional government saw a decreasing number of employees. As a result of these institutional changes, the Swedish system of government has become increasingly two-tiered, with central and local government as the dominant levels of government. Regional government has always been rather ‘anonymous’ in Sweden; the average citizen know much less about regional government and politics than about central or local government. In the late 1990s, some institutional reforms were implemented in a couple of regions, largely in response to the EU structural funds’ emphasis on the regions.

CONCLUDING REMARKS This chapter has highlighted trends and developments in public employment in Sweden. The patterns displayed by the data – the changing institutional division of labour, the relative stability in most policy sectors except

282

The state at work, 1

for structural changes and the female labour dominance in regional and local government – are not very unique to Sweden, as other chapters in this volume substantiate. However, they do suggest that public employment is a more dynamic feature of the state than is often realized. A cross-national assessment will help tease out what is more unique about the Swedish developments and what should be the focus for further analysis. Public employment in Sweden cannot be properly understood without recognizing the gender dimension of the public sector. At the middle and lower levels of public authorities at all levels of government we see organizations with a very large percentage of female employees. At the subnational level, three out of every four employees are women. Given the financial constraints that the Swedish public sector became subjected to during the 1980s and 1990s, several factors affecting the status of public employment such as wage development and an image of contributing to a larger common good took a negative turn, something which probably strengthened the public sector’s image of a low-status, women-intensive workplace. Much of the societal prestige of the public sector has a direct linkage to female employment.

NOTE 1. The United States is the best exception to this pattern. There, much of education, especially higher education, is run by private institutions. Similarly, health is still primarily run by private sector organizations.

REFERENCES Pierre, Jon (1993), ‘Legitimacy, institutional change, and the politics of public administration in Sweden’, International Political Science Review, 14, 387–401. Pierre, Jon (2007), ‘From managing the state to managing contingencies: changing roles of the senior civil service in Sweden’, in Edward C. Page and Vincent Wright (eds), From the Active to the Enabling State: The Changing Role of Top Officials in European Nations (Basingstoke: Palgrave MacMillan), 208–23. Premfors, Rune (1999), ‘Organizationsförändringar och förvaltningspolitik – Sverige’ [Organizational change and the politics of public administration in Sweden], in Per Laegreid and Ove Kaj Pedersen (eds), Fra opbygning til ombygning i staten [From Construction to Reconstruction of the State], Copenhagen: Jurist – og Ökonomforbundets Forlag, pp. 145–68.

12.

Conclusion Hans-Ulrich Derlien

A condensed summary of the ten country reports could contain the following points: From the 1950s when data became available and the time series started, until the end of the millennium, a general expansion of public employment took place with the public sector more than doubling in most countries. This trend however was not a linear one. The growth era was followed by stagnation and even slight decreases. A second trend accompanying this development was the growth in part-time employment, and this in turn was very closely related to increased female employment. Fourth, in all of the countries, part-time and female employment were linked to the most important growth areas, specifically the tremendous expansion of the educational sector, followed by health care and in many countries, personal social services. The growth of public employment in these areas was bound to affect the proportions of employment between levels of government. There is a general trend for the vertical, intergovernmental distribution of public employment to become more bulky because these services are amassed for functional reasons at the regional and local levels of government. Another aspect of the vertical distribution of employment is a trend towards agencification in particular in those countries with a British administrative culture. Privatization and agencification associated with New Public Management inspired reforms. This configuration though is much more complicated and deserves some differentiation; the in-depth analysis is reserved for Volume II.

CHANGES IN PUBLIC SECTOR SIZE In broad brush terms, all countries included in this survey show strong growth in public sector employment during the 1960s and 1970s followed by a period of stagnation or even decrease. Variations emerge in the timing and impact of the reverse. Whereas during the period 1950–2000, public employment in most countries more than doubled (Australia, Canada, France, Germany, Spain, USA), less dramatic overall growth occurred in Denmark and Sweden, and the size of the public workforce in the UK 283

284

The state at work, 1

actually decreased by 11.5 per cent and to a lesser extent this also happened in New Zealand. Following the period of growth, the UK, New Zealand and Canada began to shrink as early as 1980. In Denmark and Sweden, the Nordic welfare states, the public services continued to grow albeit at modest rates (Sweden grew by 1.6 per cent during the period 1976–95). In particular, growth in the welfare services outpaced the growth in other parts of the public sector. In the 1980s and 1990s, the number of public employees in welfare services more or less stagnated, but in relative terms the growth continued due to reductions of public employment in other areas. Most especially, the strong wave of privatization or corporatization of public enterprises and utilities (telecommunications, transport, postal services) reduced the number of public employees in these former classical public services so that public employment in education, health care and social services continued to increase in relative terms. This development as well as the following observation will be analysed in detail by Jørgen Grønnegard Christensen and Thomas Pallesen in Volume II of the study.

POLICY AREAS AND POLICY PRIORITIES In all countries, public employment in welfare services comprises an increasing share of total public employment. This can be regarded as part of the general trend towards service economies (Scharpf and Schmidt 2000). In our sample, public employment in education, health care and social services amounts to half of total public employment in the United States, Germany, the United Kingdom, France and Australia, and it is even higher in the Scandinavian countries, New Zealand and Canada. Certain policy areas, however, reveal a reverse trend. This holds in particular for the military (soldiers and administrative personnel). Especially since the fall of the Iron Curtain, military capacities shrank, not only in states that turned their armies into professional armies, but also in Germany with her tradition of keeping an army with mandatory military service. Separate from this peace gain, public workforces have been reshaped further by privatizations. In our sample, privatization in the form of the sale of assets outright appears to have been favoured more by the Westminster-style governments than by the European countries where corporatization involving some substantial continuing public sector input has been the more common approach. Prompted by European Union initiatives, services and employment in formerly centralized and monopolized infra-structure areas such as telecommunications and railway services were removed from direct government control; employees were for legal reasons excluded from public statistics and no longer counted as part of the

Conclusion

285

public service even where the newly established private law companies were fully owned by governments. Further examples include the French postal service and France-Télécom restructuring in 1991. In France most companies nationalized in 1945 and 1981 were partially or totally privatized in the 1990s. It should be noted also that those areas of public service which have to do with regulation, paperwork and red tape (administrative work in a narrower sense) have not grown proportionately with increasing employment in the personal social services areas. Although it is popular to complain about bureaucratization, this has no foundation in the employment numbers of general administration.

CENTRE–PERIPHERY DISTRIBUTION OF PUBLIC EMPLOYMENT Of course, services have to be brought to the population but this does not imply that employment contracts are necessarily with local governments. For instance, teachers may still be employed by central government as is the case in France. In most countries, however, the increased importance of education and other social services has moved the bulk of the public labour force into regional levels. It is not growth as such but growth in these specific policy areas that stimulated the expansion of public workforces at sub-national levels. Statistically the trend became more pronounced as the privatized infrastructural policy areas normally run by national governments disappeared from national employment figures. At the same time, in all countries, the central government level was the main target for cutbacks or, as in France, the level that experienced the lowest rate of employment growth. The most extensive reductions in government employment occurred at national levels in Germany, the UK, New Zealand and Australia. In the decade 1990–2000 Germany more than halved the size of its federal government workforce, largely through privatizations, although the after-effects of unification also account for the apparently severe cutback. The sheer territorial expansion to the East to include an additional five regional governments had to affect the former balance and push the national government as the previously most important public employer into an almost marginal position. Similarly severe national-level cutbacks carried out by the three Westminster-style governments might be seen as showing the influence of NPM agendas. Another aspect of the vertical redistribution of employment is a trend towards agencification, again in particular in those countries with a British administrative culture. As with privatization, agencification is also inspired

286

The state at work, 1

by New Public Management ideas that favour the removal of implementation tasks from the ministries. This trend towards de-concentration however did not affect all the countries, for the administrative tradition in continental European countries was always such to keep the ministries at the national level small with their functions concentrated on policy advice and monitoring of implementation. In this area we therefore observe a wide range of variation between the states included in the study. While the UK slimmed down her national ministries, a similar trend could not be observed in Germany or the United States. Cutbacks were less severe at sub-national levels. Continuing responsibility for the delivery of large social programmes accounted for continued growth at most regional and/or local levels. In particular France recorded a growth of almost half again at the local government level, with expansion also in its health sector, albeit at a constrained rate. In the USA, during the period 1980–98, state and local level workforces expanded by more than a quarter again, led largely by increases in the education area at one or other sub-national level, depending on the particular state. This is typical for systems of government in which the bulk of the responsibility for the delivery of labour-intensive programmes – including police services – has traditionally been located at the state and more often the local level. Helen Nelson will deal with this dimension in Volume II of this study.

PART-TIME AND FEMALE EMPLOYMENT Widespread acceptance of part-time employment in the public sector emerged in the late 1960s. A primary reason for the adoption of part-time work was the necessity to deal with labour shortages and to encourage women to enter the labour market. In the European countries, part-time employment is still predominantly a phenomenon of working mothers with school-age children. The parallels in the emerging patterns of female and part-time employment are evident in comparative perspective. Both are complex phenomena, depending on an interaction of different economic, social and political factors. The high proportion of part-time employment in Denmark and Sweden from the late 1970s to the mid-1980s has since decreased, including in the public service. The New Zealand public service also shows a substantial decrease in its part-time quota. Part-time employment is still growing in Germany and in France. Australia and the UK have experienced major increases in part-time employment, from a medium to a high level. However, in none of the countries has the feminization of the public sector been stronger than in Sweden and Denmark where about two-thirds

Conclusion

287

of the employees are women; in the UK it is just below that figure. It is around 50 per cent in Canada, France, Germany, Spain, NZ and Australia, and below 50 per cent in the USA. In countries where women have a comparatively medium participation rate, the share of women and men in the public service tends to be equalized. A comparatively low participation rate of women in the labour market in Spain coincides with a similarly low proportion of women in the public service. While women’s employment and part-time employment are predominantly in the state welfare services areas such as health, education and social services, privatization and downsizing of traditional male domains have led to an overall decrease of men’s employment in the public service in several countries. As Silke Heinemann will show in Volume II of this study, despite these trends, the general rule holds: the higher the responsibility level, the greater the gender imbalance. Further, in countries with civil service status, women are less well represented in that status group than in others. The distribution of women among career groups shows a concentration in the lower levels. Despite improvement in women’s representation, senior positions are still male dominated. The extent to which women have so far entered top positions depends on a variety of factors, not only programmes to advance female careers but also the importance of female-oriented family policy including the provision of nursery schools in a country. Comparison is difficult, but there are strong indications that the women’s share in senior civil service positions is lowest in Germany. Regarding the impact of NPM on women’s employment, there is no consistent answer. Some writers highlight the negative constraints imposed by the business approach of NPM and the relative freedom from regulation. For the same reasons, others see positive opportunities. Since the 1970s, all governments surveyed here have formulated policies to promote gender equality regarding employment and occupation in the public service. Legislation guarding against sex discrimination and providing for equal pay and maternity protection has contributed to the abolition of discrimination in recruitment, pay and conditions. The state as an employer is able to act as a frontrunner in the development of both trends, but, as shown above, countries take different approaches to fulfilling this role. The Nordic countries represent the cutting edges in the promotion of equal opportunities for women and part-time work. Germany, France and Spain are in some ways the stragglers. In the promotion of minority groups in the public service, the example of New Zealand demonstrates that stragglers might be able to learn from the experiences of other countries.

288

The state at work, 1

ETHNIC AND LANGUAGE REPRESENTATIVENESS Representativeness, apart from the gender criterion, boils down to ethnicity and language in multi-ethnic societies and immigration societies such as the USA, Canada and Australia. Federal systems might provide cultural niches with sub-national territorial borders following cultural lines, such as in Switzerland and Belgium and now Spain, in particular for the traditional provinces. In general, three types of policy can be identified: antidiscrimination; equal opportunity, geared to providing minority groups the means by which to compete; and affirmative action, requiring some form of equality of result through various preferential schemes. Iain Gow and Sharon Sutherland will go into the details in Volume II of our study.

STRATIFICATION In all of the countries, the public service is stratified according to entry requirements, which are specified by the level of education deemed prerequisites for performing the jobs in particular categories of employment and careers within these categories. For instance, university education is universally regarded imperative for jobs in the higher public service and in particular, the higher ranks of the civil service entrusted with authoritative and policy-making functions. Clerical jobs as well as jobs in the policy areas where personal social services are produced by the state normally satisfy themselves with lower educational qualifications. When inspecting the internal differentiation of the public services we discerned in all countries a core and a periphery of the service, the latter being constituted by contractual public employees and labourers who work in public sector areas typical of the welfare state and that are most likely to undergo privatization. Civil servants are most likely to be found in authoritative state sectors, in particular in government ministries where their elite may even form a particular legal subgroup (Senior Executive Service, political appointees or political civil servants). This elite, in continental countries, still tends to be recruited from the middle and upper classes, educated as generalists in special university courses, advancing through an internal career to the small number of less than 2 per cent top positions at the age of 50, united with colleagues in a certain esprit de corps (basically vis-à-vis the rest of the service) – and the elite is predominantly male. Nevertheless, this extreme type of a classical career civil servant, which was institutionalized in all the countries presently under consideration by the last quarter of the 19th century, is a diminishing presence. Few administrative elites today correspond to this Weberian type; rather, economic expertise has

Conclusion

289

come to complement traditional forms of generalist training, outside recruitment (like inter-sectoral mobility in general) is increasing, easier access to university training has broadened the recruitment pool and women are gaining ground in the higher civil service. Thus, in various respects, a blurring of the boundaries of public services can be identified. Common features of countries were observed at a general level, for instance the universal fading of blue-collar workers and a tendency in many countries of upgrading within the civil service, resulting even in a thickening at the top accompanied by increasing wage differentiation. As Hans-Ulrich Derlien and Luc Rouban will elaborate in Volume II, irrespective of training and career patterns, the existence of grand corps or not, the availability of centres of elite formation or their absence, the higher civil service might still be most likely to form an esprit de corps (at least in certain branches). At the same time, the administrative elites are interlocked with the economic and political subsystems of national societies. The intensity of links to the realm of politics, though, depends decisively on whether countries belong to the Westminster system (the UK, New Zealand, Australia, Canada) or to continental European state societies. Only the latter (and the USA) allows inter-sectoral mobility from the public service into politics. Increasing personnel exchange with the economic sector however, going hand in glove with the transfer of an economist role understanding, seems to be a universal phenomenon.

NEW PUBLIC MANAGEMENT AND PERSONNEL ADMINISTRATION From our study there is qualified support for the claims of the United Kingdom, Australia and notably New Zealand, as New Public Management pioneering and pace-setting reformers. All three reduced the relative size of their public sectors during the 1990s. In the broader comparative perspective, however, the cutbacks served only to bring the relative size of their respective public sectors closer into line with continental and North American levels. Empirically, public sector employment in Germany and Spain fit better the NPM ideal of highly decentralized public services. In both cases however the forms of decentralization have little affinity with the proposals constituting the NPM repertoire and neither country is renowned as a NPM protagonist. If the degree of centralization is measured as the relative level of central government employment, the NPM reformers – the UK, Australia and New Zealand – remain more centralized than the other federal countries and also several of the unitary states in our sample. New Zealand is not only the most centralized of all the countries

290

The state at work, 1

surveyed but has moved even further in this direction during the post-war period. The three Anglo-Saxon countries’ allegiance to NPM philosophies becomes more apparent when we consider changes in personnel management, most specifically the devolution of personnel functions to subordinate levels of government (decentralization) and agencies (de-concentration) including a trend towards more regional or sector-specific salary policies and terms of employment. The country studies reveal formal changes in the terms of contract that are undoubtedly inspired by New Public Management, in particular principal–agent theorizing. While the civil service elite in countries such as France and Germany emerge apparently untouched during the period of the investigation, in other countries, private law contracts and limited periods of appointment of the administrative elite have gained ground, most prominently in New Zealand. However none of the countries that observe the bifurcation of the civil service on the one hand and the larger public employment on the other moved towards a unitary system of terms of employment, although Switzerland, which is not included in the study, has abandoned the traditional public service status and placed all the public service on private law contracts. Undoubtedly though there is some movement going on in top civil service ranks, be it that a senior executive service was established as in the USA in 1978, be it that reshuffling of the incumbents of elite positions was eased, be it that appointments to top positions were limited to a specified period of time, be it that elements of performance-related pay were introduced. These developments will be dealt with in depth in Volume II in the chapter written by Jørgen Grønnegard Christensen and Robert Gregory.

THE STATE WANING? When considering the trends and traits elaborated so far, there is no indication in the public services of a waning of the state. On the contrary, at the most, the core policy areas characteristic of the state in the classical departments of defence, justice, police, finance or foreign policy declined in importance in relative terms; welfare state functions however are growth areas all over the world and in all the country families. Our data reflect a couple of changes in employment by policy area and in this respect a change of the welfare state, following the lines however of the EspingAndersen typology. Nevertheless, Britain with her National Health Service is so different from the United States that it becomes difficult to talk of an Anglo-Saxon family. Education even in states with a high degree of private schools and universities has been an expanding sector. However, it might

Conclusion

291

be accurate to talk of retrenchment and redressing the boundaries between state and society when it comes to public expenditures in terms of transfers. Although unemployment benefits and social benefits may have been curtailed in many countries, there is no indication that this affected the level of public employment in these areas. Ultimately it does not make a difference whether you transfer 1000 or 10 000 euros or dollars or pounds; the administrative work employed in this, unless it is not done by computers anyway, would not be affected. Our data reveal in more detail the changing nature of states in terms of their primary activities beyond the monopoly of applying physical force. As to national autonomy and the strength of national or sub-national governments in governance networks, an ultimately qualitative change, which may or may not have taken place or if it took place, possibly did not affect all policy areas, our data cannot contribute to the discussion. Also the concept of governance does not shed new light on our research topic. Governance, as far as the term is specified at all, refers rather to policy-making and the rule-making or regulatory process, which is more or less dominated in the public sector by central bureaucracies under the purview of executive governments and variations of employment at the apex of the state are statistically insignificant. One would need to go into much more detail and change the perspective from the macroscopic view employed here to a microscopic investigation; then one may observe, for instance in the European Union, a mushrooming of liaison offices in the ministries for European policy coordination, and at this level it might even be adequate to talk of a merged or interpenetrated European bureaucracy (Wessels 2000).

REFERENCES Scharpf, Fritz W. and Vivien A. Schmidt (eds) (2000), Welfare and Work in the Open Economy. Volume I: From Vulnerability to Competitiveness; Volume II: Diverse Responses to Common Challenges, Oxford: Oxford University Press. Wessels, Wolfgang (2000), Die Öffnung des Staates. Modelle und Wirklichkeit grenzüberschreitender Verwaltungspraxis 1960–1995, Opladen: Leske + Budrich.

Index accountability 84, 86–7 administrative elite 98, 181–2, 187–8, 199–201, 216–17, 288–9 agencies 27, 32–3, 43–6, 50, 55–6, 86–90, 122–4, 147, 151–3, 178, 223 agencification 27, 285 Australia 76–105 Public Service Act (1999) 100 level of government 81–6 federal 81–6 local 83, 86, 90, 93 state 81–6 policy areas 90–94 Westminster model 76–8 autonomous cities, see regional government autonomous communities, see regional government blue-, white-collar employees 130, 154, 163–5 Canada 106–39 civil service commission 121, 123 federal crown corporations 115, 121–32; see also public enterprises French Canadians, francophone 106, 134–5 level of government 109–20 federal 109–15 local 118–9 provinces and territories 115–7 policy areas 120–25 Public Service Commission 123, 126 special operating agencies 122–3 career 225, 228, 234–35, 239–44, 246 career groups 171–2, 181, 185–7, 193 career service 76–8, 103, 133–4 career system 182 central government, see country entries centre-periphery distribution 285–6

civil service 10, 23–7, 32, 35, 37, 154, 165–6, 260–64 civil service unions 51–2, 84–6, 96–100, 103, 198–201, 210–11, 216–17, 224–6, 228, 234–5, 243, 246, 260–64 clerical 130 consultants 50–51, 71, 87–8 contestability 77, 87 contracting 38 contracting-out 78, 86–8 contractual agents 235, 240, 243–5 contractual employees 288 core departmental personnel 33 core public administration 180, 183–4 corporatization 258 corps system 197, 199–201, 228, 235, 238, 243 decentralization 6–7, 142, 149, 197, 115, 228, 230–33, 246, 250–53, 262–4, 274, 281 see also devolution deconcentration 285, 289 democratization, see devolution demographic trends 222, 246, 247 Denmark 249–67 Danish National Church 251 level of government 250–61 central 250–53 local 250–53 policy areas 253–8 departmental chief executives 42, 53–4 devolution 22–3, 27, 29–30, 196, 197, 203–8, 211, 213–5, 220 downsizing 122, 129–30 elite, administrative, see administrative elite employees 252–3, 259–62, 264

293

294

The state at work, 1

Equal Employment Opportunity 65, 68, 73, 100–102 ethnic and language representativeness 288 ethnic employment 100–102 see also race ethnic minorities 34, 36 European Union 197, 219–20 executive departments 149–54 federal system 76, 79, 81–3, 88–9, 93–4 female employment 286–7 Australia 89–90, 100–102 Canada 134–6 Denmark 249, 258–61, 264 France 244–6 Germany 173, 189, 193 New Zealand 65–8 Spain 217–9 Sweden 270, 277–9, 282 United Kingdom 34, 36, 38 USA 144–5, 164–7 feminization of public sector, see female employment France 222–48 départements 228, 231–2 EDF 225 grands corps 235, 238 level of government 228–32 state 228–9 local 228–9 policy areas 239–43 left-right 224–6, 228, 247 SNCF 225 universities 236–7, 240–41 gender 144–5, 164–7, 249, 258–61, 264, 270, 277, 278, 282 see also female employment general purpose authorities 28 General Schedule 154, 163–4 geographical distribution 120, 148–9 Germany 170–95 Chancellor’s office 178–9 level of government 170–72, 176–81, 185 federal 170–72, 176–81 Länder 170–72, 176–81 local 170–72, 176–81

policy areas 172, 174–81, 189–90 status groups 170–71, 181–5, 186 unification 170–74, 176, 183–4, 193, 285 governance, term 4–5 human resources management, see personnel management independent executive agencies, see agencies industrial nationalizations 20, 38 industrial relations, see workplace relations institutional reforms 277, 281 level of government, see respective country listings local government 23, 25, 28–9, 30, 38, 43, 68–70, 83, 86, 90, 93, 118–9, 141–2, 144, 146, 148–9, 154, 157, 170–72, 176–81, 196, 203–4, 211–12, 228–9, 250–53, 274–6, 279, 281–2 ministerial bureaucracy 171, 178 ministerial departments 31, 33 minority group employment 100–102 see also ethnic employment mobility 233, 235, 289 municipalities 228, 230–32, 250–53, 265 see also local government New Public Management (NPM) 5–6, 191, 246, 285, 289 see also privatization New Zealand 40–75 crown entities 42, 44–5, 49–50, 54, 55–6, 66, 71–2 Department of the Prime Minister and Cabinet 50–51, 56, 58 Employment Contracts Act 51–2 ethnicity, Maori (tangata whenua) 40, 45, 50–51, 66–8, 69, 73 Labour Party 40–42, 48, 55, 68, 72 level of government central 40, 43–4, 69 local 43, 68–70

Index Members of Parliament 40–42, 54, 55, 73–5 ‘Muldoonism’ 72 policy areas 54–65 Public Finance Act 43 Public Service Act 43, 46 ‘Rogernomics’ 55 State Owned Enterprises Act 43 State Sector Act 42–3, 49, 51, 56, 65, 68, 70 State Services Act 43 State Services Commission (SSC) 42, 44–5, 48, 56, 58, 63, 68, 72 State Services Commissioner 53–4 state-owned enterprises (SOE) 43, 45, 50 see also public enterprises treasury 43, 45, 50, 51, 56, 61, 63 Treaty of Waitangi 67, 73 OECD 9, 53, 71, 112 outsourcing 78 see also privatization part-time employment 286–7 Australia 88–90, 101–2 Canada 133–4 Denmark 256, 258–60, 264, 266–7 France 227–8, 243–4, 246 Germany 171, 173–4, 189, 193 New Zealand 66 Sweden 278–9 United Kingdom 36–7 USA 144–5 patronage (favouritism) 154, 161, 164, 197, 199, 217 pay systems 54, 97–9, 260, 262–4 personnel costs 171, 175–6, 181 personnel management 7, 77–8, 94–100, 199–201, 220, 243 personnel policy 259, 261, 263–4 policy areas 11–2, 284–5 Australia 90–94 Canada 120–25, 132 Denmark 253–8 social and health services 253–4, 258, 264–6

295

education 253–4, 258, 264–6 hospital services 253–4 France 239–43 education 226, 228–9, 234, 236–7, 240–41, 245–6 police 234, 239, 241–2 Germany 172, 174–81, 189–90 social service 174–6, 178, 189, 193 education 174–6, 178, 185, 189, 192–3 general administration 174–5, 178, 185, 189 police 174–6, 178, 185, 190, 192 defence (armed forces) 174–6, 178, 190, 192 New Zealand 54–65 Spain 212–6 Sweden 269, 271, 274, 276 defence 273, 27–7 police 276–7 health care 268, 271, 280 New Zealand 54–65 United Kingdom 33 USA 155–63 military (defence, armed forces) 140–41, 143–4, 146, 154, 156–7, 160 higher education 148, 158 transportation 148, 162–3 health (medical care) 140, 148, 156–7, 159–60 education 140–41, 143, 148, 156–7, 158–9 police 148, 161–2 social policies 144, 154, 156–7, 161 political patronage, see patronage politicization 99, 170, 188, 217, 220 see also patronage private institutions 251, 253, 266 private sector 77–9, 88 privatization 5–6, 9–10, 284–5 Australia 77–8, 84–8, 92, 93 Canada 112, 115, 126, 129, 132 Denmark 261 France 43, 246 Germany 171–2, 175, 178, 185, 192–3

296

The state at work, 1

New Zealand 43, 46–7, 56, 62, 66, 70, 73–8 Spain 196, 203, 207 United Kingdom 20, 27, 35–6, 38 USA 158, 162 public enterprises 124–6, 251, 253–4, 257, 265 public companies 224–6, 242 see also public enterprises public corporations 32 see also public enterprises public sector management reform programme 76–8, 95–6, 99–100, 102–3 public sector unions, see civil service unions public sector workforce statistical data collection 197–9, 207, 212–3, 218–20 see also female employment; minority group employment quangos 197, 223–4, 227, 231–2 race 165–7 see also minority groups recruitment 197, 199–200, 211 reform 27, 249–51, 263–5 regional government 196, 199–200, 203–5, 207–8, 210–11, 228, 230–34 see also devolution representativeness 172, 189–91 social 181, 189–91, 222, 245 gender 173, 181, 189–90 regional 178, 189 see also female employment; ethnic employment retirement plans 226, 245, 247 school boards 142 semi-state employment 174 senior civil servants 163–4, 229 Senior Civil Service 30, 33, 36 Senior Executive Service 96–9, 102 see also Senior Civil Service Spain 196–221 Franco government 196–7 level of government 203–12 central 212

local government 196, 203–4, 211–2 sub-national level 210 policy areas 212–16 special purpose authorities 28, 142, 148 specialist 232, 243 spoils system, see patronage state, concept 2, 290 state governments, see level of government state sector reform 42–6, 48, 53–4, 66, 69, 70–72 see also reform state sector unions 51–2 see also civil service unions statutory bodies 84–6 strike 226, 234, 235 Sweden 268–82 level of government central 274–6 regional 274–6, 279, 281–2 local 274–6, 279, 281–2 policy areas 269, 271, 274, 276 Riksdag 276, 280 temporary employees 133–4 see also part-time employment United Kingdom 19–38 cabinet 33 level of government 23 central 23–4, 27, 30–31 local 23, 25, 28–9, 30, 38 National Health Service (NHS) 24–5, 29, 34 Next Steps 27 Thatcher, Margaret 21 Westminster model 40 Whitehall 31, 33 USA 140–69 agencies 151 Department of Defense 154 Department of Veterans’ Affairs 146, 154 Executive Office of the President (EOP) 155–6 General Schedule 154, 163–4 level of government 140–41, 146, 149, 154, 156 federal 140, 144, 146–7, 154, 157

Index state 141–2, 147–9, 154, 157 local 141–2, 144, 146, 148–9, 154, 157 policy areas 155–63 special district governments 142, 148 political appointees 154–5, 163 senior public servants 163–4 wage system 154, 163–4 warfare state 140, 143 user-pays 88

297

voluntary sector 78 wages and salaries 112 Wagner’s law 172, 174, 176 Weber, Max 3 welfare state 3–4, 40–41, 46, 55, 62, 71, 72–3, 140, 143, 161, 170–71, 176, 193, 197, 211–2, 222, 224, 230, 258, 268, 280 women, see female employment workplace relations 77, 95, 99