Number 4: Social policy directions across the OECD region: reflections on a decade
This report was published by the former Department of Families, Community Services (FaCS).
- 1 Introduction
- 2 Changes in social expenditures
- 3 Social policy trends across OECD countries
- 4 Conclusion
Unemployment remains a persistent problem for many who are members of the Organisation for Economic Co-operation and Development (OECD). The average rate is around 7 per cent for the OECD region as a whole, with unemployment rates above 10 per cent for a number of European countries. Government budgets have come under increasing strain from the increased income support sought by people of working age unable to find employment. At the same time, there has been considerable growth in other social security payments to people of working age, such as disability pensions, sickness benefits, lone parent pensions and early retirement pensions.
There is some concern as to whether social policies have responded appropriately to the many changes in family arrangements, such as the increased incidence of marriage breakdown and lone parenthood, the increased incidence of no-earner as well as two-earner families, and the longer average time children remain economically dependent on their parents. The tax-transfer and wage systems in many countries have created poor work incentives for families to take up low-paying jobs.
The ageing of the population in most OECD countries is leading to further pressures for change to social and health programs. This is particularly an issue for countries with very generous, earnings-related retirement pensions, which provide very high earnings replacement rates funded on a pay-as-you-go basis, often coupled with generous early-retirement pensions and limited incentives in the pension system for people to remain working. An ageing population is also expected to lead to expanding health and long-term care expenditures. The likely extent of this increase will be influenced by many factors, such as developments in morbidity, medical technology and treatment, as well as the preferences of the older population themselves.
In response to these and other pressures, many OECD countries have been reforming and adapting their social programs. This paper does not attempt to describe all of these changes, but rather provides a thematic overview of developments.1 The paper begins with a brief picture of recent trends in social expenditures, followed by discussion of some of the major policy directions evident over recent years, including examples from particular countries.
Most OECD countries have been subject to fiscal consolidation pressures over recent years (Table 1). Many European countries have sought to meet the fiscal criteria for entry into the European Monetary Union, including some countries that did not seek to join in the first round. Many other OECD countries have been pursuing fiscal consolidation to introduce greater long- term sustainability and stability to government finances.
Nevertheless, it should also be recognised that despite these commitments to fiscal prudence, there are considerable differences in budget outcomes between countries. Some countries, such as Australia and the United States, have budget surpluses, while others continue to have budget deficits (although at a lower level than previously).
Social programs, which account for a large share of government activity in nearly all countries, have necessarily been affected by these fiscal consolidation exercises. Gross public social expenditures account for around 20-30 per cent of GDP in most OECD countries.2 If governments were to meet their overall fiscal targets, they needed to moderate their social expenditures.
Continuing high unemployment and ageing populations, in particular, have challenged the affordability of the welfare state in a number of countries. Adverse employment conditions have reduced the potential scale of contributions for social insurance schemes at the same time as payment liabilities have increased for a range of income support programs. In France, Germany and Italy, fiscal policies have been conflicting with political pressures for expansion of assistance in an environment of high and persistent unemployment. In a number of the Nordic countries, such as Finland and Sweden, fiscal consolidation was being pursued at the very time unemployment was rising significantly.3 Finally, some countries with generous pay-as-you-go public pension schemes are attempting to bring their pension systems towards fiscal balance by the time the 'baby boom' generation moves through into older ages around 2010-35.
Source: OECD 1998e.
Member of EU that sought to satisfy Maastricht criteria1 for economic and monetary union
|Countries negotiating for entry into EU in the next wave||Other governments with a commitment to a surplus or balanced budget2|
|Countries||Austria, Belgium, Denmark3, Finland, France, Germany, Greece3, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden3, United Kingdom3||Czech Republic, Hungary, Poland||Australia, Canada, Iceland, New Zealand, Norway, Switzerland, United States|
To date, fiscal restraint has not resulted in falling gross public expenditures on social programs. As a general rule, public expenditures on social programs have stabilised as a percentage of GDP, or are still increasing but now at a slower rate (Table 2).4 In most OECD countries, the level of public social provision still remains high compared to previous spending levels, despite the steps towards fiscal consolidation.
This at least partly reflects the particular role of social programs, especially in providing for the poor and disadvantaged. Where there have been substantial reductions in public expenditure to bring budgets into balance, there has often been some attempt to protect social expenditures from the full magnitude of the cuts. These expenditure trends also reflect the difficult economic environment in the first half of the 1990s, when many OECD countries went into recession, and a number have experienced sluggish growth since that time.
The information on gross government expenditures on social programs (Table 2) should be used with some caution, especially when making comparisons across countries, because of the underlying differences in the structure of social protection arrangements across countries. Countries with a larger proportion of their social programs funded through private provision rather than the public sector will show up as having relatively low public sector expenditures.
A more accurate picture of the scope and generosity of total social expenditures across countries can be obtained by taking into account the assistance provided through the tax system, any tax liability on government transfers, and the scope of private social programs. When this is done (Adema and Einerhand 1998, Adema 1998), there is greater similarity across countries in terms of total net social expenditures as a percentage of GDP than is apparent from a comparison of gross public social expenditures, which is the usual indicator of the scale of social programs in OECD countries5. Unfortunately, because these estimates have been developed only recently, it is not yet possible to investigate trends over time using this measure.
|Source: OECD Social Expenditure Data Base 1998
1. Information refers to 1994.
2. Information refers to 1993.
3. Information refers to 1992.
While there are obvious differences between countries in terms of their policy priorities and the detailed design features of their social security systems, many of the social issues are similar, and there appears to be considerable convergence in directions of policy reform. Some of the distinct social policy trends evident in OECD countries include:
- 3.1 Increasing self-reliance
- 3.2 Changing financial incentives in favour of employment
- 3.3 Integrated policy packages and delivery
- 3.4 Ensuring basic provision to reduce poverty
- 3.5 Re-adjusting intergenerational burdens
- 3.6 Improving the integrity of social security
- 3.7 Improving the efficiency and quality of service provision
Increasing individual and family self-reliance has occurred in response to a number of separate pressures - in particular, social programs being scaled back as a result of fiscal consolidation, and the introduction of charges for individuals (or increased charges) to reduce unnecessary usage of services (especially in the health care and, less prominently, long-term care areas). In a number of developed countries, there are concerns about the proportion of the working-age population receiving welfare benefits (Preston 1997), and the large and increasing number of working-age households where no adult is in work (OECD 1998e,f).
Access to social insurance programs has been restricted through more closely defined eligibility criteria, reduced benefit levels and reduced duration of benefits. Public assistance is being limited, with the expectation that individuals will either manage on lower incomes or draw on income from other sources, such as increased work effort or savings. Those who have been unsuccessful in their job search now rely more on means-tested safety nets, such as social assistance benefits. These policy reforms - changing insurance eligibility criteria and benefit levels, and greater recourse to means-tested benefits - can either encourage greater self-reliance for some or lead to greater dependency, depending on the nature of social protection arrangements and the household circumstances.6
It is too simplistic to see this shift towards greater self-reliance as purely an economic phenomenon or a reaction to fiscal pressures. Acceptance of welfare can be demeaning for some people and lead to long-term isolation from usual societal activities. In some countries, such as Japan, the encouragement of self-reliance is linked to efforts to improve dignity and self-esteem.
There is also a growing acceptance that passive benefits (in-cash and in-kind) will not provide a long-term solution to poverty alleviation (see, for example, Mead 1997). Countries are placing greater emphasis on encouraging and facilitating the transfer back to employment for jobless people of working age. This is being done through policies to 'make work pay' and complementary policies to improve the likelihood of jobseekers being employed, such as through training, education, work experience and wage subsidies (OECD 1996c, 1997a). Policies expanding child-care provision and accessibility so parents with young children can participate in paid work are further examples of this approach.
Some countries with comprehensive public systems for providing sickness benefits have modified program entitlements so that employers now have greater de facto responsibility for financially supporting employees during the initial periods of sickness, after which stage public sickness benefits are payable (for example, in Austria, the Netherlands, Germany, Belgium, Sweden and the United Kingdom). This has produced not only lower public costs but also changed incentives for employers to ensure their workplaces are safe, healthy environments and that they have mechanisms to reduce the level of absenteeism.7
Greater emphasis is being placed on private provision for retirement income, either as a complement to or a substitute for public pensions, with some ongoing public involvement through tax concessions and financial regulations. However, in many OECD countries, the historical heavy reliance on public pensions to deliver retirement incomes will not be reversed quickly or easily, and the switch to private pensions will be incremental.8
In the health sector, different types of cost sharing have been imposed on patients, with the intention of reducing demand for health services in the context of predominantly universal health care arrangements and concerns to moderate health care expenditures.9 This has been most apparent in the pharmaceuticals area, where expenditure growth has been startling. Cost sharing is actively used in a number of OECD countries for hospital in-patient costs as well as general practitioner (GP) and specialist fees. It is important to note that many cost-sharing arrangements with health services either have blanket exclusion for elderly people and low- income earners, or require much lower contributions from them compared to other members of the population.
The demand for intensive hospital care for the elderly is lessening because of a decrease in acute or fatal diseases and an improvement in the health of older people. Long-term care for the elderly has traditionally relied heavily on family-based care, although there are indications of increasing demand for external services. While institutional care remains at a relatively low level,10 home and community care services are gaining increased importance. These services help people to cope with their chronic conditions in more familiar, client-focused surroundings that respect their autonomy and privacy.
While the majority of OECD countries finance public long-term expenditure from general revenues, a number of countries have either established, or are in the process of establishing, insurance schemes for future long-term care needs as a way of funding some part of the costs for people who require this service. These countries - Germany, Japan, Luxembourg and the Netherlands - all have a long tradition of social insurance arrangements. Contributions are collected from insured people, although both the Japanese and Luxembourg schemes still source around half of the funding from general government revenues.
Somewhat linked to the trend of increased self-reliance is greater consideration by governments of the financial incentives emerging from social policies, and appreciation that they can and do influence the behaviour of welfare recipients.
If people expect to receive little financial gain from working compared to receiving income support, there is little reason to expect them to want to become self-sufficient. This is behind the policy push in many countries to 'make work pay'—to ensure that social policies do not stifle financial incentives for workforce participation (OECD 1997a).
Incentives for the unemployed to work
Many jobseekers receiving social security have limited financial incentive to get a job paying average earnings, let alone to accept a low-paid job more typical of entry-level jobs and jobs available for the disadvantaged unemployed in a number of OECD countries.11 One of the most widely used measures of the incentive for the unemployed to get a job is the replacement rate, which compares the value of unemployment benefit payments to available wage rates.12 Table 3 indicates the difference between unemployment benefits and low wages for different family types, both initially and at the 60th month of unemployment.
|Country||First month of unemployment, 66.7% of APW-level||60th month of unemployment, 66.7% of APW-level|
|Single||Married couple||Couple, 2 children||Lone parent, 2 children||Single||Married couple||Couple, 2 children||Lone parent, 2 children|
|Note: Waiting periods are assumed to have already been met, refers to after-tax comparisons, and includes unemployment benefits, family and housing benefits. Information supplied by Greece suggests a net replacement rate of 50–58% of the minimum wage, depending on marital status and the length of the contribution period.
Source: OECD Data-base on Taxes, Benefits and Incentives, also published in Table 3.1 and 3.4 in OECD 1998d, Benefit Systems and Work Incentives.
Republic of Korea
Replacement rates for short-term unemployment are above 75 per cent for all demographic groups in a considerable number of OECD countries, all of them in Europe (Belgium, Denmark, France, Hungary, Luxembourg, the Netherlands, Portugal, Sweden and Switzerland).While there are a few exceptions for the short-term unemployed, replacement rates for single people are on average lower than replacement rates for married couples without children which, in turn, are on average lower than for couples with children.
Replacement rates generally fall for single people as the duration of unemployment increases, reflecting the exhaustion of time-limited, earnings-related unemployment insurance schemes and the subsequent reliance on flat-rate, means-tested social assistance. However, this is not generally true for married couples and those with children. Replacement rates for long-term unemployed families (unemployed for five years) are higher than replacement rates for the short-term unemployed in as many countries as they are lower, particularly as generous housing benefits start to operate for those long-term recipient families on social assistance.
Over recent years, there have been some reductions in the generosity of unemployment benefits in response to concerns over high replacement rates and poor work incentives for the unemployed:
- the value of unemployment benefit payments has been reduced in, for example, Austria, Canada, Germany, Ireland, New Zealand, Sweden and Poland;
- the maximum duration of unemployment benefit payments has been lowered in, for example, the United Kingdom, Canada, Hungary and Sweden. A number of Nordic countries now require a period of unsubsidised employment (rather than simply participation in a government-sponsored labour market program) before unemployment benefits can be resumed after the maximum payment duration has been exhausted.
More broadly, increased requirements are being placed on people of working age to actively search for jobs and accept suitable employment as a condition of benefit payment. Existing requirements are being strengthened for people on unemployment payments (sometimes in association with increased penalties for non-compliance) and new conditions are being placed on lone parents and social assistance recipients. These conditions are increasingly being formalised through a contract or written understanding between the recipient and the public agency outlining their respective responsibilities.
While the prevailing financial disincentives faced by the unemployed for workforce participation can be moderated, to some extent, through strict enforcement of effective job search activity (work/activity testing), this strategy will be most effective as a complement to properly structured social policies that encourage recipients to take all desirable steps to reduce welfare dependency. Relying solely on very stringent administration processes to compensate for inadequate or inappropriate financial incentives is not a suitable long-term strategy. Operating very stringent administrative processes can be very costly for public budgets and will still rely, to a major extent, on voluntary compliance by individuals to seek work.
Overall, most countries have been reluctant to reduce replacement rates through cuts in unemployment benefit payments because of concerns about the resulting impact on poverty and income adequacy. The more common approach among OECD countries has been to provide positive rewards for those who find part-time or full-time work. Within this category, a number of different approaches have been implemented:
- easing income tests to reduce the loss of benefits when people take up part-time work, through taper reductions and greater account of work-related costs in means testing arrangements;
- providing a one-off or fixed-term earnings supplement or other assistance for people who return to work and leave benefits;
- introducing or expanding in-work benefits, generally targeted to low-income families.
An extensive number of OECD countries are pursuing at least one of these strategies, although most of the recent policy attention has been on a smaller group of countries that provide in- work benefits on an ongoing basis to low-income families. These benefits, often referred to as employment-conditional benefits, are available only to those in employment in Ireland, Italy, New Zealand, Spain, the United Kingdom and the United States, and more broadly to low- income families—whether on benefits or in low to moderate income employment—in Australia and Canada.13 They can be available in the form of cash benefits or tax relief.
These in-work benefits are largely targeted at unemployed families with children, who, among the unemployed, tend to have the worst financial incentives to work.14 They reduce replacement rates (seeking to remove the unemployment trap) and increase disposable incomes at all levels of earnings affected by the in-work benefit. However, they can introduce poverty traps as the assistance is withdrawn when incomes rise (OECD 1997a).
Evidence from evaluations of the Earned Income Tax Credit (EITC) in the United States show that the EITC has in particular increased the proportion of lone parents in work, with a smaller impact on the proportion of couples in work, reflecting some positive impacts for very low income families and negative labour supply impacts over the income range where the EITC is withdrawn. It has been estimated that the EITC is paid to six million American families with earnings below the poverty line, and that the EITC is responsible for taking one million families out of the poverty zone. The recent changes in the United Kingdom to replace Family Credit with the Working Family Tax Credit are also designed to remove some very high effective marginal tax rates (of 90% and above) faced by some families, although more families will face an effective marginal tax rate of 60% and above.
Large in-work benefit schemes can be subject to abuse (OECD 1997a). Findings from a study of EITC entitlement by the US Internal Revenue Service suggested that the total credits paid exceeded entitlements by around 25 per cent. Families in the United Kingdom may be able to manipulate their incomes to gain a higher rate of Family Credit for the subsequent six months because income assessments operate for the next six months, irrespective of any subsequent changes in earnings, in order to not discourage families from increasing their earnings.
Incentives for older people to work longer
Many retirement pension systems in OECD countries have aspects that discourage people from working longer, such as attractive early retirement pensions, very limited accumulation of additional pension benefits for extra years of working, and restrictions on work participation by those who are receiving a retirement pension (OECD 1998b).
In response, countries have (often selectively) introduced a number of changes from the following menu:
- reduced the maximum limits on final benefits, or alternatively changed the overall accumulation rates and benefit calculation methods so people need to work longer to generate the same final benefit;
- increased pension accumulation rates for people in their later years of working life to improve incentives to keep working;
- introduced complementary measures to reduce the scope and incentives for people of working age to take premature retirement. These incentives include restricting inappropriate access to other (often more generous) social security benefits, such as disability pensions, sickness benefits and early retirement pensions;
- reduced the attractiveness and early availability of early retirement pensions, such as through introducing proper, actuarially-based discounting for early access to benefits, or increasing the age and working life requirements for access;
- (in Denmark only) introduced a deliberate strategy to not provide an early retirement pension until all other possible attempts for re-employment have been exhausted.
These changes have been designed to reduce future retirement pension costs in the context of ageing populations in OECD countries. This strategy is also linked to a concern with the now well-established trend of declining labour force participation rates for older men, and a desire to reverse this trend as one means of ameliorating some of the negative economic effects of ageing populations.
OECD countries are seeking more efficient and effective social programs as they continue to face funding constraints and emerging social problems that demand attention. One possible way of achieving this is to improve the extent to which social policies work together, often referred to as 'policy coherence'.15 Some of the key drivers behind the pursuit of policy coherence are recognition of the multiple difficulties faced by social security recipients and the consequent need for a number of interventions, appreciation that income support by itself is often not an effective way of helping many disadvantaged people, and recognition that there are synergies from policies being brought together into a coherent package rather than acting in isolation.
With the complexity of modern social programs (such as health, social security, employment policy, housing and education), it is likely that responsibilities for the oversight and management of social policy will be split across a number of ministers and ministries/agencies. To this can be added the potential involvement of two or more levels of government, as well as the potential involvement of non-government organisations and private sector providers (as in the retirement pension and health care sectors). One approach being increasingly adopted is the use of a lead agency that has coordination responsibility, or alternatively the merging of activities within the one administration. Improved coordination promises much in terms of better program outcomes.
Integrated policy approaches
The struggle of many OECD governments to reduce unemployment demonstrates the need for coordinated action. The promotion of high, sustainable economic growth, withdrawal of impediments to employment growth and other structural measures—such as education, training, other active labour market programs, improved workforce incentives and tight income support arrangements—are all part of the policy mix for lowering unemployment. This is mirrored in the recommendations of the OECD Jobs Study, which recognised that effectively tackling unemployment requires a broad approach incorporating macroeconomic, structural and social policies (OECD 1994b). This is becoming all the more important in OECD labour markets that now have fewer jobs for the low-skilled, where increasing globalisation and structural adjustment is a feature of the economic landscape, and where many OECD governments are struggling to achieve solid sustained outcomes on economic growth.
The working-age jobless in receipt of income support are increasingly being absorbed into active labour market programs and provided with other assistance to help them get into paid employment, rather than just being left on passive income support. The most recent data (OECD 1998f) show that across the OECD region generally, the ratio of labour market spending on active measures to passive income support benefits has increased since the release of the OECD Jobs Study in 1994.16 Sick and disabled people are being provided with rehabilitation, often at an early stage, and participation may be a requirement of income support. In addition, people with disabilities are increasingly seen as a key target group for active labour market program interventions. Lone parents are also being directed to labour market programs, education and training to improve their job prospects. Steps are being taken to improve their access to child care to allow them to participate in paid work; at the same time, access to income support benefits is being restricted.
Young unemployed people have been increasingly excluded from income support if they are simply looking for work and not participating in 'worthwhile' activities, such as education, training or work experience (for example, in Australia, Denmark, Finland, Norway, Sweden and the United Kingdom). These arrangements reduce the likelihood that young people will become passive income support recipients and dependent on welfare from a young age, and promote participation in activities that should enhance their human capital and job prospects. Nevertheless, the effectiveness of this approach also hinges on the availability of suitable interventions for a group that has generally had a poor prior educational experience.
A number of European countries have acted to limit social exclusion17 through an integrated package of measures—sometimes referred to as minimum income guarantees. In their various forms, as applied in different countries, they involve the coordination of social security measures with other measures, such as training, employment and/or full-time education opportunities. The coordination of these measures is designed to achieve the integration of disadvantaged people back into the activities of society. The objective is not just to meet their short-term income needs but also to improve their prospects for longer-term income independence. On average, these people have very poor employment prospects and require very intensive assistance. These programs generally have central government oversight of the elements, which are usually administered by a number of different ministries.18 There is also a corresponding requirement for people to meet certain conditions in return for the assistance. These conditions are often codified in the form of an agreement or contract between the two parties.
The importance of balanced, comprehensive policy responses to encourage and enable older people to work longer, as one of the key responses to ageing populations, was also recognised by OECD social policy and health ministers at their meeting in 1998. The ministers stressed the importance of policies acting on both the demand and supply side, including removing financial incentives in pension systems for early retirement, more flexible working arrangements for older workers, policies to improve the employability of older workers (such as lifelong learning and targeted labour market measures), and achieving more positive attitudes among employers towards older workers (OECD 1998g). To date, there are examples of countries taking some steps in a number of these areas, but few examples yet of countries having a consolidated package of measures to achieve substantial increases in the average retirement age. Long-term care for the elderly is another current policy focus of some OECD countries. This necessarily brings together health care, social services and accommodation arrangements. National governments at the moment are largely concerned with the issues of the overall framework and the funding/financing of these opportunities for the elderly. This includes facets of health care and institutional care, and also ways to facilitate home and community care, including support for informal/family carers of the elderly. In keeping with the broader administrative responsibilities for social services, regional governments and non-profit private organisations each have a significant role in the delivery of long-term care services.
Integrated service delivery
Some countries are strongly pursuing the amalgamation of service delivery, such as through one-stop shops. Australia established Centrelink in mid-1997 to bring together all of the national government functions related to social security, family assistance, assessment of individual labour market needs, referral to a job or employment service provider, and some rural and disaster relief measures. The process in New Zealand to set up a new combined labour market and social security agency is a similar move to establish a one-stop environment for customers, but with slightly different boundaries. The Netherlands is also pursuing this approach, with the joint project involving the public employment service and social security/assistance payment agencies.
The United States, with its long tradition of the states having considerable discretion regarding program elements and administration, provided implementation grants from the federal level to states wishing to establish one-stop centres.19 These centres are required to include provision of six federal Department of Labor programs in their plans (including employment services, unemployment insurance and training programs). Apart from this condition, they have considerable flexibility to provide information on and access to many other services from these centres, such as education, social services and community development projects.20
Over the past two decades, the growth in average incomes has slowed and there is evidence, in at least some OECD countries, that income inequalities and relative poverty rates have been increasing. The proportion of the population in relative poverty is estimated to have increased over the period from the mid-1980s to the mid-1990s in Germany, Italy, the Netherlands and, to a lesser extent, in Japan, Norway and Sweden (Table 4). Overall, the incidence of poverty has been either stable or declined in other countries. The incidence of poverty has also noticeably shifted from the older to the younger populations (young adults and those with children). Changes in income from working have resulted in widening income inequality, which has been partly offset by tax-transfer systems that reduce income inequality.21
More recent research undertaken at the OECD has focused on the incidence of poverty over a longer time period (OECD 1998h). Longitudinal data available for Canada, Germany, the United Kingdom and the United States show that a startlingly high proportion of the population (between 20 and 40 per cent) in these countries experienced at least a year of low income within a six-year period. This group can be divided into two subgroups: the majority who have only short spells on low incomes, and a smaller group who rely on low incomes for a long period. Those on low incomes for at least six years comprised between two and six per cent of the population in these countries. Some particular groups were over-represented among those with long-term low incomes: women, lone parents, elderly single people, sick people and people with disabilities. Obtaining or losing employment and improving earnings were found to be the most significant determinants of change in low-income status over the six-year data period.
Most OECD countries have expressed clear policy intentions to combat social exclusion. In Belgium, a major government report on poverty was completed recently. Ireland has put in place a ten-year National Anti-Poverty Strategy, starting in 1997, which aims to reduce the proportion of the population who are consistently poor by about five percentage points (with a target of below 5–10 per cent). In March 1998, the French Government announced a comprehensive package of measures designed to prevent and combat exclusion.
|Equivalence scale elasticity = 0.5 Changes in percentage points, unless otherwise indicated 50% median income|
1. 'Relative threshold' poverty lines are fixed in terms of real median income in each period.
2. Absolute change is the difference in the value of the index.
Source: Burniaux et al. 1998, Income Distribution and Poverty in Selected OECD Countries, OECD, Paris.
|Head count||Income gap||Sen index2|
|Level (end period)||Changes (absolute)||Changes (absolute)||Changes (absolute)|
|Australia, 1984 to 1993-94||9.5||-2.7||5.0||-0.2|
|United States, 1985-1995||17.1||-1.2||0.2||-0.4|
Social assistance is playing a more significant role in protecting the most vulnerable sections of society, especially those without recent workforce experience who have difficulty establishing an entitlement to social insurance or have exhausted such entitlements (Eardley et al. 1996).
It has become increasingly apparent that a desirable response for these social assistance recipients requires more than just cash assistance (OECD 1998a,c,e). The cash benefits need to provide adequate assistance for the individual or family, while the structure of benefits should not discourage re-employment. The depth of disadvantage experienced by many of the people receiving social assistance means that they may require other social help to deal with problems of homelessness, substance abuse, psychiatric illness, etc. With the strong link between poor labour market outcomes and the growth in social assistance recipient numbers, social assistance programs are increasingly providing access to education and labour market interventions to improve the job prospects of recipients.
As noted above, some prominent policy responses to poverty have been the minimum income guarantee and re-insertion programs introduced in a number of European countries, such as Belgium, France, Luxembourg, Portugal and parts of Switzerland. These schemes, as a general rule, provide social assistance benefits in return for participation in activities designed to improve employment prospects and social integration. In a number of OECD countries, there has been a shift towards greater expenditure on non-cash services—especially employment- related services—for the poor.
Lone parent families are one group with a high incidence of poverty in many countries. Employment rates of lone parents differ substantially across countries, reflecting different assumptions about whether participation in the labour market is expected of them, as well as the availability of affordable child care and quality labour market services. The majority of lone parents are women, many of whom may also have limited recent workforce experience, inadequate education and training, and face low wages when they get work. Prolonged absence from the labour market is increasingly recognised by governments as not in the best interests of either children or the parent, and public policy encourages the re-employment of lone parents in many OECD countries (such as the United States, the United Kingdom, Nordic countries, Australia and New Zealand).
There are recognised limits on the extent to which social programs can deliver sufficient living standards over the long run, given fiscal constraints and the likely effect on incentives for self- provision if benefit rates are too high. Reducing poverty directly through increasing the value of income support payments has a budgetary cost that governments have increasingly found to be difficult, especially in the context of fiscal consolidation.
But communities can also face severe problems if social provisions still leave families and individuals destitute. Inadequate support will lead to social exclusion, with people increasingly losing touch with mainstream society.
Many countries are still grappling with the adequacy/incentive trade-off. Balance is the key, but identifying and gaining agreement on where that balance lies remains a problem. Social research on the level of payments and other supports necessary to deliver an 'adequate' result for households may help to resolve the dilemma, but the critical assumptions underlying the research and the degree of subjectivity attached to these judgements also mean that research alone will not provide a solution.
Considerable progress has also been made in extending universal health care coverage in most OECD countries over recent decades, to the point that it is now commonplace to have universal cover. This can deliver a reasonable standard of health care for everyone, irrespective of their capacity to pay. Very low-income households have also been exempted from many of the patient contribution charges introduced in order to reduce health care usage, because of fears that these charges would discourage some from seeking essential treatment. However, despite universal health care systems, there are still apparent inequalities in health across the population, and concerns over comparatively limited usage of health care services by many poor people who are disproportionately sick compared to the rest of the population.
Social policies have an explicit intergenerational perspective. Benefits tend to be concentrated at the beginning and end of a person's life. For example, families with dependent children are likely to receive considerable net benefits from education and child-care programs, subsidised health initiatives for women and children, and direct cash benefits for their children and/or reduced taxation liabilities. People of working age are likely to be net contributors to the tax- transfer system, except for those who are economically inactive. In retirement, people again revert to being net recipients of social benefits as they access retirement pensions, and may be more intensive users of health care and long-term care services.22
The demographic developments expected in most OECD countries over the next 10–40 years show a considerable change in the composition of the population, with relatively fewer people of working age to support those aged 65 years or more (Table 5). Some governments are concerned that they will not be able to meet the budget costs of pay-as-you-go retirement pension schemes for a growing number of dependent older people, with a growing imbalance between anticipated contributions and benefits in many public pension systems.
In response to expected future pressures on public pension systems as populations age, governments have been introducing a range of pension reforms, such as:
- less generous public pension benefits, through reductions in the maximum final benefit, less generous adjustment of benefits to inflation, and increases in the level of contributions and/or years of employment to generate the same level of benefits;
- pre-funding a portion of public pension benefits (in some countries);
- expanded coverage and reliance on private pension arrangements;
- ncreases in the statutory age of retirement, especially to bring the age for women in line with or closer to that for men;
- reductions in the generosity and accessibility of early retirement pensions.
Despite changes to retirement pension arrangements over the past decade, many of which are being phased in so as not to affect those already retired or close to retirement age, more remains to be done in many countries to improve the fiscal sustainability of public retirement pensions.
As argued more fully in the 1998 report to the OECD Ministerial Council entitled Maintaining Prosperity in an Ageing Society (OECD 1998b), there is also a need to reassess the overall extent of public provision devoted to the older retired population, in view of the reductions in poverty among the older population and the extent of private financial reserves held by some aged people.
|Source: Bos et al. 1994, reproduced in OECD 1996|
In some countries, excessive government transfers to the retired population already seriously constrain the ability of governments to divert public resources to other social policy problems (Esping-Anderson 1997).
There are promising signs that people are not only living longer but are also healthier and more active in their older age, as measured by disability-free life expectancy. Many elderly people have the capacity to contribute to productive activity and caregiving roles well past current notions of the statutory retirement age, which would be enhanced by positive measures such as supporting effective life-long learning and changing community attitudes towards older people.
Nonetheless, there has been a rise in the intractable chronic conditions that accompany an ageing population. This necessitates a new approach to health care, in which the emphasis shifts from diagnosis and treatment of disease to prevention of the diseases themselves. This, along with other pressures on the health system, is leading countries to develop comprehensive, population-based health approaches to promote healthy lifestyles and thereby decrease the incidence of chronic disease. These approaches may also produce substantial benefits for the health outcomes of new generations.
As most health expenditure is already devoted to interventions during the last two years of life, recent projections of the likely expenditure effect on health costs of ageing populations are in the order of a 10–20 per cent increase—a significant but not alarming increase. Some countries are increasing their efforts to devote research to more effective treatments for some specific health problems of older people (such as dementia). This is being coupled with increased vigilance in the areas of evaluating new technologies and overall evaluation of treatment, reflecting countries' interest in supporting cost-effective health interventions.
Long-term care is receiving justifiable attention in countries preparing for a population with a larger proportion who are elderly. While family members continue to have the greatest responsibility for the care of elderly people, the share of responsibilities is changing, and governments are becoming more engaged. Policies are emerging to respond to issues of the availability and affordability of care, as well as respond to increased demands from older people themselves for increased privacy, autonomy and choice. The financing of long-term care is receiving considerable attention in a number of countries, and some have chosen to introduce additional mechanisms to improve the capacity of people to afford the care they need as they grow older.
It is easier to identify the broad approach and elements required for effective social policy reform that responds to ageing populations than it is to define the tactics appropriate for bringing about such changes. The growth of the older age population increases their electoral influence. This has led many countries to be overly cautious and limited with their reforms, especially with their pension systems. However, inaction may lead to increased concerns among the working-age population, who foresee an increased tax burden in the future as well as reduced entitlements when they reach retirement age. Recent experience has also shown that substantive reform packages can be successfully introduced, especially when they are soundly based, well explained, involve extensive community consultation and have broad political support.
Pressures on program expenditure from fiscal consolidation, as well as a general trend to improve the governance of programs funded from public sources, are leading many countries to change the way they manage social programs. This development is evident for governments ranging across the political spectrum, and provides the opportunity for some containment of expenditures while avoiding large reductions in basic social protection.
In an environment of persistent high unemployment and limited employment growth in many OECD countries, there is considerable concern about the increase in welfare dependency among the working-age population. This covers unemployment benefits as well as alternative social security payments, such as sickness benefits, invalidity pensions, lone parent benefits and early retirement pensions. In the early 1990s, around half of the OECD member countries had more people in receipt of invalidity benefits than were receiving unemployment benefits (Blöndal and Pearson 1995). While empirical analysis suggested that the generosity of sickness and disability benefits had little impact on the unemployment rate, it was found that the generosity of these benefits had contributed to lower labour force participation.
Program eligibility is being tightened for a range of payments available to people of working age, with greater attention to ensuring that potential recipients satisfy the core criteria of disability, sickness and unemployment for these respective payments. Some of the main changes in these program areas include:
- removing, or at least reducing, the usage of non-medical criteria in establishing eligibility for disability payments (for example, Australia, Canada, Norway, the Netherlands);
- more stringent assessment processes for disability pension and/or more regular reviews of people on disability pensions (for example, Greece, Italy, the Netherlands, Portugal, the United Kingdom);
- more stringent administration of rules and, in some instances, expanded requirements for eligibility for sickness benefits (for example, Hungary, Belgium, Austria, New Zealand, Spain);
- as mentioned earlier, tighter administration of the job search requirement for people receiving unemployment benefits and, in some countries, also some change to the definition of an acceptable job (for example, Australia, Austria, Canada [selected regions], Denmark, Germany, Portugal, Spain, Switzerland, the United Kingdom).
The outcome from program measures that impact on invalidity pensions is quite mixed. Some countries have been successful in reducing the number of people receiving payments, while other countries have only been able to reduce the rate of new claims, with no noticeable effect on the existing stock of recipients. At this stage, there is not enough evidence to conclude why this difference between countries exists, although possible reasons include variations in the tightness of the respective regimes, variations in the extent to which rigorous rules are actually enforced, variations in the effectiveness of accompanying active policies designed to assist labour market re-integration, and the average duration of prior receipt of payment and exclusion from the labour market.
There are obvious risks if these measures are taken to extremes, as overly rigorous and inflexible rules can lead to people in need being excluded from assistance, contrary to the policy objective of many countries to reduce social exclusion. Some balance and good judgement needs to be exercised by administrators to have effective systems of control while at the same time providing appropriate support for people genuinely in need.
Many OECD countries are giving increased attention to fraud detection and prevention as one key element of administrative change. The focus of these activities generally reflects a broadly based approach that seeks to reduce the extent of all unnecessary expenditures. While available estimates of the likely extent of fraud in most developed social security systems place it around the level of 3–5 per cent of total expenditures, this still represents a significant amount of public money in view of the scale of public social expenditures.
There are many dimensions to the anti-fraud strategies implemented by OECD countries:
- more stringent verification of personal and financial circumstances at the initial application stage;
- special fraud investigation units, often with wide-ranging powers, undertaking selective reviews of mainly high-risk recipients;
- increased cross-matching of computer records held by different government agencies and cooperative ventures between social security and tax authorities;
- a separation of anti-fraud and payment agency functions, in order to emphasise the anti-fraud dimension and more clearly delineate responsibilities.
Increased computerisation is providing greater opportunity to check claims for social security payments and reduces the likelihood of multiple or incorrect payments, supplemented by increased personal review activity. Such approaches have correctly withdrawn assistance from some people who were ineligible—such as people who had not declared that they already had a job—while in other instances people have been directed to more appropriate social security payments. Savings from fraud control have contributed to fiscal consolidation targets and may have limited the scale of action required to meet fiscal targets that would have otherwise reduced the generosity and/or availability of social programs.
Governments are responding to community pressure to improve services at the same time as they are facing pressures to constrain program costs. In most OECD countries, the community has witnessed considerable advances in the range and quality of services over recent years. Undoubtedly, this has influenced public perceptions over the quality of service that should be available from the public sector.
Increasing computerisation and the development of communication technologies have provided many opportunities for cost-effective advances in the administration of social programs. These include quicker and more accurate assessment of program eligibility, more reliable disbursement of benefits, and greater accessibility to assistance for people who reside outside major population centres.
Use of the non-government sector
There is increased use of the not-for-profit and voluntary sector to tap into the networks and experience of other agencies already heavily involved in the delivery of social services. This can entail block funding of these agencies, payment from government for the provision of specific services, or active partnerships with government—each of these models has been used effectively in OECD countries. In some instances, countries have resorted to greater competition to provide greater consumer choice and pressure for cost reductions. However, there are also limits to the extent to which government can outsource core social protection activities and maintain a nationally consistent level of protection.
Private social expenditure is more significant in a number of areas, such as health and retirement pensions, as some people seek improved quality of provision over and above that available from the public system (Adema and Einerhand 1998, Esping-Anderson 1997). In some social policy areas—such as health and education—there can be considerable competition between the public and private sectors, such as in the United States. Australia has introduced a competitive market for employment services, in association with a substantive change to the public employment service so there is fair competition between the respective agencies.
In the context of greater reliance on the private, not-for-profit and voluntary sectors to deliver social programs, the quality of regulation has become increasingly important as a means of providing some form of external accountability. This may involve codes of practice and professional standards expected of non-government agencies, such as those providing health care and long-term care services. In other cases, such as private retirement pensions, extensive regulatory and supervisory mechanisms should be in place, reflecting the scale and importance of the investments for the future wellbeing of the contributors.
Greater attention to quality
Commitment to public standards of services (for example, the Citizens Charter in the United Kingdom) can provide a yardstick for judging performance and managing public expectations, as well as greater transparency of performance. This approach has also been seen in (for example) health care reform, where a majority of the OECD countries now have national commissions on quality standards, and several countries are beginning to develop a 'consumer bills of rights'.
Some progress has been achieved in making governments more client-oriented in several areas, such as improved access to governmental services, transparency of operations and decision- making, and provision of grievance procedures to provide a feedback mechanism for government services. One-stop shops have been pursued in some countries as a means of achieving greater integration of services, including across levels of government. In the social protection area, a separation between agencies providing services and policy development can focus the energies of the former on the improvement of services (for example, in Australia, New Zealand and the United Kingdom). However, the formal separation of the policy making and policy execution functions has some risk that policy could become out of touch with reality, unless mechanisms for providing feedback from operations are strengthened.
Improving the quality of care and services in the health care arena is receiving greater attention. Countries are increasingly focusing on improving the return on their investment, rather than simply pursuing price and supply control measures to restrain expenditures in their health care systems. Competition has been introduced or expanded in several countries to encourage players in the health care arena to increase the adaptability of their systems to attract more clients. Countries are also transferring health services to the local level to increase the flexibility of care. The decrease of acute conditions of the elderly, and the expanded role of social services, has been accompanied by a shift to a more client-focused orientation as consumers and their families demand more from service providers.
Devolution and decentralisation
Another distinct trend is towards 'devolution' and/or 'decentralisation', which aim at more efficient and effective public administration.23
Within the health and welfare sectors, there are many examples of decentralisation of social programs (OECD 1998e, OECD 1997d). The growing importance of means-tested social assistance in the broad scheme of social security necessarily introduces greater emphasis on local control and responsibility, given the way social assistance is structured in a number of OECD countries.24 Greater regional responsibility is also being introduced for housing assistance in Mexico and Poland, and Ireland is also actively considering this shift in responsibilities from the national government to the regional level. In the 1992 Ädel Reform in Sweden, management of long-term care institutions and day care facilities was shifted from county councils to municipalities, with the possibility of a similar shift of authority for visiting nurse services and other medical care services. Responsibility for funding care for the elderly in county hospitals has been shifted from the counties to the municipalities.
The Personal Responsibility and Work Opportunities Act of 1996 in the United States transferred more responsibility for welfare provision to the states. In exchange for accepting a fixed block grant over a five-year period, the states were given much more flexibility to determine the eligibility of recipients and how to administer the benefits, subject to meeting minimum federal requirements. Additionally, the states themselves could devolve responsibility by contracting with private companies for a broad range of services, including eligibility determination and work retraining programs. As welfare numbers have fallen over recent years and income support payments declined, the states have found themselves with much greater scope to provide active labour market intervention for welfare recipients within their budget allocations. In the mid-1990s, Sweden experimented with a variant of block funding and coordinated responsibility for people on sickness benefits (the FINSAM experiment). Five counties were able to direct notional social insurance funds to increased rehabilitation effort, with the intention of reducing incapacity rates. The results were positive, producing reduced social expenditures on the sick in these regions.
Monitoring and evaluation
Overall, there is insufficient investment in the research and evaluation of social programs, especially compared with the overall level of program expenditures. Without comprehensive analysis and evaluation to back up policy deliberations, there is a considerable risk that governments may divert scarce public resources to poorly performing programs and not act quickly to amend programs in ways that improve their effectiveness.
At an international level, there is a good understanding of the social policies operating in different countries, but this now needs to be complemented by a shared understanding of the outcomes and effects of these policies. The OECD is well placed to contribute to that process, especially in terms of developing common frameworks for the assessment of social policies and monitoring reforms, but individual countries also need to have robust monitoring and evaluation cultures to complement any international activity.25
It is evident that there has been a degree of convergence between many countries in the direction of social policy reforms over recent years, and this paper has sought to highlight some of these major trends. This should not be interpreted as meaning that these countries have similar social programs. There are many remaining differences, which at least partly reflect historical, social and cultural factors.
Further reforms to social programs are required in many countries. There is considerable evidence of a staged and cautious approach to social policy reform. Substantive social policy reform is possible, even in some potentially contentious areas, but this requires clear enunciation of the justification for change, and clear exposition of the path of reform and its implications. Experience also shows that political consensus is an important ingredient in the implementation of major reforms. Social programs will need to respond to new and emerging challenges. In this vein, it is likely that welfare systems in OECD countries have not yet been fully adapted to reflect the growing number of double-income families. Two regular income earners can provide the family with their own economic security and can reduce the risk that unemployment, disability or death will have drastic economic and income effects on the family unit. With a greater proportion of families reflecting this work pattern, there may be some pressure exerted by families who wish to opt out of some forms of social protection. The one caveat is that those with two incomes may become accustomed to a higher living standard, and still seek income protection via public arrangements.
- This paper draws extensively from information collected for a survey of social programs and recent social policy reforms in OECD countries (Kalisch et al. 1998, OECD 1998e). This OECD work responded to the 'Initiative for a Caring World' proposed by the former Japanese Prime Minister, Mr Hashimoto, at the Lyon G7 summit in June 1996, where he called for an extensive sharing of experiences among OECD countries in their social policy activities.
- The level of public expenditure is determined by a number of factors: the generosity and expansiveness of social provisions, and economic dimensions that influence the need for and take-up of social programs.
- In Sweden, previous cuts to social security are now being partially reversed as the budget has come into balance.
- The latest OECD social expenditure data (from about 1993 to 1995 or 1996) are now showing a few more countries with recent limited declines in total non-health social expenditures, such as in Australia, Canada, Spain, Finland, Ireland, Italy, Netherlands, Norway and Sweden.
- The OECD has recently increased the number of countries for which this data is available, in the OECD Social Expenditure (SOCX) database, and is in the process of extending the information to include comprehensive estimates for Australia.
- Atkinson (1993) and Ingles (1997) outline some of the adverse effects from excessive targeting of benefits available for families.
- This issue, while of interest, has little policy application to Australia where employers are already largely responsible for the ongoing income of employees while they are sick, through the system of occupational sick leave entitlements.
- Recent OECD work on resources in retirement also suggests that there is already considerable substitution between public pensions and other marketable wealth available to retirees (OECD 1999, p. 116). Retirees have relatively consistent levels of overall retirement income across countries (at around 70–80% of previous salary for an average wage earner), despite clear differences in the importance of the sources (public pensions, private pensions and other savings) of their retirement income.
- There are three main types of cost sharing in place: copayments, where a patient pays a set amount for each service; coinsurance, where patients pay a set proportion of the total cost of services; and deductibles, where patients pay the first set amount and health insurance pays the rest.
- In most OECD countries, the share of the population aged over 65 years in institutions varies between 5 and 7 per cent and has not changed much over recent years.
- Incentives can be even lower than is presented here once some consideration is given to the additional costs incurred by those who work (which is thought to be greater than job search costs) and the loss of leisure time.
- This is a very simplified explanation of replacement rates, which can be constructed from the perspective of the financial incentives for the employed to become unemployed or the perspective of the financial incentive for the unemployed to find work. Martin (1996) provides a useful discussion of replacement rates and their use for the OECD Jobs Study.
- A number of these countries also have other benefits and/or tax relief for low-income families (for example, New Zealand, Italy).
- Spain's Employment Credit is payable to all eligible earners irrespective of family composition, while the United Kingdom is currently conducting an experiment in selected regions of an in-work benefit for people without children called Earnings Top-Up. The Netherlands has also recently announced its intention to introduce an income tax credit for all people in employment with earnings over a very low level.
- In their 1992 meeting, OECD social policy ministers proposed new orientations for social policy, with one of these orientations being policy coherence through 'a renewed focus on the means by which the strands of policy—from setting goals, to formulating policies, implementing them, and, thereafter, administering programs—can be pulled together across social, labour market, education and training, and economic policies and across levels of government' (OECD 1994a).
- There is no clear evidence of an increase in spending on active measures as a proportion of a country's GDP since 1994.
- Social exclusion is a term that is increasingly used in OECD countries to encapsulate aspects of poverty and low income, but also include the dimension of non-participation in usual activities, such as access to work, cultural activities and health care services.
- One exception is in Switzerland, where the selective latin cantons that have introduced these types of programs were the main driving force behind the initial development and administration of the reintegration measures.
- Planning and implementation grants have been available since 1994, on the basis of bids by states in three rounds of competitive tendering, with the funding provided through the US Federal Department of Labor.
- The variation between one-stop centres across different states provides an opportunity to assess the merits of particular approaches, which vary in terms of coverage of services and the degree of integration of service delivery within the centre.
- The latest analysis by the OECD shows that transfers to the elderly generally have no impact on reducing inequality, explained partly by the strong link from prior earnings to pension payments in many old-age pension schemes.
- Falkingham and Harding (1996) provide one example of an attempt to measure the lifetime distribution of social security and taxation arrangements, for Britain and Australia.
- Devolution refers to the shift of responsibilities within the central government, whereas decentralisation refers to the shift of responsibilities to local levels of governments.
- The decentralised nature of social assistance in Sweden, Switzerland, Poland and Italy is being modified, to some extent, by the introduction (or possible introduction) of common national standards for social assistance.
- Some recent examples where the OECD has played a valuable monitoring and assessment role include the social assistance reviews, and the follow-up to the OECD Jobs Study. The OECD is in the process of establishing a process for monitoring progress with retirement pension reforms, as well as developing a broad range of comparative social indicators.
Adema, W. 1998,'What do countries really spend on social policies? A comparative note', OECD Economic Studies, no. 28, 1997/1, OECD, Paris.
Adema, W., Einerhand, M., Eklind, B., Lotz, J. & Pearson, M. 1996, Net Public Social Expenditure, Labour Market and Social Policy Occasional Paper, no. 19, OECD, Paris.
Adema, W. & Einerhand, M. 1998, The Growing Role of Private Social Benefits, Labour Market and Social Policy Occasional Paper, no. 32, OECD, Paris.
Atkinson, A.B. 1993, On Targeting Social Security: Theory and Western Experience with Family Benefits, Welfare State Programme Discussion Paper, no. WSP/99, Suntory-Toyota International Centre for Economics and Related Disciplines, London School of Economics.
Blöndal, S. & Pearson, M. 1995,'Unemployment and other non-employment benefits', Oxford
Review of Economic P olic y, 11(1), pp. 136–69.
Burniaux, J.-M., Dang, T.T., Fore, D., Förster, M.F., Mira d'Ercole, M. & Oxley, H. 1998, Income Distribution and Poverty in Selected OECD Countries, Economics Department Working Paper, no. 189, OECD, Paris.
Eardley, T., Bradshaw, J., Ditch, J., Gough, I. & Whiteford, P. 1996, Social Assistance in OECD Countries: Synthesis Report, Department of Social Security Research Report, no. 46, HMSO, London.
Esping-Anderson, G. 1997,'Welfare states at the end of the century: the impact of labour market, family and demographic change', in OECD 1997b, Family, Market and Community: Equity and Efficiency in Social Policy, OECD, Paris.
Falkingham, J. & Harding, A. 1996, Poverty Alleviation Versus Social Insurance Systems: A Comparison of Lifetime Distribution, NATSEM Discussion Paper, no. 12, University of Canberra.
Ingles, D. 1997, Low Income Traps for Working Families, Centre for Economic Policy Research
Discussion Paper, no. 363, Australian National University, Canberra.
Kalisch, D.W., Aman, T. & Buchele, L.A. 1998, Social and Health Policies in OECD Countries: A Survey of Current Programmes and Recent Developments, Labour Market and Social Policy Occasional Paper, no. 33, OECD, Paris (provided to OECD social policy and health ministers at their meeting in June 1998 under the title The Caring World: National Achievements).
Martin, J.P. 1996,'Measures of replacement rates for the purpose of international comparisons: a note', OECD Economic Studies, no. 26, 1996/I, pp. 99–115.
Mead, L.M. 1997,'Raising work levels among the poor', Social Policy Journal of New Zealand, Issue Eight on Beyond Dependency, pp. 1–28.
OECD 1994a, T he Future of Social Protection, OECD, Paris.
OECD 1994b, T he OECD Jobs Study: Facts, Analysis, Strategies, OECD, Paris.
OECD 1996a, Ag eing in OECD Countries: A Critical Policy Challenge, Social Policy Studies, no. 20, OECD, Paris.
OECD 1996b, Caring for F r ail Elderly P eople: P olicies in Ev olution, Social Policy Studies, no.19, OECD, Paris.
OECD 1996c, Enhancing the Ef f ectiv eness of Activ e La bour Market P olicies, OECD, Paris.
OECD 1997a, Making W o r k P a y: T axation, Benefits, Employment and Unemployment, OECD, Paris.
OECD 1997b, F amil y , Market and Community: Equity and Ef f icienc y in Social P olic y, Social
Policy Studies, no. 21, OECD, Paris.
OECD 1997c, Implementing the OECD Jobs Strategy: Member Countries' Experience, OECD, Paris.
OECD 1997d, Managing Across Le v els of Government, OECD, Paris.
OECD 1998a, T he Battle Against Exclusion: Social Assistance in A ustr alia, F inland, Sw eden and the United Kingdom, OECD, Paris.
OECD 1998b, Maintaining Prosperity in an Ag eing Society, OECD, Paris.
OECD 1998c, T he Battle Against Exclusion: Social Assistance in Belgium, the Czech Re pub lic , the Netherlands and Norway, OECD, Paris.
OECD 1998d, Benefit Systems and W o r k Incentives, OECD, Paris.
OECD 1998e, T he Caring W o r ld: An Analysis, paper prepared for the June 1998 meeting of OECD social policy and health ministers, subsequently published under the heading A Caring World: the New Social Policy Agenda, OECD, Paris, 1999.
OECD 1998f, Employment Outlook, OECD, Paris.
OECD 1998g, Communiqué of the Meeting of the Employment, Labour and Social Affairs
Committee at Ministerial Level on Social Policy, June 23–24 1998, SG/COM/NEWS(98)70, Paris.
OECD 1998h, Economic Outlook 64, December 1998 (Chapter VI:'Low income dynamics in four OECD countries'), OECD, Paris.
OECD 1999, OECD Economic Surveys: 1998–99: Australia, OECD, Paris.
Preston, D.A. 1997,'Welfare benefit reform', Social Policy Journal of New Zealand, Issue Eight on Beyond Dependency, pp. 29–36.
Saraceno, C. 1997,'Family change, family policies and the restructuring of welfare', in OECD
1997b, F amil y , Mar k e t and Community: Equity and Ef f icienc y in Social P olic y, OECD, Paris.