52000DC0163

REPORT ON SOCIAL PROTECTION IN EUROPE 1999

(presented by the Commission)

REPORT ON SOCIAL PROTECTION IN EUROPE 1999

TABLE OF CONTENTS

1. The changing context of social protection systems

2. Trends in social expenditure

3. Financing social expenditure

4. Recent changes in policy

4.1. Key objective: to make work pay and to provide a secure income

4.2 Key objective: to make pensions safe and pension systems sustainable

4.3 Key objective: to promote social inclusion

4.4. Key objective: to ensure high quality and sustainable health care

5. Data issues

6. Social protection in the candidate Countries of Central and Eastern Europe

Executive summary

This is the first report following the European Commission proposal in July 1999 to set up a "Concerted Strategy for Modernising Social Protection" (COM(1999) 347), an initiative that should be seen in the context of continuing developments in the political, legal, economic and social framework within which social protection policy operates, both at the Member State and European Union level. Rapid progress towards full economic and monetary union, the concerted employment strategy and Union enlargement have significant implications for social protection and make it increasingly a matter of common concern for Member States and, therefore, a matter for closer co-operation at the European level.

The responsibility for organising and financing social protection lies with Member States. However, the instruments of the EU policy co-ordination and surveillance, such as the Broad Economic Policy Guidelines and the Stability and Growth Pact, constitute broad frameworks also for social protection and employment policies through setting targets for budgetary consolidation and providing guidelines for public policies. The new policy-oriented initiative of social policies, which takes into account the changing economic context and the need for overall policy co-ordination, aims at strengthening the co-operation in this area in parallel with, and complementary to, the employment process. The 1999 Communication identified four broad objectives to which social protection systems should respond, i.e.:

- to make work pay and provide secure income;

- to make pensions safe and pension systems sustainable;

- to promote social inclusion, and

- to ensure high quality and sustainability of health care.

Within the overall framework, a new Social Protection report will assume a key role in monitoring recent developments and thereby in assisting Member States in policy deliberations. In order to be able to keep track of the reforms taking place in national systems, the Commission proposes to publish the report annually in future, in a form still to be discussed with Member States. The present report, in consequence, should be seen as an intermediary one. Nevertheless, it focuses on the four broad key objectives mentioned above.

The report reviews recent developments in social protection expenditure and receipts, and as such, updates previous reports. The limitations for policy analysis of the data currently available at the European level, from ESSPROS (the European System of Integrated Social Protection Statistics) and the new European Community Household Panel (ECHP), are also discussed. The report points to the need for improved and, above all, more timely data if developments in social protection policies are to be analysed properly in the future.

Gender issues are addressed as an important issue which cuts across all four objectives, and the report emphasises the need for social protection systems to respond better to changing gender realities, particularly the need to support increased participation of women in the work force. The report also examines, for the first time, social protection systems and the main developments in these in recent years in the Central and Eastern European countries which have applied to join the European Union.

The report, in addition, contains new empirical findings based on results from the first two waves of the ECHP of what men and women actually received in 1993-1994 when they were unemployed compared with their take-home pay when they were working. These findings, which aim at contributing to the ongoing debate on the financial disincentives to work, should be interpreted with some caution, compared with evidence from other studies and will need to be further investigated as data from new waves become available for analysis.

To make work pay and provide secure income

A prominent feature of policy across the Union in the past few years has been a widespread re-examination of conditions for benefit payment. As a result, measures have been taken in many Member States to tighten the qualifying conditions for eligibility to benefits and to increase the activation of the unemployed to find a job. Measures have not been confined to unemployment benefits. In particular, changes have also been applied to disability and early retirement schemes, which had come to be used extensively as effective substitutes for unemployment compensation in a number of countries There have also been extensions of these schemes in the countries with the least generous systems. In order to increase the flexibility between labour force participation and retirement, several countries have increased the scope for more flexible retirement, such as through part-time retirement schemes. Changes in benefit systems have also been combined in some countries with attempts to make work more rewarding financially by introducing measures to increase the net income of those taking up a job, both through allowing them to continue to receive some support after they begin to work and by extending in-work benefits.

The reports examines changes in family-related benefits, one of the major aims of which is to provide both financial and practical support to men and women with children in balancing their professional and family responsibilities. The provision of family-related support facilities is also a necessary complement to active labour market measures. Support services for child care and for people with disabilities or who are elderly and frail are of the utmost importance for improving the income security of people with families, especially women. Availability of child-care support has been increased in recent years in a number of Member States, especially in those where it has been limited, while parental leave entitlement has been raised throughout the Union, which is particularly important to enable people with young children to pursue working careers and to facilitate the new gender balance in the labour market.

To make pensions safe and pension systems sustainable

The significant growth in the number of people above retirement age in all Member States in 10-15 years time has focussed attention everywhere on the consequences of this for old-age pensions systems. A new balance is being sought between the aim of providing sufficient and safe income to the elderly and that of making systems sustainable. While the measures introduced differ in detail, in practice, their common emphasis is on limiting the future transfers which will be necessary, particularly those which governments are responsible for, or increasing the finance available. They include increasing the official age of retirement or the number of years of contributions required to qualify for a full pension, reducing the pension paid in relation to past earnings, or relating it more closely to contributions, and creating special funds to finance future transfers. The latter amounts to adding beside existing 'pay-as-you-go' systems, where present payments are financed from present contributions, 'funded' ones where present contributions are fixed in relation to future liabilities.

To promote social inclusion

A major feature of social policy in recent years has been the emphasis on active and preventive measures as well as adapting social benefit systems to make work pay designed to get men and women into work. This has been motivated, on one hand, by the aim of increasing the numbers contributing to the generation of income rather than being supported by that produced by others, and on the other hand, by the aim of reducing dependency and social exclusion. Although the main emphasis is on employment, since it is recognised that having a stable job is an important condition for social inclusion, it is also recognised that there are other aspects to be addressed if social exclusion is to be tackled effectively. These include access to education and vocational training, decent housing and sufficient levels of health care and social services. In several countries, recent measures have actually been taken to ensure better access to health care to low income households. In practice, despite the emphasis on active measures throughout the 1990s, there has been relatively little shift in actual spending from passive to active measures. The amount spent on active programmes per person unemployed was the same in 1998 in relation to GDP as in 1994 and less than in 1990. This indicates that there is still scope to extend opportunities to benefit from active measures, especially towards the most disadvantaged groups.

To ensure high quality and sustainable health care

The pressure on social protection systems across the Union has manifested itself in a growing need for health care, reinforced by the effect of rising real income and advances in the range of treatment which can be provided in pushing up demand, and so putting substantial upward pressure on expenditure. The common response of governments has been to seek to limit the growth of spending and its effect on public budgets, whilst seeking to maintain the quality of service and access to care. This has naturally led to a focus on ways of increasing productivity or improving cost effectiveness.

All European social protection systems have to fulfil common broad objectives. They also face common challenges and need to be reformed to adjust to them. This report is a contribution to support the exchange of experiences and the ongoing debates on modernising and improving social protection systems, and thus, complements the recent 1999 Employment in Europe Report. The new political, legal and economic developments at EU level have increased the need for economic and structural policies to interact and reinforce one another in order to support the European social model. In this framework, social protection systems, together with employment policies, should act as key productive elements in order to support economic progress, a high level of social protection and to maintain social cohesion.

The European Commission has launched a new initiative in relation to social protection by suggesting a "Concerted Strategy for Modernising Social Protection" [1]. Systems of social protection throughout the Union are undergoing a process of change in response to the major demographic, economic and social developments which have occurred in recent years in all Member States. These pose new challenges to governments in meeting the twin objectives of providing an extensive and high level of assistance to people in need of support and of ensuring that their economies remain competitive and capable of generating sustained, and sustainable, improvements in living standards and a high rate of net job creation.

[1] "A Concerted Strategy for Modernising Social Protection" (COM(1999) 347).

There is a variety of reasons for the Commission to launch a new initiative now: the completion of EMU through the establishment of a single currency, including the requirements for budgetary discipline and sound public finances, the entering into force of the Amsterdam Treaty putting the European Employment Strategy and the fight against social exclusion on a new legal basis and the intensive preparations for Union enlargement. These change the political, legal, economic and social framework within which social protection policy operates at both Member State and Union level and mean that developments in systems are a matter of common concern, requiring closer co-operation between Member States.

The aim of this new, policy-oriented, initiative is to strengthen co-operation in this area in parallel with, and complementary to, the process in respect of employment. Policies on social protection are a key adjunct to the European Employment Strategy and a major means of helping Member States to pursue the measures set out in the Employment Guidelines which are the practical manifestation of the Strategy. A number of the 1999 Guidelines entail, either explicitly or implicitly, changes to systems of social protection and the Commission's proposals for the Guidelines for 2000 retain or reinforce this approach.

The Communication identifies four key objectives, designed both to secure a high level of social protection which people in the Union can count on when they are in need and to reinforce the economic and employment strategy. They are:

- to make work pay and provide secure income;

- to make pensions safe and pension systems sustainable;

- to promote social inclusion, and

- to ensure high quality and sustainability of health care.

The present Report focuses on changes in systems of social protection which have either been made in past two or three years or which are in the process of being introduced, in the light of those broad objectives, and as they are central to the modernisation of social protection systems. As such, it updates the analysis contained in the previous three Social Protection in Europe reports (for 1993, 1995 and 1997).

At the same time, the report devotes some attention to financial aspects, such as the level, growth and structure of expenditure on social protection, the different sources of revenue and the interlinkages between financing social protection and employment policy, as well as the level of benefits and incentives to work.

Gender issues, as made clear in the Communication, cut across all four objectives, and the present report emphasises the need for social protection systems to respond better to changing gender realities, particularly the need to support the participation of women in the work force. In addition, the report, for the first time, examines social protection systems in Central and Eastern European countries (CEECs) and recent developments in these.

Outline of the Report

The present report summarises the main findings of the full Report of the Commission Services.

The report, first, examines the principal demographic, economic and social developments which shape both the context within which social protection systems operate across the EU and the demands made upon them (in Chapter 1 of the Report of the Commission Services).

Secondly, it analyses the scale of expenditure on social protection in the different Member States, its composition and the changes which have occurred over the 1990s (in Chapter 3 of the Report of the Commission Services).

Thirdly, it examines the relative importance of different sources of finance for spending and the extent to which they have altered during this period (in Chapter 3 of the Report of the Commission Services).

Fourthly, it reviews more largely the changes which have been made to systems of social protection across the Union over the past two or three years in relation to the key objectives set out in the Commission Communication. It focuses, in particular, on those which are designed to support the employment strategy, to reduce the number of people dependent on benefits and to expand the number in work. It also examines the evidence on the level of unemployment benefits relative to earnings from work and assesses the extent of any financial disincentive which this may create for not actively looking for a job (in Chapters 2 and 4).

Fifthly, it gives an overview of developments in social protection systems in candidate CEECs, drawing on recently completed studies carried out under the PHARE Consensus programme (Chapter 5).

1. The changing context of social protection systems

The major demographic, economic and social changes over the past 20-25 years across the Union have had profound implications for social protection systems and will continue to do so in future years. Four major trends are of particular importance: the ageing of the population in all Member States and the prospective increase in the rate of growth in the number of elderly people from 2010 onwards; the growing participation of women in the labour force and the changing gender balance; the persistence of long-term unemployment, especially among older workers, and the trend towards earlier retirement; and the increase in the number of households in relation to the growth of population, with the rise of people living on their own and of households with no-one in work.

These trends, moreover, are occurring in a context of increasing globalisation and a faster pace of technical advance, which have accelerated the speed of structural change in the economy, put greater pressure on businesses and workers alike to adapt to changing market circumstances and new methods of working and reduced the ability of governments to manage economic developments. At the same time, they have increased the importance of securing financial stability and have, therefore, limited the scale of government borrowing and constrained the growth of public expenditure. Public policies are directly affected by fiscal consolidation required by the Stability and Growth Pact, as well as by the Broad Economic Policy Guidelines which provide a framework for reforming public policies in line with broader objectives, in particular, in support of employment and job creation. These evolutions have concentrated policy attention on social protection, which represents a large component of overall spending.

Demographic trends: the implications of an ageing population

The growth in the number of people of 65 and over is set to accelerate over the next 10-15 years as the post-war baby boom generation reaches their late 60s, so imposing increased pressure on pension systems. At the same time, the even larger expansion in the number of people of 75 and over will increase demand for long term care markedly, just when increased female participation in the workforce is also likely to reduce the available pool of unpaid family carers.

Whereas, at present, there is on average in the Union, one person aged 65 and over for every four people of working age (the old-age dependency ratio), this is projected to rise to just under one in three by 2020 and to almost one in two by 2040. In contrast to the past, this growth in the old-age dependency ratio is unlikely to be offset very much by the continued decline in the number of children and young people under 15, which is forecast to decline more slowly than the number of people of working age. The overall dependency ratio which has fallen slightly in the Union over the past 20 years, is projected to recover by 2010 and then increase markedly, from its present ratio of 5 people aged under 15 or 65 and over per 10 of working age, to 6 per 10 by 2025 and 7 per 10 by 2040.

These prospective increases in dependency ratios, which are the conventional way of assessing the implications of demographic trends, are liable, however, to give a misleading impression of the scale of the problem which they pose for systems of social protection and of the most appropriate policy response to them. So far, attention in most Member States has focused on curbing the growth of pension commitments and trying to ensure that there will be sufficient funds available to effect the transfers involved. While this is important, it is equally important to take account of the fact that a large proportion of people of working age are not, in practice, in employment and, therefore, play no role in generating the income from which pensions for those in retirement have to come. Instead, they rely, like those of 65 and over on support from those who are working. In other words, any assessment of the prospective difficulties of funding social transfers in future years has to distinguish between the potential work force - those aged 15 to 64 - and the actual people in employment.

In 1998, only just over 60% of working-age population were in paid employment. The rest were either unemployed (around 7%) or (the great majority) not in the labour force at all, most of them women but around a third of them men. Among men, a major cause is early retirement, among women, other factors are also important such as earlier barriers to female participation and caring obligations. In consequence, the effective dependency ratio - the number of people aged 15 and over not in work - was substantially higher than the conventional ratio cited above, even if children under 15 are left out of account. In fact, slightly more people in the Union of 15 and over are not in employment than those who are, and over 60% of them are under 65. This ratio was much the same in 1998 as in the mid-1980s, as the fall in unemployment and rise in women's participation has more than offset the increased number aged 65 and over. If the decline in the number of children is also taken into account, it means that the overall effective dependency ratio (everyone not in work, including those under 15, relative to those employed) fell over this period.

This illustrates the potentially misleading impression gained by focusing solely on the old-age dependency ratio and demonstrates the importance of taking due account of the number of people in employment which will be critical to overcoming the problems created by supporting an ageing population. In other words, the success of economies in effecting the transfers which will be necessary to ensure a decent standard of living for those of 65 and over ultimately depends on their success in increasing the proportion of people of working-age actually in work and, therefore, on their performance in creating jobs and wealth.

Participation of women in the work force

One of the main features of social change in the Union over the past two decades has been the increase in the relative number of women pursuing working careers. In practice, women account for all, or nearly all, of the expansion in jobs which has occurred over this period in all Member States.

The biggest increase in participation has been among women of child-bearing age, but there remain substantial differences in participation rates for this group between Member States, partly explained by differences in underlying labour market conditions. However, the statistics suggest that female participation is related both to marital status, especially in Southern countries where participation of married women is much lower than that non-married ones, and to whether or not women have young children, particularly in Northern Member States, where availability of child-care support outside the extended family is decisive.

The effect of the apparent deficiency in child-care support in a number of Northern Member States shows up most notably in the much lower rate of participation of single mothers than married mothers. This is particularly the case in the UK and Ireland, where under half of single women aged 25 to 49 with a child under 5 were in the work force, much less than anywhere else in the Union and well below the rate of participation of married women in a similar position. There is, therefore, an evident need in many Member States for an extension in the availability of child-care facilities as well as for support arrangements to help women caring for disabled or frail elderly family members.

As well as the increase in the participation of women, the kind of job they do is also important. There is a strong trend across the EU for both sexes towards part-time working as well as towards temporary jobs. Women, who represent the main potential source of future labour force growth, tend to be concentrated in part-time jobs (around 33% of women in work in 1998 worked part-time). Yet many part-time jobs are effectively excluded both from the obligation to pay social contributions and from entitlement to benefit. These developments, therefore, have important implications both for the financing of social protection systems and for the availability of income support.

The information presented in this report emphasises the challenges posed for social protection systems by the need to respond better to the complexity and diversity of gender relations in the modern European social economy. Most systems are still based primarily on a male breadwinner model which does not reflect the new realities: increasing participation of women in the labour force which needs to be further encouraged, more diverse models of child, and family, care requiring appropriate support, the dependence of a growing number of households on the earnings of women, more diversity in patterns of work for both men and women, a reduction in the effectiveness of derived rights as a means of ensuring support because of the decline in traditional family structures, as a result particularly of the rise in divorce and separation, and widening concern about the persistence of barriers and disincentives to women pursing working careers.

The high level of early retirement and the persistence of long-term unemployment

Inadequate growth of employment in the Union since the mid-1970s has led to both a persistently high level of unemployment and a low rate of participation in the work force among older people, as well as low participation among women. This poses serious problems for both systems of social protection and labour market policy.

Although the official male retirement age in most Member States is presently 65 or over, there has been a slow but persistent decline in the number of men aged 55 to 59 in the work force since the mid 1980s and a more marked decline for men aged 60 to 64. In 1998, under 70% of men and only 40% of women aged 55-59 were economically active. Only 32% of men, and just 15% of women aged 60-64 were still in the workforce. Reversing the trend towards early retirement among men and increasing the number of older women who are economically active has to be a major aspect of any policy aimed at relieving the pressure on systems of social protection.

Large numbers of older people are also long-term unemployed, a state which often precedes early retirement. In 1998, over 8 million people in the EU had been unemployed for a year or more, some 5% of the labour force. Over 5 million of these had been out of work for at least two years. These figures are similar to those at the end of the recession of the early 1990s, and not much lower than in the mid-1980s. The high rate of long-term unemployment is a major source of pressure on systems of social protection across the Union, as well as of social exclusion. It is clear, however, that the nature of the problem differs between different parts of the Union, which means, in turn, that the appropriate policies also differ.

The changing structure of households and the growth of those with no-one in work

There is a continuing increase in the number of households across the Union, despite the low growth of population, and an even larger increase in the number of people under 65 living alone, a disproportionate number of these being out of work. The decline in the average size of household implies a growing demand for social services since it makes it more difficult to provide care and support from within the family.

In the Union as a whole, just under 25% of households in 1998 consisted of only one person of 15 and over, and in some 16% of these, there was a child under 15, virtually all of them being brought up by lone mothers. Both figures were significantly higher than 12 years earlier. There is, however, a marked difference between the prevalence of single-person households, as in the average household size, between the North and South of the Union. In Finland, Germany, the Netherlands and the UK, some 30% or more of households consisted of only one person of 15 or over, 2-3 times the proportion in Southern Member States.

People of working age living alone are, moreover, far more likely to be unemployed in the North of the Union than those living with other people. In the Netherlands, Germany, Belgium and the UK, for example, 25% or more of men aged 15 to 64 who were unemployed lived alone, twice the proportion of those in employment. In the Southern Member States, only around 4% or less of men in this age group who were unemployed lived by themselves, much the same as in the case of those in employment. In most Northern countries, moreover, the figure for those unemployed has increased since the mid-1980s as well as during the most recent period when unemployment has been falling.

The relative pressure imposed on systems of social protection by unemployment, or, more generally, by people not being in work, is, therefore, not fully reflected in the figures for unemployment rates. In the South of the Union, high rates of unemployment combined with low levels of labour force participation impose much less pressure than they would in most Northern Member States, because of the larger size of households and the support available to most of them from within the family. The trend towards smaller households and more people living alone means, therefore, that falls in unemployment do not necessarily give rise to proportionately lower needs for social and financial support. This is illustrated most forcibly in Belgium and the UK, where lower rates of unemployment have not been accompanied by a corresponding decline in the number of households with no-one in work.

2. Trends in social expenditure

In relation to output, overall spending on social protection in the Union, gross of any taxes or charges levied on benefits, amounted to 28½% of GDP. The level of spending ranged from over 30% of GDP in France, Belgium, Germany, Netherlands and the three Nordic countries, with Sweden having the highest level at around 35% of GDP, to 22-23% of GDP in Greece, Spain and Portugal and only 18½% of GDP in Ireland. In very broad terms, the countries with the highest level of GDP per head tend to spend proportionately more on social expenditure, but the relationship is by no means systematic, with expenditure in Sweden and Finland, in particular, being higher than would be expected on this basis and expenditure in Italy (25% of GDP) and Ireland being lower.

Net social spending

The figures for expenditure on social protection, gross of taxes and social charges, however, tend to overstate both the cost entailed for governments, in terms of the revenue they need to raise to finance spending, and the value of benefits going to those receiving support. This is because the taxes and social charges which are levied on benefits, and which vary significantly between Member States, reduce the amount of transfers involved. On the other hand, the figures understate the amounts involved, since they exclude the social transfers which are made through tax concessions or reduced contributions which have the same effect on government budgets and the net income of recipients as expenditure on benefits.

Both of these elements are difficult to estimate, which is a major reason for their exclusion up to now from the ESSPROS database. Nevertheless, estimates produced by OECD, which have to be regarded as preliminary, are at least indicative of the size of the amounts involved. These are relatively large for the Nordic countries and the Netherlands, where charges are levied on benefits to a greater extent than elsewhere. In net terms, therefore, the cost of social protection and the sums received by beneficiaries are reduced in these cases to a level more similar to that in other Member States. In Sweden, net expenditure on social protection is estimated to have amounted to around 30% of GDP in 1996, much the same as in Germany, and in Denmark and Finland, to 26-27% of GDP, much the same as in the UK and slightly above the level in the Netherlands. As a result, the gap between spending relative to GDP in these countries and that in Italy and Ireland is narrowed considerably.

These preliminary results confirm the importance for policy analysis of extending the ESSPROS database to include estimates of net expenditure, incorporating the amounts transferred by means of tax concession and reduced contributions, to give a complete view of the implications of social protection for national budgets and the redistribution. To this end, the Commission, through Eurostat, plans to set up a task force with Member States to develop a special module of the database containing such estimates.

Growth of social expenditure, 1990-1996

While spending on social protection has continued to grow in recent years in nearly all parts of the Union, the rate of increase in real terms has slowed down in recent years in most Member States since 1993. This coincides with the end of recession of the early 1990s, which pushed up expenditure on unemployment benefits in particular, but also on disability and housing benefits.

Whereas social expenditure increased markedly between 1990 and 1993 (from an average of 25½% of GDP to just over 28½%), it fell slightly relative to GDP the following years, partly because of the slowdown or decline in unemployment-related expenditure, but mainly because of a slowdown in real spending in most areas. Only in Portugal, Greece, Germany, Austria, Belgium and Luxembourg did spending continue to rise relative to GDP over the three years 1993 to 1996, in the first two, reflecting the ongoing development of the social protection system, in the second three, the failure of their economies to fully recover as in most of the rest of the Union. The decline in social expenditure relative to GDP in the rest of the Union was most marked in Sweden, Finland and Netherlands, the three countries with the highest level of spending in 1993.

The pattern of social expenditure

Spending on old-age and survivors' benefits accounts for by far the largest element of social protection, amounting to an average of 43% of the total spent in the Union in 1996, or to just over 12% of GDP, with health care the second largest, accounting for just over 21½% of the total, or just over 6% of GDP. Outlays on disability and unemployment benefits were of a similar size, each accounting for just over 8% of total outlays, though in 9 of the 15 Member States in 1996, transfers to people with disabilities exceeded those to the unemployed. Expenditure on the family and children, which covers maternity benefits and child-care support as well as child benefits, is only slightly smaller, on average, than outlays on disability and unemployment benefits, while spending on the other three kinds of benefits -sickness, housing and social exclusion - taken together is much the same size as this, averaging around 7½% of overall spending in total.

Despite the ageing of the population across the Union, the main change in the pattern of spending on social protection over the period 1990 to 1996 was not, as might be expected, a relative growth of old-age pensions or health care, spending on both of which increased at much the same rate as the total, but other elements of expenditure, in particular, disability and housing benefits and social exclusion. Some of this rise is attributable to the recession of the early 1990s and the increase in unemployment which accompanied it, but spending on the first two items has continued to rise relative to that on other elements since 1993. The main item on which expenditure has fallen since 1990 is sickness benefits, which seem to have been a focus for curbs on spending in a number of Member States.

Expenditure relative to beneficiaries

While no information on the number of beneficiaries is included in the ESSPROS database, some indication of the possible effect of changing numbers on expenditure can be gained by examining other evidence on this - specifically, data on the number of people over the official age of retirement and the number unemployed.

Over the period 1990 to 1996, expenditure on old-age benefits in the Union (including survivors' benefits) increased on average by 3½% a year in real value terms. The number of people above retirement age went up by 1½% a year, implying a rise in the average benefit paid per person of 2% a year, which is broadly in line with the long-term growth of GDP per head. Over this period, however, because of the depressed rate of GDP growth, it meant that average pensions rose by slightly more than GDP per head. Apart from the Netherlands and Greece, where there was a small fall, average pensions in these terms increased in all Member States, most especially in Denmark, the UK (both by 4% a year) and Portugal (8% a year).

Spending on unemployment compensation (excluding redundancy compensation and benefits in kind) went up by just under 4% a year in real terms in the Union in the period 1990 to 1996, less than the increase in the number of unemployed as defined according to ILO conventions, which rose by just over 5½% a year. Compensation per person unemployed, measured in these terms, therefore, fell by almost 2% a year, which may reflect a reduction in the number of unemployed covered rather than a decline in the average benefit paid as such, though it may indicate a tightening in the system of transfers, which would be in line with policy efforts in many Member States over this period. The fall was apparent in most countries and was particularly marked in Spain, Greece, France, Belgium, Germany and Sweden. On the other hand, there was a substantial increase in the implied value of compensation per person unemployed in Ireland, Portugal and the Netherlands.

3. Financing social expenditure

On average social contributions financed some 63% of the total expenditure for social protection in 1996, though the proportion ranged from two-thirds or more in France, Belgium, the Netherlands, Germany, Italy and Spain to just over half in Luxembourg, Sweden and Finland, around 40% or less in Portugal, the UK and Ireland and only 25% in Denmark.

Since methods of financing are rooted in the way that systems have developed historically and are linked closely with the way that entitlement to benefits is controlled, they tend to be slow to change. Despite the increasing concern to limit or reduce the charges falling on income from employment, the relative importance of social contributions over the Union as a whole, declined only marginally between 1990 and 1996 (from just under 64½% of the total to just over 63%).It should be emphasised that while social contributions represent the major part of total taxes on employed labour, there are other taxes accounting for more than one third of this total in EU.

The reduction in the relative importance of social contributions occurred in the first three years of the period, and after 1993 there was even a small increase in their relative importance. This was also the case as regards employers' contributions, their share of total funding falling from 42% in 1990 to just under 39½% in 1993, but remaining at this level over the subsequent three years while revenue from taxes increased from 30% of the total in 1990 to 33% in 1993 and then fell to 32% in 1996. Nevertheless, the share of revenue collected from social contributions changed significantly in a number of Member States over this period, though not uniformly downwards and more in countries where the share was relatively small initially rather than in those where it was large. In Portugal, it declined from 57% to 43% over the period and in Ireland and the UK, by some 4 percentage points. On the other hand, social contributions became a more important source of funding in Denmark (where their share rose from 13% to 25%) and the Netherlands (from 59% to 67½%), as well as, more modestly, in Finland, Belgium and Austria. Overall, therefore, there is little sign of any convergence in the proportion of revenue yielded by contributions over this period.

Since the share of wages and salaries in GDP has tended to fall over the 1990s, particularly since 1993 and the end of the recession, the revenue from contributions has risen significantly in relation to wages and salaries - and therefore labour costs - from an average of 33½% of labour costs in 1990 to 37% in 1996, while employers' contributions rose from 21½% to 23%.

Nevertheless, since 1996, there is evidence of increased efforts in a number of Member States, especially those where rates of contribution are high, to shift the source of financing of social protection from contributions to taxes, in some cases to 'earmarked' taxes levied specifically to fund certain benefits. This is particularly the case in France, where the CSG (Contribution Sociale Généralisée), introduced as a 'solidarity tax' in 1991 to replace part of the social contributions and extended several times since then, now is higher than income tax. At the same time, the CRDS (contribution au remboursement de la dette sociale) an additional charge levied at a rate of 0.5% on income, introduced in 1995 to contribute to reducing the deficit of the social protection system, has been extended until 2014, while the tax levied on tobacco for a similar reason has been increased.

As a consequence of this shift, social contributions have declined as a source of revenue since 1996 whereas earmarked taxes have increased, accounting in 1998 for more than 15% of total funding for social protection in France. Similarly, in Belgium, employers' contributions have recently been reduced, while in Germany, the contribution rate for pensions was reduced in 1999 and a proposed 'ecological tax reform' will provide extra finance for social spending. Moreover, in Italy, it is planned to shift the financing of family benefits, including maternity allowances, from contributions to general taxation.

4. Recent changes in policy

The changes in systems of social protection which have occurred in recent years have centred on the same key objectives emphasised in the Commission Communication published earlier this year and outlined above. In most cases, the aims have been: to increase the incentive for people not in employment to find jobs if they are capable of working and to support them in their efforts to so, while trying to ensure that those unable to work have a reasonable level of income; to reduce the numbers dependent on State benefits and to encourage and help them to become integrated into society; to seek to safeguard future pensions both by limiting the transfers which will be required and by increasing the finance which will be available to fund them; and to contain the cost of health care whilst improving the standard of service provided.

Through the shift in the emphasis of policy from income support alone to active measures, in line with the Employment Guidelines, social protection is coming to represent a more important factor in overall economic and employment policy in the Union, supporting rather than obstructing the action taken in other areas.

4.1. Key objective: to make work pay and to provide a secure income

There is a widespread recognition across the Union that systems of social protection need to ensure that people have a financial incentive to look for work, that they do not spend so long out of work that it harms their chances of finding a job and creates an attitude of dependency and that their employability is enhanced rather than damaged during their spell of unemployment. This has led to efforts to ensure that levels of benefits in relation to prospective earnings do not represent a financial disincentive to find a job and that people are helped to do this through appropriate active measures.

At the same time, benefits still need to provide an acceptable level of income while people look for work. The challenge facing Member States is to set benefit levels and the regulations governing entitlement to them so that these two objectives are achieved simultaneously. In practice, both access of those unemployed to benefits and the amount they get differ significantly between Member States, the main divide in the case of the former being between the North and South of the Union. This seems to be confirmed by a recent empirical study of replacement rates, based on what people actually receive when unemployed as compared with their net earnings when in work as reported in the first two waves of the European Community Household Panel (ECHP). The results, which relate to 1993 and 1994 are exploratory. They will need to be compared with evidence from other studies and to be further investigated as data from new waves become available for analysis. It should also be kept in mind that it is always difficult to gather high quality information on income and transfers through interviews.

The findings show that in Italy, Greece and Portugal, over half of those in the Union who were unemployed for at least three months over the two years did not receive any unemployment benefit at all and most of these no social assistance either, while in Spain around a quarter received no form of social transfer. In Denmark, Germany and Belgium, by contrast, only 10% or less of the unemployed received no unemployment benefit and hardly anyone no State support of any kind, the latter being the case also in France, Ireland and the UK, despite fewer receiving unemployment benefit as such.

The average level of unemployment benefit (including those receiving no benefit at all) varied from almost 80% of net earnings when the person was working in Ireland and 70% or above in Germany, Belgium and Denmark to under 25% in Italy and the UK and under 10% in Greece, with France (around 60%), Spain and Portugal (around 35%) in between. For the UK, in particular, however, these figures understate the amount received because they exclude housing benefits paid to households, which are important in the UK.

These averages give only a very partial indication of the extent of the financial incentive for people unemployed to look for work. A more relevant indicator is the proportion receiving relatively high benefits compared to earnings in employment, though there is no general agreement on what should be regarded as 'high'. In most Member States, for example, only a small proportion men and women who were unemployed had a replacement rate of 80% or above. Except in Denmark (27%) Germany (25%) and France (24%) for men, and Denmark (49%), Germany (21%), France (30%), Belgium (23%) and Ireland (21%) for women, less than 20% of unemployed had a rate as high as this in all Member States. This proportion was only around 10% in Spain and well under 10% in Italy, the UK and in Greece both for men and women. Except in Denmark, over 60% of men and women either had a replacement rate of under 65% or did not receive any benefit at all.

This evidence suggests that the financial disincentive to work was relatively weak for a majority of unemployed in most Member States in 1993-1994. This broad conclusion has to be taken with some caution. Further investigation should notably pay attention to specific groups, such as low paid people, who may face financial incentives to be unemployed or not actively look for work. Furthermore, since 1993-1994, the discussion on incentive issues has led several governments to tighten benefit entitlement conditions.

A prominent feature of policy across the Union in the past few years has been a re-examination of benefit entitlement conditions . As a result, measures have been taken in many Member States to tighten the qualifying conditions for eligibility for benefit and to increase the activation of the unemployed to find a job.. These measures have not been confined to unemployment benefits but have also been applied to disability benefits and early retirement schemes, which had come to be used as effective substitutes for unemployment compensation in a number of countries.

Changes in benefit systems have also been combined in several countries with attempts to make work more rewarding financially by introducing measures to increase the net income of those taking up a job, both through allowing them to continue to receive some support after they begin to work and by extending in-work benefits.

In work benefits

This approach has been pushed furthest in the UK, where Family Credit, aimed at supplementing the net income of people in low-paid jobs with children, was replaced in 1999 by Working Families' Tax Credit, guaranteeing all people in employment with dependent families, including lone parents, a higher minimum level of income than before and giving them an allowance towards the costs of child care. In addition, a similar measure has been introduced for people with disabilities and a pilot scheme of the same kind has been launched for people over 50 who are unemployed, with the aim of encouraging them to work.

The only other country in the Union in which in-work benefits are important is Ireland, where Family Income Supplement has existed for some time and has covered an increased number of people in recent years as qualifying income levels have been raised and the conditions of eligibility and receipt made more favourable. Moreover, recipients of unemployment benefit can now retain their dependent child allowances for 13 weeks after taking up a job to increase their net income and provide an extra incentive to start working.

Elsewhere, in-work benefits have also been introduced in a number of countries to assist people make the transition from unemployment into work but equally to encourage them to take part-time or temporary jobs, as in Austria, Germany and Portugal, where the main aim is to help the unemployed gain work experience so as to improve their future employability.

Changes in unemployment compensation schemes

While the changes made to unemployment benefits over the 1990s have concentrated on containing the costs to public budgets, in recent years a wide concern has also been to integrate income support with active labour market measures to a greater extent and to encourage those in receipt of benefit to find a job as soon as possible. In a number of Member States, including Denmark, Germany, the Netherlands and the UK, the definition of what constitutes a suitable job for a person unemployed to take up has been broadened, to encompass jobs which they are capable of doing with a minimum of training rather than merely those they were doing before. Refusal to accept a suitable job after being unemployed for a time or unwillingness to participate in active labour market programmes can lead to loss of benefit.

At the same time, changes have been introduced, or are planned, in a number of countries to improve the safety net provided by unemployment benefit schemes. In Italy, an insurance-based unemployment benefit scheme, with means-tested support after entitlement has expired, is being introduced to provide wider protection than up to now, while in Portugal, the period of entitlement to unemployment benefit has recently been lengthened for people in all age groups.

While the recent focus of policy attention across the Union has been on ensuring that work pays, there are still circumstances in some countries where people may have little financial incentive to be in paid employment. This applies particularly to people, usually women, whose spouse is unemployed and in receipt of means-tested unemployment benefit related to household income, such as in the UK and Belgium. Any additional earnings in these cases are liable to be offset by a loss of benefit. In both the UK and Belgium however, measures have been taken to lower the possible deterrent to working.

Changes in disability benefits

Disability benefits have been used as a more 'acceptable' substitute for unemployment benefits in a number of Member States, especially in respect of the long-term unemployed. As expenditure and the number of claimants has grown, however, there have been widespread moves to try to restrict eligibility to those who are genuinely incapable of working because of their disability. In the Netherlands and the UK, where, more people than are unemployed are being supported on benefit, reforms were introduced in the mid-1990s to make medical testing procedures more stringent and to widen the definition of the jobs that the person concerned can potentially do.

In addition, in the Netherlands, some of the onus of policing the system was shifted on to employers from 1998 on, by charging them differential contribution rates according to the risk of their employees becoming claimants, giving them an incentive to reduce the numbers either by dissuasion or by making it easier for people with disabilities to continue working. In the UK, further measures are planned to tighten the qualifying conditions even more, while, as noted above, those moving into employment are now guaranteed a minimum income through a tax credit, though at the same time, benefits to those who are really incapable of working have been raised.

Elsewhere, such as in Austria, Greece and Italy, eligibility for benefit has also been made more restricted, while in Sweden, where spending is comparatively high, the system is in the process of being reformed. In contrast, in Ireland, where spending on disability benefits is relatively low, social assistance disability allowance was extended in 1997 to those in part-time residential care and benefits were increased to couples who both have disabilities.

Early retirement pensions

Along with recipients of disability benefit, the number of people taking early retirement has also risen over the past 15-20 years, as noted above, for much the same reason. Indeed, the two are alternative responses to the problem of long-term unemployment, with the UK and the Netherlands, for example, opting to expand disability benefits, France, Germany and a number of other countries, to rely more on early retirement. The rising cost has led to requirements for eligibility being tightened in the case of both types of scheme and levels of benefit being reduced in the case of early retirement.

In Denmark, Austria and Italy, the years of contribution required to be eligible for early retirement have been increased, while in Germany, the minimum age for claiming a pension is in the process of being progressively raised (to 63, or 62 if long-term unemployed)..

In Luxembourg, however, the possibility of early retirement has been extended, though with the aim of encouraging people to continue working on a part-time basis rather than retiring completely. This is similar to schemes operating in France, Germany, Denmark and Austria, where employees over a certain age can opt to reduce their hours of work progressively until they retire and be compensated by receiving a partial pension. In addition, as in Austria and Germany, employers are entitled to a reduction in social contributions if they take on someone who has been unemployed at the same time.

Family benefits and other arrangements to reconcile work and family

Social protection for families with children has three main aims: to help people bear the additional costs of having children through child benefits or allowances; to support women when they give birth through maternity allowances and entitlement to leave, which being increasingly extended to both parents; and to provide help with child care and with caring for family members with disabilities or who are elderly and frail. All three are important for improving the income security of people with families, especially women, and helping them reconcile work with their other responsibilities. Child-care support arrangements and the right to leave to take care of children - as well as other members of the family who may need long-term care - are crucial for many women who wish to pursue working careers. This is especially the case for those bringing up children on their own, who are increasingly numerous. As recognised explicitly in the EU Employment Strategy, women still face special difficulties in gaining access to the job market, advancing their careers and reconciling professional and family life, and under the Employment Guidelines, Member States have committed themselves to ensuring that women have equal access to labour market measures and job opportunities.

In Germany, child benefits were raised significantly at the beginning of 1999 and are due to be raised again in 2000, along with child tax allowances, so increasing the net income of families. In Luxembourg, support are been targeted more on low-income families through reducing child tax allowances (which tend to favour higher income earners) and increasing child benefits by an equivalent amount. Similarly, in Italy, a new means-tested benefit was introduced in 1999 for all families with more than 3 children below 18.

The main developments in this area, however, have been aimed at increasing parental leave possibilities, encouraged by the recent Community Directive. The entitlement of people with children to take leave has been extended, or is in the process of being extended, in many parts of the Union, including Austria, Luxembourg, Denmark, Belgium, Ireland, the Netherlands, Italy, Spain and in the UK.

At the same time, the availability of child-care support has been increased in recent years in a number of Member States, most especially in ones where support has been limited in the past. Support facilities are often a necessary complementary component of active labour market measures to make it possible for many people, especially women, to pursue a working career and to become properly integrated into society. In Spain, tax deductions allowable for child-care costs for children under three were raised significantly in 1998. In Greece, a pilot project has been launched to extend the time children are able to spend in nursery and elementary schools as well as in State day-care centres and in Portugal, the network of nurseries has been expanded. In the Netherlands, from 1999, lone parents looking for a job or undergoing training are entitled to reimbursement of after-(primary) school child-care costs if their income falls below a certain level. In the UK, a new child-care tax credit was introduced in October 1999 for low-income families, including lone parents, while there is an objective to expand the number of out-of-school child-care places.

4.2 Key objective: to make pensions safe and pension systems sustainable

The significant growth in the number of people above retirement age in all Member States in 10-15 years time has focussed attention everywhere on the consequences for old-age pensions and their funding given that old-age and survivors' benefits already represent 43% of the total social protection expenditure in the Union, or to just over 12% of GDP. Though the measures introduced differ in detail, they have a common emphasis on limiting the future transfers which will be necessary or increasing the finance available. They include increasing the official age of retirement or the number of years of contributions required to qualify for a full pension, reducing the pension paid in relation to past earnings, or relating it more closely to contributions, and creating special funds to finance future transfers. The latter amounts to adding beside existing 'pay-as-you-go' systems, where present payments are financed from present contributions, 'funded' ones where present contributions are fixed in relation to future liabilities.

In Austria, Italy and Luxembourg, where rules for civil servants' pensions were more favourable, they were reformed to bring them closer to those in the private sector, while in Spain, the basis for calculating pension entitlement is in the process of being progressively extended from salary in the best 8 years to that in the best 15 years. In Italy, the highest pensions were frozen for 1998, while in Germany, the basis for indexing pensions has been changed from wage increases to price inflation for both the years 2000 and 2001, so effectively reducing expenditure. On the other hand, the 1999 reform, aimed at a gradually lowering pensions from 70% of the average net wage to 64% has been suspended until the end of 2000.

In order to increase available finance, the contributions paid by higher and middle income earners in Spain have been increased without compensating rises in future pension entitlement, while in Italy, contributions paid by the temporary self-employed have been raised. In Denmark, contributions were effectively increased from 1998 by incorporating an extraordinary charge of 1% levied in 1997 into the system for funding supplementary pensions.

In the Netherlands, where a ceiling has been imposed on the rate of contributions, the Government has set up a special fund, into which it contributes an amount each year calculated to meet the peak in expenditure in the years after 2020. In France, there is a proposal to increase the contributions required for a full pension from 40 to 42.5 years, and, as in Spain, a fund has been created to consolidate the finance available for pensions. In Ireland, a Social Welfare Pension Fund Reserve was set up in 1998, using money from tax revenue and privatisation of telecommunications, to fund the future cost of pensions. The high rate of economic growth, however, has enabled the full rate of social contribution to be reduced and the basic pension to be increased significantly in both 1998 and 1999.

At the same time, measures have been taken in a number of countries, such as Italy, to encourage the development of occupational and personal pensions to relieve the future burden on the State.

Despite the focus in most countries on making pension systems sustainable, improvements have, nevertheless, been made in a number of Member States to the entitlement of disadvantaged groups. In Italy and Spain, minimum or non contributory pensions were raised. Credit for periods spent looking after children or adults in need of care has been extended in Luxembourg, Germany, Austria and the UK, while people working part-time have been accorded increased pension rights in Luxembourg, Germany and Spain and those with shorter working careers than the norm in Finland.

4.3 Key objective: to promote social inclusion

A major feature of social policy in recent years, as noted above, has been the emphasis on active policy measures designed to get people into work. This has also been motivated by the aim of reducing dependency and social exclusion. Although the main focus is on employment, since it is recognised that having a stable job, or at least having the option to work, is a major condition for social inclusion, it is also recognised that there other aspects of exclusion which need equally to be addressed if the problem is to be tackled effectively. These include better access to education and vocational training, to decent housing and to reasonable levels of health care and social services.

In all Member States, governments have committed themselves to preventing exclusion and increasing the effort devoted to active measures, aimed at providing job search assistance, improving employability and, where necessary, direct access to jobs. These measures have been directed primarily at the unemployed, especially the long-term unemployed and young people, but also other disadvantaged groups, such as women returning to work after interrupting their working careers because of caring responsibilities. They have also been aimed at other people out of work, particularly people with disabilities but capable of working, older people who have taken early retirement and lone parents with children.

In the case of young people and the long-term unemployed, following the adoption of the Employment Guidelines for 1998 and 1999, all Member States have introduced measures to ensure that everyone out of work for more than a certain time (6 months in the case of those under 25 and 12 months for those of 25 and over, though 3 months and 6 months in Sweden, Portugal and Luxembourg) has access to an active measure of some kind or a subsidised job. It has been accompanied by efforts to tailor measures more closely and systematically to individual needs in order to increase their effectiveness. In many countries (such as Belgium, France, Ireland, Finland and the UK), the same kind of approach has been applied to people with disabilities, and conscious efforts have been made to integrate them into active labour market programmes. In the Netherlands and the UK, the same kind of assistance has recently been introduced for lone parents, particularly those with children over 5.

In several countries including Belgium, Denmark, Germany, France and the Netherlands, measures regarding job search, career guidance and the provision of training, or retraining, have been supplemented by the significant growth of subsidised jobs or direct job creation schemes. These have been directed, in particular, at the long-term unemployed and others with difficulties finding work.

The activation of measures has extended to those on minimum income schemes, who, in a number of countries, are being encouraged to participate in programmes designed to improve their employability. This is the case in Belgium, for example, where training is now provided for those on minimum income support and in Denmark, where from 1998, receipt of income support has been made dependent on people participating in schemes aimed at increasing their self-reliance. Moreover in Italy, recipients of the prospective means-tested benefit guaranteeing a minimum level of income who are capable of working will be obliged to participate in a training course.

Efforts to ensure that people are better off when working than when not have been extended from unemployment insurance benefit schemes to minimum income schemes in several Member States. In France, for example, since 1998 the possibility, which already existed, for some means-tested benefits (allocation de solidarité spécifique, revenu minimum d'insertion) to top up wages has been increased. The periods over which this is possible have been extended and the levels of income permitted raised. This possibility has also been extended to specific allowances for lone parents and widows (allocation de parent isolé, allocation d'assurance veuvage). Similar steps have also been taken in other countries, such as Belgium, Germany, Ireland, the Netherlands and Denmark, while in Italy, a pilot scheme for a new means-tested benefit to guarantee for the first time a national minimum level of income incorporates an earnings disregard of 25% so as to maintain a financial incentive to work.

In practice, despite the emphasis on active measures over the 1990s, there has been relatively little shift in actual spending from passive to active measures. According to OECD data on labour market expenditure, the relative increase in spending on active measures in the Union since 1994 and the end of the economic recession has been small and these accounted for the same share of total expenditure in 1998 (just under 37%) as in 1990. Moreover, the amount spent on active programmes per person unemployed was the same in 1998 in relation to GDP as in 1994 and less than in 1990. This indicates that there is still scope to extend opportunities of benefiting from active measures, especially towards the most disadvantaged groups.

4.4. Key objective: to ensure high quality and sustainable health care

The pressure on systems of social protection across the Union has also manifested itself in a growing need for health care which in 1996 accounted for just over 21½% of social protection expenditure or 6% of GDP. This pressure is reinforced by the effect of rising real income and advances in know-how and in the range of treatment in pushing up demand, so putting substantial upward pressure on expenditure. The common response of governments has been to seek to limit the growth of spending and its effect on public budgets, whilst seeking to maintain the quality of service and access to care. This has naturally led to a focus on ways of increasing productivity or improving cost effectiveness.

To this end, governments across the Union have tightened control over overall spending, either directly where there are national health systems or indirectly by limiting contribution rates in the case of insurance-based systems. At the same time, in many countries with national health services (Spain, Italy, Sweden, the UK and Ireland), more responsibility for managing services and determining the allocation of budgets has been devolved to the regional or local level, where needs can be more closely assessed. In addition, there has been a widespread tendency to extend co-payment for drugs and treatment, increasing the cost borne directly by the consumer to encourage them to economise on their use of the service as well as to provide additional finance. This has gone along with the establishment of approved lists of drugs and the introduction of generic medications to limit the cost of prescriptions.

In some Member States, however, there has been a reaction to continuous increases in charges left to patients. In Sweden, in particular, where these were raised markedly during the 1990s, there has been growing concern that low income households, especially those with children, might forego treatment to the detriment of their health. Charges for children were, therefore, abolished in 1998, leading to counter-concerns about possible excessive demands on the service and cutbacks in other public services to compensate for the reduced funding. The response has been to transfer more ear-marked revenue to the local authorities responsible for the provision of services and to propose shifting resources from defence to health care. In France, because of concerns about access to health care, basic health-care insurance is to be extended to everyone and complementary health insurance is to be provided free to low income households, or to some 6 million individuals. In Luxembourg and Belgium, measures have also been taken to ease access to health care and to lower charges for poorer households, while in Spain, all foreigners, legal and illegal, are to be granted access to health care on equal footing to nationals.

In the Netherlands, a 20% charge on all treatment and drugs, except visits to hospitals and GPs, levied in order to encourage people to use services more prudently, was withdrawn at the beginning of 1999 because it seems to have had little effect on dampening demand while involving heavy administrative costs. In Germany, where expenditure on health care is among the highest in the Union, a number of measures introduced to increase the cost of treatment falling on patients were withdrawn or suspended at the beginning of 1999, but ceilings on doctors' fee were maintained. Proposals under discussion, designed to contain cost increases and keep down contribution rates, include limiting patients' freedom of choice over provision, restricting the drugs that can be prescribed and imposing ceilings on expenditure of general practitioners, leading to claims that doctors might refuse to treat elderly or chronically ill patients because of the cost.

The share of private expenditure in total expenditure on health care in the Union has increased slightly over the 1990s, according to OECD data, from 22% in 1990 to almost 24% in 1997, so adding just over 30% to the outlays included as part of social protection. The importance of private spending on health care varies markedly between Member States, from 40% of total spending in Portugal and 30% in Italy to around 15% in Sweden, the UK, Belgium and Denmark and only 8% in Luxembourg. In countries such as Italy, Finland and Sweden, constraints on public health expenditure seem to have led to a growth of private provision between 1990 and 1997 - directly as well as indirectly, since the increased use of co-financing was one of the ways of achieving a reduction in public spending.

5. Data issues

Wherever possible the analysis in the report is based on quantitative information about the scale of spending on social protection and on particular types of benefit, about the sources of finance and recent changes. This is important in order to present an objective view of developments and the factors underlying these and so help to improve understanding of the changes occurring across the Union and of their implications, not only for economic policy and government budgets, but also for social issues, such as the relief of poverty and social deprivation.

Nevertheless, the quantitative analysis which it is possible to undertake in this area is limited by the availability of data. The new ESSPROS database of social protection expenditure and receipts, which Eurostat has compiled in recent years on a more coherent and satisfactory basis of classification and the recently established European Community Household Panel (ECHP) surveys provide an invaluable insight into the features of social protection systems across the Union and into the developments taking place. There are, however, still important gaps in coverage.

In particular, apart from problems of comparability in certain areas, the ESSPROS data do not as yet include details of the social transfers delivered through tax concessions rather than benefits or of the taxes and charges levied on benefit payments which reduce both the net income paid to recipients and the cost to governments of funding the transfers entailed. Nor do they include details of the number of people receiving benefits and, equally importantly, of those who do not, which makes it difficult to say anything about the effectiveness of systems in providing support to those in need or about the average amounts they receive. The lack of data on beneficiaries is in the process of being rectified for the unemployment function at least, by the development of a labour market policy module of ESSPROS.

The ECHP represents an alternative means of examining these kinds of issue since it provides information about income and the various sources of this, as well as about the personal circumstances of the individuals covered, including their employment status, and details of the households in which they live. The main problem with the Panel, however, is that it is a small sample, comprising some 60,000 households across the Union as a whole and less than 200,000 people, which may bring risks of representativeness where small groups of people are concerned. This restricts the analysis which can be undertaken, especially in areas of social protection which apply to only a small minority of the population.

Moreover, there are serious delays in data becoming available, which limits the usefulness of both sources as tools for monitoring changes. The data available being analysed relate, at a minimum, to the situation around 3 years previously, time enough for significant changes to have occurred. While both are relatively new sources of data, which explains for some of the delay, it is important that they become available on a more timely basis if they are to fulfil their full potential as tools for policy analysis.

6. Social protection in the candidate Countries of Central and Eastern Europe

There is widespread interest in the way that social protection systems are developing in the candidate countries of Central and Eastern Europe (CEECs) that have applied to join the European Union, which is likely to increase as their accession draws closer [2]. The significance of the current enlargement - the largest yet in terms of people - is one of the factors affecting the future of social protection policy in the EU identified in the recent Commission Communication A Concerted Strategy for Modernising Social Protection.

[2] Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia.

A strategy of the kind the Communication proposes and the challenges to which it is a response are not the sole province of the present 15 Member States. New countries joining the EU must embrace the full acquis or body of agreements, commitments and policies that Member States have in common and which they are developing together. In relation to social protection, this means both the binding Directives and Regulations in force at the time and the common principles or values to which Member States subscribe in ensuring that their systems are geared to tackle emerging problems and to take advantage of new opportunities. Underlying this is the notion that well-designed and responsive systems of social protection have a fundamental role to play in supporting economic policy and improving economic performance. Further shared ideas and ideals may well emerge during the course of Member State deliberations on the agenda proposed in the Communication, though it is important in all this not to forget that, while Member States may share a common vision of the purpose of social protection and the broad way it should evolve, they have total freedom to decide the specific changes they make and the new measures they introduce.

The four objectives set out in the Communication also offer a way of articulating what social protection systems in candidate countries must achieve in way of securing the economic, social and political stability necessary to ensure the fullest gains from participation in the internal market and EMU. Enabling men and women to make a secure living through working, mitigating the effects on them when they cannot, guaranteeing a safe and decent income during retirement and ensuring access for all to quality heath care and the essentials of life can only strengthen political stability and support for enlargement. The benefits they bring to individual lives will also translate into benefits for the economies of the countries concerned. Ensuring sustainable funding for social protection, not least for health care and pensions, will be essential under the economic discipline associated with EMU. It is also essential that social protection systems are efficient and effective, which means putting in place appropriate administrative structures. This is a key aim in the EU's approach to supporting candidate countries in their preparations, helping them strengthen the administration needed for implementation, through, for example, training and sharing experience.