Détails du document
Conference Paper
Publication Year:

French HTML
Reconciling equity and decentralization in income security

International Conference on Federalism Mont-Tremblant, October 1999 BACKGROUND PAPER RECONCILING EQUITY AND DECENTRALIZATION IN INCOME SECURITY Peter Whiteford Department of Family and Community Services Australia 1. Introduction (1) This paper attempts to map a range of issues to be considered when assessing the advantages and disadvantages of decentralisation strategies, in particular in the context of income support policies. The paper addresses a set of inter-related questions: • What is the relationship between decentralization and inter-regional and intra-regional equity? • Is it possible to set adequate national standards while still respecting diversity? • What are the benefits and costs of diversity on the delivery of services? The main objective of the paper is to provide a starting point for discussion of these issues. As such the paper does not reach firm conclusions, indeed the conclusion to this paper provides an extensive set of further questions to be answered. It is intended that these questions provide a basis for a possible future research agenda on the relationship between decentralisation and equity in income support. Such research should be widely based and bring together the practical experiences of the diverse nations encompassed by the Forum of Federations. The paper is structured as follows. The next section discusses the meaning of decentralisation in broad terms. This is followed by a discussion of the scope and definition of income support to be considered for the purposes of this paper. This section also discusses some of the equity objectives of different income support strategies. The following section provides a range of data on regional and jurisdictional income disparities within a number of federations, plus other countries for which information is available. The paper then discusses whether there are potential conflicts between equity and decentralisation. This is followed by a comparison of administrative practice in the income support systems of a number of countries. The paper concludes with a discussion of future research issues. Before turning to the substance of the paper, some major caveats are required. The paper is written from the perspective of a researcher involved in income support policy issues who has experience of comparing income support programs in wealthy OECD countries, and some limited experience of analysing income support policies in the more diverse countries of the Asia-Pacific rim. I am not a student of federal-fiscal relationships, and I have had no specific involvement in analysing issues relating to the decentralisation of policy and programme administration in federations. Moreover, I come from a federation in which income support is highly centralised. The federal countries listed by the Forum of Federations include some of the richest societies in the world, such as the United States, Switzerland and Canada, as well as some of the poorest, such as Ethiopia. The economic and social context of income support thus varies widely across federations (2). Moreover, the issues to be discussed are also relevant beyond these existing federations, in particular in considering income support policy across the European Union and in countries moving towards some form of devolution, the United Kingdom, for example. As a consequence of this diversity, it is impossible to produce a paper that gives full justice to the range of issues and individual experiences. At the same time, I have attempted however inadequately to try to include material from countries other than the rich nations of the OECD. The extent to which I have been able to do so is very limited. To quote Elster (1992, p. vii) “ I am stretching my competence thinly over a large number of areas. It is not just that my treatment of the issues is selective: my knowledge is based on what may well be, in some cases, idiosyncratically chosen, unrepresentative, or dated sources.” 2. Defining Decentralisation As noted in a recent OECD report (1997), issues of inter-governmental management are high on the policy agenda in many countries. For example, Belgium has gone through the process of transformation from a unitary state to a federal state. The 1995 referendum on Québec brought devolution to the top of the political agenda in Canada. Regional councils were established in Finland in 1994. A Devolution Commission was established in Ireland in 1995. The National Development Plan 1994-2000 for Mexico put decentralisation high on the policy agenda. Recent developments in the United Kingdom suggest that the UK may be moving towards becoming some form of federal state. What is meant by decentralisation? A useful set of definitions is provided in the 1997 OECD report, Managing Across Levels of Government. The OECD noted that any move along a centralisation-decentralisation spectrum can take a range of forms. Since differing definitions may be used in different national contexts, some formal definition is initially required (and can then be discussed). The OECD definitions are as follows: • Decentralisation: the transfer of responsibility to democratically independent lower levels of government, thereby giving them more managerial discretion, but not necessarily more financial independence; • Deconcentration: the transfer of responsibility from central ministries to field offices or more autonomous agencies, thereby coming closer to citizens while remaining part of central government; • Delegation: the formal relationship in which competence for an aspect of the regulatory process is given by one level of government to another; and • Devolution: the umbrella term covering all forms of transfers of responsibility. 3. Why decentralise? Interest in decentralisation arises from a number of differing concerns and perspectives. First, many governments face pressures associated with budgetary deficits. Decentralisation may offer the opportunity to reduce unnecessary duplication and overlap, so resulting in better value for taxpayers’ money. Related to these concerns is the interest in improving service delivery across levels of government and improving coordination between public and private service providers. It can be argued that decentralisation may lead to increased policy flexibility and improved client and service orientation. In addition, interest in policy experimentation in the field of income support suggests that there may be advantages in decentralisation, as providing a basis for evaluation of alternative policy reforms. Second, many governments and commentators are also concerned with improving the quality of democratic participation. If services can be made more responsive to local needs, it may be possible to further empower citizens to participate in more decision making. Hentschel, Lanjouw, Lanjouw and Poggi (1998) also note that underlying some of the current arguments in favour of decentralising poverty alleviation efforts is the notion that local communities are best placed to identify the interventions that are most beneficial to the poor in their area. This position is based on the contention that at the local level public resources are less likely to be captured by some non-poor households (Hentschel et al., 1998, p. 29). Third, further interest in decentralisation arises because of a renewed interest in the spatial and geographical dimensions of social problems – what Kleinman (1998) has called « the new politics of place and poverty ». There has been a striking increase in interest in area-based initiatives in many countries, including the United Kingdom, where one focus of government concern with social exclusion is recognition that deprivation is geographically concentrated. A further consideration from the perspective of developing countries is that location may provide a good proxy for greater need, and thus provide a basis for targeting scarce resources (3). A related concern is that existing social programs have not proved effective in combating social exclusion, or arguably have actually increased problems of exclusion by promoting a culture of « welfare dependency ». From this perspective what is needed are initiatives to promote the active participation of low-income individuals and families in valued social and economic activities. This promotion of independence is almost invariably focussed at some regional or territorial level, since social participation normally has a spatial dimension. It can also be argued that there are parallels between such concerns in wealthy societies and issues relevant to developing societies. For example, Jalan and Ravaillon (1997) discuss « spatial poverty traps », as situations where otherwise identical households experience differences in growth in living standards as a result of living in different areas, with different endowments of « geographic capital ». If spatial poverty traps can be shown to exist, then according to Jalan and Ravaillon (1997) there may be efficiency and equity arguments for investing in these areas by developing local infrastructure or encouraging labour export to better endowed areas. Interest in decentralisation can also be viewed from the perspective of resisting positive moves towards centralisation. This appears to be relevant to consideration of the concept of subsidiarity, as notably discussed in debates on European integration. According to Streeck (1995), the concept of subsidiarity in Catholic social doctrine, where it originated, implied a duty on the part of higher levels of governance to enable smaller units at lower levels to conduct their affairs in responsible social autonomy. Where the outcomes of this lead to shortfalls from generally desirable standards, higher levels of governance had either to adjust the lower unit’s capacities or take on responsibility for outcomes. Streeck (1995) argues that when the European Commission reintroduced the term in the 1980s, it reinterpreted the concept to mean a general presumption of the precedence of lower levels of government over higher levels. He argues that this is ultimately a principle of laissez faire with respect to whatever the lower level of government chooses to do. Thus, the concept has been transformed from one requiring a strong central state mandating and enabling, to one that implies a weak state acting only at the request of sovereign constituencies. He also argues that subsidiarity is usually invoked in EU usage to claim precedence for two very different kinds of lower-level self-governance, territorial and functional. 4. The scope of income support All societies have complex systems of social protection. Social protection has developed over time from family and community-based support, through charity and mutual associations to formal government-provided or financed social security programs. Differing forms of social protection continue to coexist even in countries described as “welfare states” (4). What differs across countries is the mix and form of social protection mechanisms. These mechanisms can support families and communities either through the direct provision of cash assistance (social security), or indirectly by mandating or regulating private social protection (for example, private and occupational pensions or employer-provided sick leave), or by subsidising private expenditures (for example, through the tax system). This paper concentrates on government systems to provide for social security, specifically through the provision of cash transfers to individuals (5). Table 1 provides a listing of social security programs existing in federal countries in 1997. Typically, these systems have a range of objectives, such as assisting individuals and families experiencing contingencies like unemployment, ill health or disability, work injuries, sole parenthood or old age. The difficulty in assessing these objectives and how well they are met is that in most cross-sectional analyses it is virtually impossible to distinguish different types of redistribution. In particular, it is not possible to distinguish interpersonal redistribution (from rich to poor) from intrapersonal redistribution (across the lifecycle or self-insurance). These methodological complications mean that it is difficult to define “equity” for the purposes of this paper. In particular, there is no single available index or measure by which it is possible to rank the redistributive effects of income support systems in different countries (6). Despite this, the design features of income support systems will provide fairly reliable broad indications of how much redistribution is being sought in different countries. In general terms, those systems that are closer to the pure insurance model could be expected to be less redistributive to poor individuals than demogrants (or universal flat-rate benefits), which will in turn be less redistributive than income-related benefits or social assistance (holding levels of spending constant) (Barr, 1990). TABLE 1: SOCIAL SECURITY PROGRAMS BY TYPE, SELECTED FEDERATIONS, 1997 Notes: O: Medical care and/or hospitalisation coverage provided in addition to cash benefits. NA: data not available in source. -: No program listed in source Source: Social Security Administration, Social Security Programs Throughout the World – 1997. TABLE 2: OBJECTIVES FOR INCOME SUPPORT Source: Derived from Hills, 1990 and Barr and Coulter, 1990. There are other sorts of redistributive objectives or redistributive effects to be considered. For example, in recent years a good deal of attention has been given to issues of gender equity in social security. Different social security systems will redistribute resources from men to women to differing degrees. These variations will reflect the specific rules determining contributions, entitlements and benefits in different programs, as well as differences in the social and economic circumstances of women, such as their labour force participation rates, differences in hours of work, and differences in wage rates. In practice, social security systems can also redistribute to different generations. For example, when social insurance systems were first established in many countries, benefits were paid out to retirees before they had accumulated full contribution histories. As a result, « early » members of social security schemes contributed less than they received in benefits. This is possible because most systems are largely funded on a pay-as-you-go basis, where current entitlements are paid for by current contributions. This form of redistribution can be the consequence of either explicit or implicit policy decisions. Redistribution through social security across regions either within federations or in unitary states appears to largely reflect implicit policy decisions rather than explicit design features. While it is common for governments to explicitly design systems of inter-governmental transfers so that richer regions subsidise poorer regions, it is very uncommon to have geographical redistribution set out as a basic objective of transfer systems for individuals. While the extent of any geographical redistribution will necessarily be the consequence of explicit design choices, those choices are usually made without conscious reference to the desirability or otherwise of redistributing from rich to poor regions. It can also be argued that a completely non-redistributive social security system may appear to redistribute income across regions under certain conditions. For example, imagine a country that had a compulsory defined contribution pension system under which the benefits received by individuals are solely related to their own individual contributions and the earnings of their personal funds. Strictly speaking, such a system would involve no inter-personal redistribution. However, should pensioners under such a scheme choose to move to a different province on retirement, then it will appear that contributors in one region are paying for beneficiaries living in another region. Measuring the extent of redistribution across regions is therefore potentially fraught with methodological complexities. 5. Regional differences in income levels, inequality and poverty Putting these complexities aside for the moment, Table 3 provides a range of comparisons of income differentials across regions in a number of federal states plus the United Kingdom. The most important point of these comparisons is the differences between regions in different societies, rather than the level of the measures for any country or jurisdiction. For example, in Australia the Gini coefficient in 1996-97 was 1.15 times higher in Western Australia than in Tasmania. In Great Britain, the level of inequality as measured by the Gini coefficient was 1.31 times higher within London than in Scotland, while in the USSR in 1989 inequality within Azerbaijan was 40 per cent higher than in the Ukraine. If further disaggregation is undertaken for Australia, the range for the Gini coefficient increases to 1.30 from 1.15. It is plausible that in other countries moving the focus to smaller areas would also increase the disparity between inequality measures. TABLE 3: COMPARISONS OF INCOME DIFFERENTIALS IN DIFFERING JURISDICTIONS AND REGIONS, SELECTED COUNTRIES TABLE 4: TRANSFER SPENDING PER CAPITA BY REGIONS IN FEDERAL COUNTRIES Note: Figures for the United States are average payments per recipient in the state, not average per capita for the state. Sources: Australia: Bray and Mudd, 1998; Canada: Banting, 1995; South Africa: Financial and Fiscal Commission, 1996; USSR, Falkingham, 1999; United States: Statistical Abstract of the United States. Table 4 provides a range of estimates of spending on cash payments by regions in four federal countries. All the figures have been standardised to the national average level of spending per capita. It is important to note, however, that the figures for the USA refer to average spending per recipient. This may reduce disparities in the USA relative to the other countries, as it removes differences between states in the proportion of the population who receive transfers. It can also be seen that social insurance benefits in Canada have a much wider range than income support benefits. This is surprising given that income support is income-tested, and could therefore be expected to be more redistributive than social insurance. This may reflect the then national character of unemployment insurance, while income-tested payments could vary across provinces. It can also be seen that state unemployment insurance benefits in the USA in 1996 were more disparate than social security retirement pensions, but that the then AFDC program provided an extremely wide range of average benefits, which were more than 6 times higher in Alaska than in Mississippi. This shows that the US situation needs to clearly delineated from the situation in Canada or Australia. In Canada and Australia average benefits are highest in poor regions (Tasmania, Prince Edward Island and Newfoundland) and lowest in richer regions such as the ACT and Alberta. In contrast, in the USA average benefits are lowest in poor states (Louisiana and Mississippi). Table 5 provides estimates of differences in poverty rates across a range of federal countries in the 1980s and 1990s. This can be regarded as a slightly better measure of the effects of the system of cash transfers in reducing disparities (although we do not know what poverty rates were “before” transfers in these countries and regions. The poverty lines used for each country vary, of course, so again it is the difference between jurisdictions that is important rather than the level of poverty. In addition, the poverty lines for Australia, India and the USSR are constant across all regions, while the poverty lines for Canada and the USA vary according to urban or rural location. 6. Equity and decentralization – do they need to be reconciled? This paper addresses issues associated with the reconciling of equity and decentralisation in income support policy. This specification of the topic implies that there is some tension if not actual conflict between equity and decentralisation. One argument for the existence of such tension has been put in a recent article in Prospect, where Richard Thomas (1998) argues that devolution in the United Kingdom will increase the tension between equality and democracy. He argues that « devolving power will make the high levels of redistribution between regions and nations in Britain more transparent. … Once devolution makes the invisible redistributive hand plain for all to see, there is likely to be a backlash.” Thomas quotes research by the Institute of Fiscal Studies (IFS) which estimates changes in income tax rates if redistribution between regions in the United Kingdom ended. In London and the southeast, the standard rate of income tax could fall and in Wales and Scotland tax rates would rise. Standard and higher rates of income tax in Scotland and Wales would need to be twice as high as for London and the southeast, under the scenario of no inter-regional redistribution. The basis for the estimate is the difference between regions in per capita GDP and consequently in levels of taxation, as well as differences in social security receipt. After calculating the scale of current redistribution the IFS is quoted as concluding that, « It is quite possible that when the scale of these flows is made explicit, they will not attract the same public consent as the implicit transfers currently have. This would imply that decentralisation could substantially worsen the position of areas that currently benefit from fiscal transfers » (1998, pp. 14-15). Further evidence of large redistributions across regions is available from Canada. Courchene (1993) quotes figures for Canadian unemployment insurance in 1988 indicating that for each dollar of contributions Newfoundlanders received $4.16 in benefits, while Ontarians contributed nearly twice as much as they received in benefits. Similarly, he estimates that in 1988 Quebec benefited from nearly $900 million in net inward transfers due to unemployment insurance. Table 6 provides what are now very dated estimates of the redistributive effect of public finances between regions in a selection of federations and unitary states in the late 1970s. The table suggest that public finances broadly defined substantially reduced differences in average per capita incomes across regions, and consequently in regional income inequality. The extent of redistribution appears to have been greater in the four unitary states than in the federal states, although the effects of the Australian system were about as great as the most redistributive unitary state (France). The least redistributive unitary state at the time was the United Kingdom, which at the time apparently was not as effective as Germany in reducing regional income inequality. Table 7 provides calculations for Australia of an index of net redistribution, which is calculated as the ratio of average income taxes paid by a state or region to the average income support benefits received by the region, then standardised as a proportion of the overall Australian average ratio. This therefore can be interpreted as an index of how much more or less than the average some regions « put in ». On this basis there are substantial differences between Australian states and regions. The Australian Capital Territory, the Northern Territory and Western Australia are the highest net contributors (8), while South Australia and Tasmania are the highest net beneficiaries. The table also suggests that Sydney « contributes » about two and three quarter times the contribution of non-metropolitan Tasmania. At a more disaggregated level, Mosman a wealthy suburb of Sydney contributes at nearly five times the average Sydney rate. The index also suggests that Roxby Downs, a remote mining town contributes at more than 100 times the contribution rate of Inala or Elizabeth, both capital city areas with high concentrations of public housing. Other small regions that are large net benefit recipients include some rural areas, illustrated here in the case of Moe. These examples therefore suggest large effective redistributions across regions can occur, at least in the case of the United Kingdom and Australia and Canada and plausibly in many other societies. If the scale of these redistributions were reduced, would equity be challenged? The difficulty in answering such a question is that one needs some form of benchmark to assess equity. One obvious issue to be considered is that the cost of living may vary across regions as may average incomes. From an “absolute” perspective on living standards, it may therefore be possible to justify differences in levels of benefits across regions as reflecting differences in needs and therefore as being fair. Even from a relative adequacy perspective differences in benefits could be justified if they were set as a constant proportion of community incomes, where community incomes also vary across jurisdictions. Similarly, it could also be argued that uniform national benefit rates in a country where wage levels vary substantially may create unintended incentive effects. An obvious research question here is to determine the size of differences in the cost of living across regions in different countries. 7. Decentralisation and Administrative Practice – Lessons from OECD Countries By definition, the term decentralisation suggests a process of change. Actual examples of decentralisation in practice would therefore require examples of reforms by which responsibility for income support or aspects of income support policy was taken from central levels of government and given to lower levels of government, either state/provincial or local/municipal. There appear to be two outstanding examples of delegation of responsibility for income support policy in the last decade. The first of these is the changes to the then Canada Assistance Plan (CAP) from the middle of the 1980s onwards, its subsequent replacement by the Canadian Health and Social Transfer (CHST) from 1995, and other changes to Unemployment Insurance (UI) and employment development services from 1996. The second is the set of changes to welfare arrangements in the United States culminating in the replacement of the Aid to Families with Dependent Children (AFDC) program with Temporary Aid to Needy Families (TANF) from 1996. There is already a very large literature on the nature of these changes both in Canada and the USA and a good deal of debate about the fairness of the outcomes of these changes (9). The complicating factor in assessing the lessons of these changes is that they involve a good deal more than just delegation or decentralisation. For example, the equity implications of the US changes also need to be assessed in relation to the effects of time-limiting entitlements to benefits, as well as changes to eligibility for payments. Given that the full implications of time limitations are not yet available and the USA continues to enjoy robust economic growth it would appear premature to reach judgements about the outcomes of these policy changes. An alternative source of information for assessing the implications of decentralisation is provided by the comparative experience of different countries where differing levels of devolution of income support functions can be identified. Having said this, it appears to be uncommon in much of the comparative literature to separately discuss decentralisation issues. Some information and points for discussion are available from a number of recent studies of social assistance benefits in OECD countries. Table 8 summarises information on the degree of centralisation/ decentralisation in the administration of social assistance in OECD countries in the early 1990s. The material comes from the research of Eardley, Bradshaw, Ditch, Gough and Whiteford (1996), which was the first study of social assistance in all OECD countries. TABLE 8: ADMINISTRATION OF SOCIAL ASSISTANCE SCHEMES, OECD COUNTRIES, 1992 Source: Eardley et al, 1996, pp. 63-64. The OECD has subsequently undertaken further in-depth analysis of the social assistance schemes of Australia, Finland, Sweden and the United Kingdom (1998a) and Belgium, the Czech Republic, the Netherlands and Norway. These reports provide a more detailed analysis of administrative issues than does the earlier, more broadly based study by Eardley et al. (1996). The OECD (1998a) point out that the form of benefit delivery in different countries reflects the nature of the social policy system. The strength of the locally based social assistance schemes of Finland and Sweden derives from the role of local government in providing a universal safety net, so that few people find themselves in severe hardship without some form of intervention. However, the corresponding disadvantage of these systems is that local governments have found themselves overloaded with routine cases of income support. The OECD therefore suggested that in the case of Finland and Sweden that administration of basic allowances be integrated with state social insurance schemes, leaving local governments to provide the additional tailored services for the more difficult cases (1998a, p.128). The OECD also note that the centralised Australian system has achieved a “high and impressive level of efficiency and customer service … a considerable achievement, which has eluded most OECD Member countries” (1998a, p. 128). However, they also note that both the centralised systems of Australia and the United Kingdom are weak when it comes to relations with social services at the local level, so that identifying and assisting difficult cases may be less effective. The OECD also noted that in their discussions in each country they detected a tendency for local governments in the UK and state governments in Australia to regard maximisation of benefit claims as a plus for their region. The OECD’s subsequent (1998b) discussion of Belgium, the Czech Republic, the Netherlands and Norway provides a more detailed assessment of administrative issues. In the case of Norway it is noted that social workers respond to the local democratic process by following “scale rates” set by local politicians, but that in complex cases, there appears to be little correlation between actual payments and these scales. Rates also differ markedly across areas, and apparently reflect the personal judgements of social workers as to who is more “deserving”. While this does not appear to be controversial, the OECD argues that the system may become discredited if arbitrary decisions continue to be made outside of local political control. It is also noted that in Norway, financial responsibility for social assistance payments is completely devolved to the local level, so that benefits are tightly administered. In the Czech republic local governments are merely payment agents of central government, and have no independent financial responsibility. Because costs are shared in Belgium and the Netherlands, financial incentives to control access and reduce fraud are considered to be relatively low. The OECD also argue that locating financial responsibility for assistance with the institution most able to move clients off benefits is likely to lead to an appropriate level of resources being devoted to reintegration than if responsibilities are separated. Locating both responsibilities at the local level may not be feasible in all countries. While financial risks may be mitigated by grants from central to local government, this could conflict with other grant systems. The OECD also argue that introducing purchaser/provider relationships into the employment services can provide incentives for employment agencies to tackle clients with the most severe labour market problems (OECD, 1998b, pp. 156-157). Outside the OECD an interesting perspective on the financing issues has been provided by the Financial and Fiscal Commission of South Africa (1996), which is worth quoting at length. The case of welfare funding needs specific discussion, as it requires an explanation of the difference between delegated and devolved or decentralised functions. Delegated functions are those which are carried out by the provinces on behalf of the national government, in which case the province has no policy control or decision making authority. Neither the extent, nor the manner of service provision is discretionary. A good example is the delivery of state pensions for the aged – the national department has set statutory norms and standards including the amount to be paid to every eligible individual, the criteria for eligibility as well as the mechanism for means testing. Qualifying individuals have an entitlement to this stated amount. The province has no choice, hence pensions are a delegated function. In contrast, a devolved function is one where the subnational government is able to apply its own policies within the broad constraints set by national government. The provincial government must, for example, apply certain funds for a function or functional programme but other than this constraint, it is able to exercise its prerogative. The implication of welfare being a delegated function is that national government should be obliged to fund the full cost of pension pay-outs (1996, p.5). 8. Conclusions: Towards a Research Agenda As discussed at the beginning of this paper, it is not intended to give definitive answers to the question of whether decentralisation of income support can be reconciled with equity. The contexts facing different federations (and other societies) vary widely. Consideration of decentralisation strategies therefore needs to take into account the differing interests and concerns underlying calls for decentralisation. The balance of objectives for decentralisation may differ significantly between (and within) countries. In addition, different individuals within different countries will define equity in different ways. What follows therefore is a set of questions about these issues addressed to the audience at this conference. Understanding of these questions can only be advanced by a process of dialogue within the individual countries considering decentralisation. It seems reasonable to argue, however, that these dialogues can be assisted by bringing together international experience in dealing with these issues. It would also seem desirable to consider the preparation of a set of research papers analysing in detail administrative practices in different countries and the national experience of each country according to the degree of devolution of responsibility for income support. A fundamental set of questions to be discussed surrounds the meaning(s) of equity in different countries. How much difference across jurisdictions or regions is acceptable in your country? What is the source of the differences between jurisdictions in you country? Are these disparities driven by differences in the average incomes of regions or differences between the degree of inequality within regions? What are the differences between wages and productivity and the levels of labour market participation? What is the role of differences in demographic structure, the urban/rural mix, and what is the effect of ethnic and religious differences? How much redistribution across regions actually occurs in your country because of the effects of the income support system and the means of financing it? What data are available on this redistribution and how can we improve measures (and underlying concepts) of this redistribution, if any? Can we think of a standard way of measuring inter-regional redistribution across a range of countries? What are the problems or issues with your income support systems that you believe are priorities for reform? Is your system of income support too expensive? Administratively inefficient? Is it too mean or inadequate? Are there problems of coverage? Is your system appropriately targeted to the poor (if that is an objective)? Do you believe that there are adverse incentives built into your system? Does it create welfare dependency or encourage long-term unemployment? How will decentralisation or devolution affect these concerns? What are you seeking to achieve if you are a proponent of decentralisation or devolution? Is your concern with the cost and efficiency of current arrangements? Are you concerned with encouraging greater democracy and community participation? Is the present system a failure? Do you believe that decentralisation will encourage experimentation and policy flexibility? What are the relative administrative and financial capacities of different jurisdictions? Are there other ways of achieving your goals? What are your national goals in relation to state formation? Are you seeking to integrate the jurisdictions within your country? Do you believe that greater unity will be achieved by encouraging devolution within your federation? Footnotes 1. The views expressed are my own and should not be taken to represent the views of the Australian Department or the Minister for Family and Community Services. 2. Appendix Table 1 provides a range of background data on income levels, income inequality and poverty, and the level of social security spending and taxation revenue in federal countries in the 1990s. 3. It should be noted that this is in the context of poor alternative mechanisms for targeting to the poor. 4. Recent estimates for Australia suggest that 46 per cent of persons aged between 15 and 24 years receive private transfers in the form of free accommodation, while nearly 40 per cent of those aged 75 years and over receive private transfers in the form of personal care and home help (King and McDonald, 1999). It is also clear that in all wealthy “welfare states”, children benefit from extremely significant private transfers. 5. In this context, I do not include cash transfers to reimburse individuals for medical expenses, which I define as health expenditures. 6. Note that there are many studies that measure the distribution of cash disposable income “after” taking account of social security programs. Because these studies usually rely on a counterfactual where the existence of the social security system has had no impact on the “original” distribution of income, they cannot be considered to produce valid estimates of how much redistribution is actually achieved by cash transfers. See Whiteford (1997,1998) for further details. 7. This is simply because the overall Gini coefficient for any country is calculated by ranking all incomes sequentially across the whole country. 8. Many Australians may disagree with the view that the ACT and the Northern Territory are the highest net contributors, since the source of their high average incomes is the high level of public sector employment, funded by Australian taxpayers. 9. For a very good overview of Canadian changes see Battle (1998); for discussions of the US changes, see many of the recent publications of the Institute for Research on Poverty, including Haveman and Wolfe (1998), Kaplan (1998) and Mead (1999) at References Adema, W., Einerhand, M., Eklind, B. Lotz, J. and Pearson, M. (1996), Net Public Social Expenditure, Labour Market and Social Policy Occasional Paper, No. 19, OECD, Paris. Ardington, E. and Lund, F. (1995), Pensions and Development: How the social security system can complement programmes of reconstruction and development, Development Bank of Southern Africa, Development Paper 61, Halfway House. Atkinson, A.B. (1992), Towards a European Social Safety Net? Discussion Paper WSP/78, Welfare State Programme, STICERD, London School of Economics, London. Australia, Department of Family and Community Services (1999), Social Indicators for Regional Australia, mimeo, Canberra. Banting, K.G. (1995), “The Welfare State as Statecraft: Territorial Politics and Canadian Social Policy”, in Leibfried, S. and Pierson, P. (eds), European Social Policy: Between Fragmentation and Integration, The Brookings Institution, Washington D.C. Barr, N. (1992), ‘Economic Theory and the Welfare State: A Survey and Reinterpretation’, Journal of Economic Literature, 30, June, 741-803. Battle, K. (1998), “Transformation: Canadian Social Policy since 1985”, Social Policy and Administration, 32,4, December, 321-340. Caledon Institute of Social Policy (1999), Poverty Eases Slightly, Courchene, T.J. (1993), “Path-Dependency, Positive Feedback and Paradigm Warp: A Schumpeterian Approach to the Social Order”, in E. Reynolds, E. (ed.), Income Security in Canada: Changing Needs, Changing Means, Institute for Research on Public Policy, Montreal. Dalaker, J. and Naifeh, M. (1998), Poverty in the United States: 1997, Current Population Reports P60-201, Economic and Statistics Administration, Bureau of the Census, Washington DC. Eardley, T., Bradshaw, J., Ditch, J., Gough, I. and Whiteford, P. (1996), Social Assistance Schemes in the OECD Countries: Synthesis Report, Department of Social Security, Research Report No. 46, HMSO, London and OECD, Paris. Eardley, T., Bradshaw, J., Ditch, J., Gough, I. and Whiteford, P. (1996), Social Assistance Schemes in the OECD Countries: Country Reports, 1996, Department of Social Security, Research Report No. 47, HMSO, London and OECD, Paris. Elster, J. (1992) , Local Justice: How Institutions Allocate Scarce Goods and Necessary Burdens, Russell Sage Foundation, New York. Falkingham, J. (1999), Welfare in Transition: Trends in Poverty and Well-Being in Central Asia, CASEpaper No. 20, Centre for Analysis of Social Exclusion, London School of Economics, London. Ferreira, F.H.G. and Litchfield, J.A. (1996), Growing Apart: Inequality and Poverty Trends in Brazil in the 1980s, Distributional Analysis Research Programme, Discussion Paper No. 23, LSE/STICERD, London. Glennerster, H., Lupton, R., Noden, P. and Power, A. Poverty, Social Exclusion and Neighbourhood: Studying the area bases of social exclusion, CASEpaper No. 22, Centre for Analysis of Social Exclusion, London School of Economics, London. Hentschel, J. Lanjouw, J.O., Lanjouw, P. and Poggi, J. (1998), Combining Census and Survey Data to Study Spatial Dimensions of Poverty: A Case Study of Ecuador, World Bank Working Paper, Washington D.C. Jalan, J. and Ravaillon, M. (1997), Spatial Poverty Traps? World Bank Working Paper, Washington D.C. Khan, Q. (1998), Social Protection in a Changing Economic and Social Environment – The Australian Experience and its Lessons for World Bank Group Client Countries, Human Development Sector, World Bank, mimeo. Leibfried, S. and Pierson, P. (1995), European Social Policy: Between Fragmentation and Integration, The Brookings Institution, Washington D.C. Lund, F. (1992), « Social Security and Social Policy in ‘The New South Africa' », in Social Security: 50 Years after Beveridge, Volume 2, Competing Models of Social Security: A Comparative Perspective, University of York, 27-30 September. National Council of Welfare, Canada (1998), Poverty Profile 1996, Ottawa. Organisation for Economic Co-Operation and Development (1996), Social Expenditure Statistics of OECD Members Countries, Provisional Version, Labour Market and Social Policy Occasional Papers, No. 17, OECD, Paris. Organisation for Economic Co-Operation and Development (1998a), The Battle Against Exclusion, Volume 1, Social Assistance in Australia, Finland, Sweden and the United Kingdom, OECD, Paris. Organisation for Economic Co-Operation and Development (1998b), The Battle Against Exclusion, Volume 2, Social Assistance in Belgium, the Czech Republic, the Netherlands and Norway, OECD, Paris. OECD (1998c), Income Distribution and Poverty in Selected OECD Countries, Economics Department Working Paper No. 189, OECD, Paris Reynolds, E. (ed.) (1993), Income Security in Canada: Changing Needs, Changing Means, Institute for Research on Public Policy, Montreal. Social Security Administration, United States, Social Security Programs Throughout the World – 1997, Research Report No. 65, Social Security Administration, Washington DC. South Africa, Financial and Fiscal Commission (1996), The Financial and fiscal Commission’s Recommendations for the Allocation of Financial resources to the National and Provincial Governments for the 1997-98 Financial Year, Financial and Fiscal Commission, Pretoria. Stewart, K. (1997), Are Intergovernmental Transfers in Russia Equalizing? Innocenti Occasional Papers, Economic and Social Policy Series No. 59, UNICEF, Florence. Streeck, W. (1995), “From Market Making to State Building? Reflections on the Political Economy of European Social Policy”, in Leibfried, S. and Pierson, P. (eds), European Social Policy: Between Fragmentation and Integration, The Brookings Institution, Washington D.C. Thomas, R. (1998), « Equality versus Democracy », Prospect, November, pp. 14-15. United Nations Development Programme, Human Development Report 1996, Oxford University Press, New York. Whiteford, P. (1997), « Targeting Welfare: A Comment », Economic Record, 73, 220, March, 45-50. Whiteford, P. (1998) « Is Australia Particularly Unequal?” in P. Smyth and B. Cass (eds), Contesting the Australian Way: States, Markets and Civil Society, Cambridge University Press, Melbourne. Wurzel, E. (1999), Towards More Efficient Government: Reforming Federal Fiscal Relationships in Germany, OECD Economics Department Working Paper No. 209, OECD, Paris. APPENDIX TABLE 1: INCOMES, INEQUALITY, POVERTY AND PUBLIC SPENDING IN FEDERATIONS, 1990S Note: -: data not given in source. Source: United Nations Development Programme, Human Development Report 1996. Forum of Federations / Forum des fédérations