Accept Refuse

EUR-Lex Access to European Union law

This document is an excerpt from the EUR-Lex website

Document 52013SC0038

COMMISSION STAFF WORKING DOCUMENT Evidence on Demographic and Social Trends Social Policies' Contribution to Inclusion, Employment and the Economy Accompanying the document COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS Towards Social Investment for Growth and Cohesion – including implementing the European Social Fund 2014-2020

/* SWD/2013/038 final */

In force


COMMISSION STAFF WORKING DOCUMENT Evidence on Demographic and Social Trends Social Policies' Contribution to Inclusion, Employment and the Economy Accompanying the document COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS Towards Social Investment for Growth and Cohesion – including implementing the European Social Fund 2014-2020 /* SWD/2013/038 final */


This Staff Working Document (SWD) provides the Social Investment Package's underpinning evidence on the long-term trends and short-term social challenges that Europe is facing.

The document demonstrates that large differences exist between Member States, but that they face similar long-term challenges. Demographic changes, an ageing population and decreased fertility are leading to a smaller working-age population and increasing dependency ratios. This threatens the solidity of financing for social protection budgets, which are generally financed through labour taxation. At the same time, the economic, social and financial crisis has aggravated this situation. Further, it has increased (long-term) unemployment, social exclusion and poverty across Europe.

This SWD highlights the three functions of social policy expenditure — social investment, social protection and economic stabilisation — and shows how they complement and reinforce each other. It analyses the size and structure of these budgets, their efficiency and their adequacy. In doing so, it makes the case that the size, structure and design of social policies all matter for the performance of welfare systems. Expenditure on social policies that focus on social investment is shown to be linked with greater returns to employment. To help illustrate this point, the paper also provides detailed analysis of social policy budgets and their financing before and after the crisis.

The SWD focuses further on the risks people face at different stages of their life courses, and points out that a difficulty experienced at one stage of a person's life can often interact with other difficulties later, and may exacerbate them. Successful policy interventions mitigate disadvantage and prevent it from becoming compounded. Often it is more efficient and effective to prevent than to cure, better to anticipate and prepare, than it is to repair. Evidence in this document shows that certain policies with a focus on prevention and on developing human capital can result in considerable savings later on. This knowledge specifically builds a case for early child education and care (ECEC) and active inclusion measures.

Lastly, the paper notes the main obstacles to obtaining timely and comprehensive social statistics. These jeopardise the way Member States can design their policy interventions, how those efforts are monitored in the European Semester and how EU funding is directed. In response to the issue, it sets out the way forward to improve the availability and timeliness of good social data.

1..... Introduction.. 2

2..... The structure of social protection budgets. 3

3..... Long-term trends. 7

3.1     An ageing population. 7

3.2     A trend towards a more active society. 9

3.3     Geographic trends and growing regional imbalances. 10

3.4     More diverse populations, a more diverse workforce, more diverse lives. 11

3.5     Changing family structures. 13

3.6     The development of information and communications technologies. 13

4..... The social impact of the economic and financial crisis: divergence between Member States and growing inequality in the European Union.. 15

4.1.    Growing divergence between EU countries. 16

4.2.    Trends towards greater social inequality in most countries. 18

5..... The evolution of poverty and social exclusion.. 22

5.1.    The multiple facets of poverty and social exclusion. 22

5.2.    The gender dimension of poverty and exclusion. 25

5.3.    The depth and duration of poverty. 26

5.4     The most severe forms of poverty and social exclusion: focus on homelessness, the Roma and migrants  29

5.5     The economic returns on social investment and the costs of unemployment, poverty and social exclusion  30

6..... Risks over the life course and opportunities for intervention.. 33

6.1     Children  34

6.2     Youth    38

6.3     Working-age populations. 40

6.4     Older people. 45

1. Introduction

European societies have been changing and will continue to change profoundly and rapidly in a number of interlinked areas with major implications for social outcomes and thus for social policy. These areas include demography, family structure, employment, poverty and social inclusion. At the same time the social policy responses and programmes have themselves been undergoing reform in response to economic and social change. The interaction of all of these defines the context for future social policy-making. Social developments in the European Union are also increasingly influenced by the global economy. Many countries outside the EU are experiencing similar challenges, and are engaging in efforts to extend social protection, particularly in emerging economies.

Demographic changes, an ageing population and decreased fertility mean fewer people of working age and increasing dependency ratios. Though a longer life and better health are of course intrinsically desirable, a smaller population of working age relative to the elderly will not only impact the design of Member States' health systems, but will also threaten the solidity of financing for social protection budgets, which are generally financed through labour taxation. Meanwhile, changes in family sizes and structures have led to weaker support systems, as social protection models have failed to adapt. The workforce will become increasingly diverse as the population of working age shrinks. Member States are introducing measures to enable more women, older people and non-EU nationals to join the labour market. In a very positive development, workers are increasingly mobile, but this sometimes creates problems as children and older people are left without family support. 

In the decade before the crisis, economic and employment growth in general improved overall living standards. Many governments were able to devote more resources to social policy intervention. However inequalities often grew, so poverty and social exclusion remained a major issue in most EU countries. This included people in employment, though the pattern was very uneven across Europe. The economic crisis has exacerbated economic and social pressures. Recent data[1] point to higher levels and deeper forms of poverty and social exclusion. The proportion of people at risk of poverty or social exclusion has risen in a number of Member States since 2008.

The crisis has also brought about increasing divergence between Member States and between regions of the EU. Unemployment and poverty are major issues in most EU countries, but there are substantial differences among them. The risks of falling into and the chances of getting out of poverty also vary across Member States, and significant differences can also be observed in terms of efficiency of spending when it comes to poverty reduction.[2] The structure of employment across the EU also continues to become more polarised between high-income and low-income jobs, resulting in increases in inequality and thus in relative poverty. Not only do these trends run counter to European values of fairness and dignity for all, they also pose a threat to our economy, as poverty and social exclusion bring with them significant social and economic costs.

Some Member States have sought to confront the long- and short-term challenges with reforms to their social models, changing both the sizes and the structures of their social protection systems. Those that have been most successful in meeting these challenges are those that have maintained adequate social protection systems, reformed their labour markets and adopted policies strongly emphasising social investment, which seeks to empower individuals by facilitating the development of their human capital. These social investment policies reinforce social policies that protect and stabilise by addressing some of the causes of disadvantage and giving people tools with which to improve their social situations. Education and training in particular play a key role in breaking the intergenerational transmission of poverty and improving people's outcomes.[3] Measures to reform labour markets to make them more inclusive can also improve individuals' opportunities to make best use of their skills.

This paper looks at current social trends and challenges that the EU is facing in greater detail, and outlines how these have affected poverty and social exclusion. It puts a strong focus on the specific risks people experience at different stages of their life courses, and shows that a difficulty experienced at one stage of a person's life can often interact with other difficulties later, and may exacerbate them. Social policy interventions, especially those with a social investment focus, may help to mitigate disadvantage and prevent it from becoming compounded. Such interventions take various forms, such as early child education and care (ECEC), up-skilling opportunities, and inclusive labour market measures.

To help illustrate this point, the paper also provides detailed analysis of social policy budgets and their financing before and after the crisis. It discusses the three functions of social policy expenditure — social investment, social protection and economic stabilisation. It shows how these functions complement and reinforce each other. It then analyses the size and structure of these budgets, their efficiency and their adequacy. In doing so, it highlights that it is not only different levels of social policy expenditure, but also different types of expenditure, that affect social outcomes — among other factors. Expenditure on social policies that focus on social investment is shown to be linked with greater returns to employment. It is also shown that social policy interventions to tackle disadvantage throughout the course of life appear to yield the highest returns if started as early in life as possible.

Finally, the paper notes the main obstacles to obtaining timely and comprehensive social statistics, which may impede the ability to take stock of and compare the living conditions of people across the EU, identifying those most excluded from society and monitoring the full impact of social policies. Improved monitoring will help Member States better target their policy interventions, which will help them improve delivery on the Europe 2020 employment, education and poverty targets. It will also better orient the direction of EU funding.

2. The structure of social protection budgets

Following Musgrave's classical framework,[4] which defines the three main functions of public intervention in the economy as stabilisation (aimed at securing economic stabilisation, in particular of GDP but also of employment and price levels), distribution (aimed at securing adjustments in the distribution of income and wealth, not least an equitable distribution of incomes) and resource allocation (aimed at securing adjustments in the allocation of resources and in particular the efficient use of resources), social policies can be considered according to the three functions of social investment (primarily linked to the allocation function), social protection (primarily linked to the distribution function, considered in a very broad approach covering in particular distribution of incomes over the life course) and stabilisation of the economy.

Investment function. A key function of social budgets is to strengthen people’s skills and capacities, in order to prepare them for confronting or preventing risks over the life course and improving their future prospects. In other words, social policies show their effects not only immediately at the time they are implemented, but also have lasting impacts: with the characteristics of an investment by offering some returns over time, for instance in terms of increased employment or labour incomes, thus enhancing growth. In particular, social policies ‘prepare’ individuals, families and societies to adapt to various risks and transformations (such as changing career patterns, new working conditions or an ageing population, cf. Vandenbroucke et al., 2011)[5] and can thus reduce the need for responses intended to ‘repair’ adverse situations.

For example, good quality childcare and early childhood education have been proven to have a strong impact on improving children’s chances of finishing their studies and finding employment, or of avoiding extreme risks such as delinquency and drug abuse. Preventive health care and health and safety at work help people to avoid sickness and maintain their productivity. Retraining and lifelong education help people to maintain employability and to obtain better jobs over their working careers. Promoting a healthy lifestyle, rehabilitation and improving access to assistive devices enable older people to lead independent lives as they become frail and develop functional limitations.

Box — A growing focus on a social investment approach

The social investment approach stresses the case for considering certain parts of employment and social policies — and possibly other policy areas, such as education — as entailing investments improving prospects for future employment and social participation, together with more social cohesion and stability (Van Kersbergen and Hemerijck, 2012), thus stressing the life course dimension of social policies and their long-term benefits for society.

The focus on social policies as an investment in developing human capital has been gaining ground in Europe since the early 1990s, in particular with the work of Esping Andersen (1992).[6] The last two decades have recently been analysed as a period of emergence of a ‘Social Investment State’ in Europe (Van Kersbergen and Hemerijck 2012),[7] while the roots of the social investment approach have been located in work on social protection as a productive factor developed in Sweden by Alva and Gunnar Myrdal in the 1930s (Morel, Palier and Palme 2012).[8] It has recently been argued that deficits in social investment in some Member States can contribute to explaining the current disequilibria observed in EMU (Hemerijck and Vandenbroucke, 2012).

Protection function. Social budgets also have the function of supporting and protecting people when they have to deal with the reality of some social risk. For instance, social insurance pools risks and redistributes contribution revenues in benefits to those affected by those risks.

Typically, unemployment benefits provide people with an income as they look for jobs, and social assistance does the same if they are unemployed for a prolonged period, while it also supports those who for one reason or another cannot go out to work. Social protection provides insurance against social risks over the course of life (e.g. health and long-term care, sickness benefits, disability and survivors' pensions) and also redistributes income from those actively in work to those who have retired (e.g., through pensions).

Stabilisation function. In periods of economic downturn, the need for social intervention rises rapidly, while receipts also automatically decline, which has a substantial countercyclical macro-economic effect, dampening fluctuations in GDP and in household incomes.

Typically, unemployment benefits (and to a lesser extent other benefits such as social assistance) automatically rise in scale in response to a downturn, while retraining schemes, which are part of active labour market policies, also increase in volume. This cushions the impact of an economic downturn or crisis on those directly affected (e.g., by losing their jobs or working fewer hours). Pensions also stabilise incomes towards the end of life and offer a second line of automatic stabilisers, as they remain stable while active incomes decline.

In the current economic context, public budgets are under pressure due to fiscal consolidation concerns. While some Member States have more fiscal space, most Member States have little room to spend more on social protection.  It is therefore essential to ensure the best use of existing resources and to avoid potential lasting adverse effects of solely savings driven adjustments on the short term and longer term employment and growth prospects. In this respect, well designed social systems combining the social investment dimension with the other two functions of protection and stabilisation, can contribute to increase the effectiveness and efficiency of social policies.

In practice, social policies actually very often support two or all three of these functions, which are mutually reinforcing. It would thus be misleading to allocate individual items of expenditure to a specific function (investment, protection or stabilisation), although some areas can be more directly linked to one of the three social functions. Typically, protection of human capital during adverse periods enables former investments made in human capital to be preserved.

For instance, unemployment benefits can function as protection (by providing the actual benefits) but also as investment (by preparing future labour market prospects through effective targeted activation measures and preservation of human capital during unemployment spells) and can also play a key role in automatic stabilisation. Furthermore, active inclusion policies refer to a protective pillar (adequate minimum incomes) and an investment pillar (access to quality services), alongside a labour market pillar (inclusive labour market measures). Pensions have a key protection function and also help stabilise household incomes and internal demand in recessions, while also having a social investment aspect, in that they help older people maintain themselves in a state of independence. Likewise, the provision of healthcare, while having a key protection function, also fulfils a crucial investment function that aims at the early detection and prevention of diseases, such as cancer screening,

While there is thus no exclusive correspondence between the various social policies (or risks[9]) and the three functions of investment, protection and stabilisation, specific social policies can however be more specifically linked to one of these functions, depending in particular on their design, on the specificities of national contexts and circumstances in time. For instance, preserving human capital, the family’s relation to the economy and the link to employment can generally be more directly linked to the social investment function (Vandenbroucke et al. 2011), relating more particularly the social investment dimension to some policies such as childcare, active labour market policies, rehabilitation, education or training.

Nevertheless, the expected returns from specific policies also depend on the national contexts and sequencing. For instance, the impact of childcare provision on the employment rates of women can depend on labour market factors such as the institutional labour market context (e.g., availability of part-time contracts and flexible working time arrangements) and in particular gender equality. Likewise, reforms to rebalance social spending, for instance, between old-age and childcare and unemployment benefits and to revise the design of unemployment benefits in specific countries (e.g., through greater activation measures, reducing work disincentives by aligning tax/benefit policies, etc.)  are also expected to achieve efficiency gains in terms of both economic and social effects.

For instance, in some southern European countries (ES, IT, EL, PT) the stabilising impact of social expenditures is relatively weak. These countries also acknowledge a low poverty reduction impact (excluding pensions) of social expenditures and low levels of childcare provision. Given the rigid constraints on public finances, a careful rebalancing of expenditure, most often rather than an increase, would help to reinforce both the poverty reduction impact of social benefits, as well as the investment in childcare and the stabilising effect of expenditures. This could include reviewing the structure of spending, currently skewed towards old age benefits (in comparison to other countries), making sure that unemployment insurance covers the most vulnerable workers (those who are first to lose their jobs), or adjusting the design of schemes to make them more responsive to the economic cycle (e.g., automatic adjustment of the duration of unemployment benefits).

A revision of the mix between cash and in-kind benefits can also lead to efficiency gains. Both cash and in-kind benefits have pros and cons and an optimal choice depends on the type of benefit and national circumstances. In-kind family benefits such as childcare are more employment-friendly than cash family benefits and ease the participation of women into the labour market, generating virtuous effects on female employment. Childcare services also contribute to preventing child poverty and the intergenerational transmission of disadvantage by securing parental incomes and giving access to quality education and care to all children. Cash housing benefits are more efficient in cushioning situations of temporary financial distress, while in-kind housing benefits can hamper workers’ mobility and create ghetto effects.

Promoting an ‘active’ welfare state should also contribute to the well-functioning cyclical behaviour of social spending, whereby expenditures increase more in response to a shock, and then decrease in times of recovery. This is an essential aspect of the sustainability of social systems. A recent study shows that countries that invested heavily in active labour market policies (ALMPs) before the crisis saw their employment levels less severely impacted during the crisis.[10] During the years 2000-2010, a number of countries (the Nordic countries, Germany and the Netherlands, for example) engaged in successful reforms to modernise their welfare states, aimed at reabsorbing high levels of long-term unemployment and/or swelling numbers of people on long-term illness or disability benefits resulting from the recession of the 1990s.

Improving the efficiency and effectiveness of education systems (accompanied by demand-side policies such as enhanced research and development (R&D) spending and innovation) is a key priority for increasing human capital and boosting productivity. Some proposed interventions do not entail higher spending, while at the same time increase future returns from human capital investments. Depending on the country, such interventions may include: curricula reforms in primary and secondary schools to improve basic and traversal schools; promotion of vocational, scientific and technological education with a greater responsiveness to the labour market and attention to innovative sectors; modernising assessment processes and wider use of standardised exams; and avoiding early tracking of students[11] [12].

More cost-effective provision and use of health services can also be done through better health promotion and disease prevention in and outside the health sector; creating financial incentives to encourage patients to register with a general practitioner (GP) or family doctor and using a referral system; use of less expensive equivalent (generic) drugs; ensuring a balanced mix of staff skills and anticipating staff needs due to ageing; promoting the cost-effective use of medicines; reducing the unnecessary use of specialist and hospital care while improving primary healthcare services; improving data collection and using available information to underpin the improvement of the performance of health systems; and using health technology assessment more systematically for decision-making processes.[13]

3. Long-term trends 3.1       An ageing population

People in the EU are fortunate to have a life expectancy among the highest in the world. This is a great achievement. This century-long trend seems set to continue, with life expectancy increasing by about a year every five years.

Fertility dropped from 2-3 children per woman in the 1960s to well below the population replacement level (2.1 births per woman) in the 1990s. Then, in the past decade or so, fertility started rising again in most Member States. There are signs that it could rise further, to just below the replacement level.[14] Moreover, low fertility rates in past years have resulted in smaller cohorts of women (and men), so even somewhat higher rates would result in lower numbers of births as compared to some decades ago. That would mean that rapid ageing and rising dependency[15] will occur, and in the long run the population will shrink — bar a large influx of migrants. As evidenced in the 2012 Ageing Report, Labour supply is projected to contract by 11.7% between 2020 and 2060, equivalent to 27.7 million people (24 million compared with the 2010 level).[16]

This will create challenges for economic growth and welfare in two respects:

· First, with more dependents and fewer workers, financing social policies will become more difficult.

· Secondly, with fewer workers, some economic activities will have to be rationalised; unless work-efficient alternatives are found, staffing social services will become more difficult.

Figure 3.1 — Population by sex and age group and working status, EU-27, 2010 and 2030[17]

2010 || age || 2030

|| 90 80 70 60 50 40 30 20 10 0 ||

Economic dependency ratio: 1.29 || || Economic dependency ratio: 1.44

Source: Eurostat, LFS and projections, computations

However, a significant part of the longer lives are not lived in good health, as measured by the healthy life years (HLY) indicator (see Figure below). There is an average difference of 14.3 years between life expectancy and healthy life years for men, and 19.5 years for women. The spread of values for HLY at birth among EU member states are much greater than for life expectancy, being 19 years for women and 18 years for men[18].

Figure 3.2 — Life expectancy (LE) and healthy life years (HLY) at birth by gender, 2008-2010 Average

3.2       A trend towards a more active society

Rising average ages are accompanied by rising life expectancies. From a societal viewpoint, this is important, as older people can remain active and independent for longer.

In the past decade, national pension systems have been reformed[19] and more older people have stayed involved in work and society. These were major achievements of what we call active ageing policies. At the same time, more women have entered the labour market while fertility has increased, albeit slowly, showing that policies to reconcile work and family life can work. These gains, together with increased immigration, have contributed to employment and economic growth[20] and partly offset the challenge of ageing. Figure 2.2 below illustrates these trends.

Figure 3.3 — Employment rates by age group and gender, EU-27, 2000-2011

Source: Eurostat, EU-LFS

3.3       Geographic trends and growing regional imbalances

With 10 new Member States joining the EU in 2004 and two more in 2007, more people are on the move. Growing numbers of people from the new EU-12 Member States choose to live in a different country in the EU-27. Their ranks increased almost ten-fold from 2000 to 2010, to 4.5 million.[21] Young adults from the EU-12 brought extra manpower to the EU-15, relieving saturated labour markets and gaining valuable experience. Over the same period, there was a slight rise in the number of EU-15 nationals living abroad permanently, though there are signs that more and more of them move abroad for short periods.[22]

A recent study indicates that people moving from the EU-2 (Bulgaria and Romania) during 2004-2009 may have increased aggregate EU-27 GDP by about 0.2% in the short term and 0.3% in the long term.[23] The study finds that for the receiving EU-15 countries, the long-term impact is even stronger (0.4%), exceeding the economic loss in the sending countries.

The downside of mobility is the cost for sending countries. For Bulgaria, Latvia, Lithuania and Romania, this has meant losing about 5 %[24] of their population in the past decade, a trend expected to continue for the next 20 years.[25] The losses will be heaviest among young adults (a one-third loss) and highly-skilled workers. Though these populations are currently among the youngest in the EU, especially Romania, by 2050-2060 they will all rank among the oldest. This is also a result of a long process of de-industrialisation following the collapse of the Soviet Union, and it will further exacerbate existing inequalities among Europe's regions. Closing the large decentralised industrial plants, mostly in rural regions, that were artificially supported in the times of planned economies has created an exodus towards the cities, especially among young adults in search of employment. Those who leave face difficult conditions at the margins in cities, while those who stay behind experience an aged and declining population and a weaker economy.

People in rural regions are likely to suffer the most from population decline and ageing. These are caught in a vicious cycle where population loss and ageing discourages investment; as a result, talented young adults are even more motivated to move to more developed regions. Many emigrating young adults care for dependent people in need of assistance, such as children or elderly relatives. When these dependent people also move, they often face integration problems (such as learning a language in new school); when they are left behind, they often lack the necessary care.

But the impact is not limited to the rural areas. In northern and western Europe, the employment rate in urban areas is below the national average. This leads to an 'urban paradox'. People arrive to job-rich city centres for work, but the resident population suffers from a lack of jobs. High unemployment among residents further puts pressure on social cohesion in cities and is a contributing factor in social problems such as inadequate housing, crime and delinquency.

3.4       More diverse populations, a more diverse workforce, more diverse lives

The working-age population is shrinking, but it is also becoming more diverse. At the same time, by 2030, the rate of women's participation in the workforce is expected to grow only marginally. The participation of older workers will grow more rapidly (those aged 50-64, to increase from 30 to over 35%) and that of workers with a foreign background[26] will increase considerably (see Figure 3.3). A greater number of older workers will also mean further adaptations in the workplace.[27]

Figure 3.4 — Share of population with a foreign background, EU-27[28]

Non EU-citizens have represented two-thirds of the EU population growth in the past decade[29] and half of employment growth in the past five years.[30] As a result, out of just over 30 million non-nationals in the various EU Member States, 20 million are non-EU citizens. Up to 10-15 million more people were born as non-EU citizens and later acquired citizenship of an EU Member State.

Migrants from outside the EU are generally younger than the population they join, so they rejuvenate it. These migrants also tend to be better educated than older workers, though they are generally less well educated than local-born labour market entrants,[31] so they play a more significant role in maintaining the size of the labour force than in up-skilling it.[32]

There are also country-specific aspects to increasing diversity. The Roma, for instance, are projected to increase from 5 % to 16 % of the Hungarian population of active age by 2050.[33] As low education and low employment are characteristics of many Roma at present, fostering their integration is a priority.

People’s lives are also becoming more varied. The traditional pattern whereby people, predominantly men, first studied, then worked, then retired is slowly being supplanted. People, especially women, tend to stay in education longer and wait longer before having children. Within their working lives, many people have spells of childcare and lifelong learning, reducing or discontinuing work temporarily.[34] People still migrate for long periods, but now many more have short work or study experiences abroad.[35] People are also leaving the labour market later. These new, more varied patterns of life mean that social protection systems have to adapt to the new realities.

3.5       Changing family structures

In recent decades, ever-growing numbers of women have been getting an education and going out to work. Among those born after 1960, more women than men obtain university degrees in almost all EU Member States. Each new generation of women has valued employment and economic independence more and managed to reconcile having children with employment more successfully. Those who work as young adults are also more likely to be working as they grow older, with employment rates growing for women of all ages.

Changes in values, opportunities and increasing mobility have led and continue to lead to changing family structures. Since 1970, there have been fewer marriages, falling from eight to five a year per thousand people, while the number of divorces has risen from one to two per thousand.[36] This has probably increased the number of single-parent and recomposed families. More children, now over a third, are born outside marriage. At the same time, fewer women (traditional informal carers) are available to look after dependants as demographic dependency rises. Overall, the male breadwinner model, on which much family policy has been grounded, is no longer predominant. This means adapting social policies to a new reality, ranging from pensions, education and care, health and elderly care and measures to reconcile work and private life.[37]

3.6       The development of information and communications technologies

Information and communication technologies (ICTs) are increasingly being used across Member States. In 2011, 77 % of households in the EU had access to a computer. The internet was available in 73 % of EU households in 2011, a sharp rise from 41% in 2004.[38] An estimated 180 million people are also using ICTs at work.[39]

In spite of increasing levels of regular internet usage in Europe, large disparities still persist between EU countries and between different socio-economic groups. Age and education are the main factors determining the probability of using the internet. Other socio-economic conditions such as household income or rural/urban living area also strongly influence internet use.[40]

Furthermore, when looking at young people, according to Eurostat, in 2011, 91% were regular internet users, or accessed the internet at least once a week. However, research on the relationship between youth, ICT and inclusion[41] points to worrying trends: (a) the lack of ICT skills of young people, for example digital reading or ICT skills to meet the labour market demands; (b) the fact that socio-economic status affects the frequency of ICT use, the skills and the types of ICT uses, where disadvantaged youth adopt more risky behaviours online and are more likely to be exposed to unsuitable and harmful on-line experiences; and (c) digital exclusion of those with lower socio-economic status is affected by cost, peer pressure, social context, attitudes, and school divides.

The development and increasing use of ICTs has the potential to promote cultural understanding between citizens, to seed innovation in institutions and to create competitive advantage for businesses and employees in the future. These technologies continue to play an important role in improving labour productivity and they also provide responses to major societal challenges. ICTs increase the effectiveness of existing processes and create enormous potential through 'network effects' ('collective intelligence') in order to find effective solutions to societal challenges.

ICTs are a key enabling tool in the development of new forms of work. Thanks to their pervasiveness, people can work remotely, on the go from virtual work stations and indeed also according to their own rhythm. This flexibility allows workers to take up employment positions which otherwise would have been too far away or not compatible with other personal obligations such as child or elderly care or with limitations in activities imposed by a disability.

The ICT sector has increasingly accounted for a substantial part of EU GDP and employment. For the EU as a whole, in 2009, the EU ICT sector represented 4.0% of EU GDP (€ 470 billion) and accounted for 2.7% total employment, amounting to 6.1 million people.[42]

ICT skills have become crucial for employability of individuals, supporting individual empowerment (development of self-confidence and self-efficacy) and the development of other skills, such as transversal skills (social networking, collaboration, problem solving, language skills), job search skills, and e-learning skills.[43]

There is also an increasing demand in the labour market for ICT skills, which not only increases the employability of those having them, but also gives people access to more and better jobs — more creative jobs where they can further develop their skills, enhance their career prospects and earn higher wages. Moreover, ICT skills facilitate access to the labour market as they help people to search for jobs more effectively and can reduce the duration of unemployment.

Furthermore, ICTs have dramatically changed the way job seekers look for jobs and employers recruit and select the best candidates. ICTs support the role of labour market intermediaries (social actors, employment services, etc.) that can offer new online assessment, labour market information and employment services to job seekers and employers, improving service quality, cost-effectiveness, supply-demand matching and reach.[44]

Demand for ICT specialists and people who have strong ICT skills is growing, while shortages of ICT practitioner skills have become endemic. The employment of ICT practitioners is growing so fast that there are not enough workers to fill all the vacancies available in the sector, even with increased unemployment during the economic crisis. According to recent estimates,[45] there will be up to 700,000 unfilled ICT practitioners' vacancies in the EU by the year 2015. The Employment Package underlines the potential for employment growth by developing the ICT sector and ICT skills.[46]

Outside of ICT specialist skills, most jobs nowadays already require some kind of computer-related knowledge. It is forecast that by 2015, 90% of jobs will need at least basic computer skills.[47] Possessing ICT skills is increasingly important in order to lower unemployment risks even among senior workers,[48] and ICTs can also be used to improve other work skills. ICTs can help increase high level skills and up-skill existing skills, for example, by enabling education and training through open and easily available learning resources, open educational practices and the use of open networks.

When exploring the links between ICT and the social inclusion of disadvantaged populations, an analysis of 61 initiatives using ICT to support the inclusion of disadvantage youth at risk of exclusion[49] shows that the short term effects or 'outcome' of ICT-based initiatives are associated with multi-dimensional impacts, typically combining empowering/skilling, social inclusion and social capital and employment-related benefits, supporting protective factors and resilience.

4. The social impact of the economic and financial crisis: divergence between Member States and growing inequality in the European Union

In the decade before the crisis, economic and employment growth in general improved overall living standards. Many governments devoted more resources to social policy intervention. Despite the clear redistributive effect of social protection, inequalities have often increased. Poverty and social exclusion remain major issues in most EU countries, though the pattern varies substantially across different countries.

The outbreak of the financial and economic crisis in late 2008 put further strain on households in the EU. The timid recovery generating some employment growth after 2010 has not proved durable and the ensuing economic slowdown in 2011 gradually turned into mild recession in the EU, as the escalation of the debt crisis in several Member States led to necessary policy shifts toward fiscal austerity by and large across the EU, with inevitable adverse effects on aggregate demand. As a result, the employment recovery from the 2008-2009 recession has come to a standstill. The number of people in employment in the EU grew by a modest 0.3 % of the working-age population in 2011, owing to the better first half of that year, and then has stagnated until autumn 2012.

Over the period 2008-2011, the employed population shrank by 2.0 %. Unemployment is again rising to historically high levels. The average EU unemployment rate is over 10 %. The rates for long-term and youth unemployment are especially worrying. In autumn 2012, 25 million Europeans were unemployed. More than 10 million of them had been unemployed for over a year, while in the United States and Japan unemployment has been slowly declining.

4.1. Growing divergence between EU countries

Employment and social indicators point to growing divergence among EU countries. Some southern and peripheral European countries have seen their unemployment increase sharply and their GDP contract considerably, while most of the countries of northern and central Europe have so far shown greater resilience. The latter tend to combine higher levels of competitiveness with better functioning labour markets and more robust welfare systems. The (in)ability to cope with the shock was frequently compounded by the initial public debt and deficit levels, as well as the property market situation, and subsequent developments followed by the reaction of financial markets.

EU countries have seen very different trends in unemployment, as shown in Figure 3.2. After converging in the years up to 2003-2004, in favour of southern and peripheral countries, it changed in favour of northern countries again in 2008, both within and outside the euro area. In 2011, the gap between the north and the south/periphery in the euro area reached 7.7 percentage points, but was smaller among non-Euro area countries (1.8 %). Labour market divergence has become a major issue for the EU.

Figure 4.1 — Diverging unemployment rates by groups of Member States since 2000, 15-74 age group, 2000‑2011

Source: Eurostat, EU LFS.

Much of the post-2008 period in most Member States has been marked by wage deceleration, including pay freezes and sometimes pay cuts. In 2011, this deceleration was increasingly concentrated in Member States with high unemployment and current account deficits, in efforts to reduce the wage bill in the public sector and regain competitiveness in the private sector. Changes in pay per employee ranged between plus 3 % in Finland, Belgium and Germany and a similar rate of negative growth in Greece.

Figure 4.2 — Compensation per employee, productivity and ULC in 2011 (annual % growth)[50]

Source: Eurostat, National Accounts

There is growing divergence between the countries where the social impacts of the crisis were somewhat limited, and countries where living conditions have deteriorated markedly. The fall in household disposable income was most significant (above 4%) in the southern countries, Ireland, Hungary and the Baltic States, the result of further deterioration in labour market conditions, together with weakening of the cushioning impact of social expenditure over time (see Figure 4.3).

In the Baltic States the rebound of the economic and labour market situation has contributed to stabilising income levels overall after 2010, but long-term unemployment and poverty remain at high levels.

In stark contrast, northern and continental countries fared far better. The combined effect of robust automatic stabilisers[51] (reinforced by initial discretionary measures) and more resilient labour markets in general helped mitigate the impact of the recession on overall household incomes and private demand. Still, while household incomes continued to rise during the crisis, there were still some groups badly affected by rising unemployment.

Between 2007 and 2009, automatic stabilisers and stimulus packages contributed to sustaining household incomes in most countries. Between 2009 and 2011, household incomes fell in a number of countries, especially where the recession was prolonged, due to: the weakening of the effect of automatic stabilisers (e.g. people losing entitlement) and the phasing out of discretionary measures taken at the onset of the crisis, combined with the introduction of fiscal consolidation measures (cuts in benefits and increases in taxes).

Figure 4.3 — Evolution of Gross Household Disposable Income (GHDI) in real terms (2005=100) 2005-2011 (2012 forecast)[52]



Source: AMECO

4.2. Trends towards greater social inequality in most countries

The evolution of incomes since 2008 results both from long-term trends[53] and from the specific impact of the recession. The initial deterioration of labour market conditions primarily affected workers at the lower end or in the middle of the income distribution. This impact was mitigated by the working of automatic stabilisers and initial discretionary measures. Over time, long-term unemployment and the weakening of the protective role of social transfers led to an increase in the share of people on low income in a number of countries. In addition (see section 7) fiscal consolidation measures have had differing redistributive impacts, in some cases exacerbating and in some cases mitigating inequality.

The crisis has not affected the whole population uniformly. It has in several respects worsened the position of people already at heightened risk, thus contributing to social polarisation. Young people have been particularly hard hit, and more than one in five active Europeans aged 15-24 are unemployed. Non-nationals, the low-skilled and men are also badly affected by deteriorating labour market conditions. There has been a levelling down of gender gaps in employment, unemployment, wages and poverty over the crisis. This does not, however, reflect progress in gender equality as it is based on lower rates of employment, higher rates of unemployment and reduced earnings for both men and women.

Even before the crisis, there was a trend towards polarisation of incomes. The polarisation of jobs took root in the decade before the crisis. Most of the jobs created between 1998 and 2007 were either low paid or highly paid[54]. This led to slow erosion of the number of middle-paid jobs (Figure 3.5). The growth of service jobs at the extremes of the wage distribution aggravated this trend.

The recession reinforced the wage polarisation observed in the past. During the current recession, the massive job losses in manufacturing and construction — and the ensuing collapse in middle-ranking jobs — drove wage polarisation for the most part. Job growth in the top wage quintile persisted strongly throughout the recession, mainly because of the rise of knowledge-intensive services (KIS), including public services (mainly education and health) and private services (business services). One of the main factors behind the "disappearing middle" was the polarisation of service sector employment, characterised mainly by employment growth at the top and bottom, which is likely to continue in the long run. This suggests that the service sector cannot be relied upon to fill the gap created by the decline in manufacturing. This trend toward further segregation means that those in lower-end employment are at growing risk of limited career mobility possibilities, exacerbating problems of job-skills mismatches and over-qualification.

Figure 4.4 — Net job creation (difference between the number of jobs created and the number of jobs lost — expressed in thousands of jobs) by wage level

               Before the crisis                                           During the crisis

Source: ESDE 2011

Another factor driving rising inequalities is the growth in involuntary part-time and temporary contracts. Persistent labour market segmentation is an important determinant of growing earnings inequality, not just because non-standard workers tend to work fewer hours per year, but also because they are generally paid less per hour after taking into account differences in education and experience. Non-standard contracts can also provide opportunities to get back into the labour market and act as stepping stones towards better jobs, as observed in countries with better functioning labour markets.[55]

In most countries, inequalities between workers were further aggravated by the polarisation of jobs between job-rich and job-poor households. Before the crisis, growth in employment was mainly driven by women entering the labour market. The new jobs mostly went to second earners in households that previously had only one earner. Households that included no earners were benefiting less. This is illustrated in the graph below, which shows that while the EU unemployment rate was reduced by 2 percentage points between 2004 and 2007, the share of people in jobless households was reduced by only 1 percentage point.

Figure 4.5 — Employment, unemployment and share of working age adults living in jobless households, EU-27, 2000-2011

Source: Eurostat — LFS

Welfare systems play an important redistributive role, but in the past decade, government transfers, the tax system and public services have not always been able to mitigate the rise in earnings inequalities (see ESDE 2012). Moreover, the current need for fiscal consolidation and the growing pressure of an ageing population are putting European welfare systems under stress, so there need to be efficiency gains.

The effect of the first phase of the crisis (2008-2009) was uneven across the income distribution. In a third of the EU countries the impact was regressive. In Ireland, Spain and Lithuania, the poorest segment of the population saw their income drop more than the rest of the population. In France, Denmark, Sweden and Slovenia, the incomes of the poorest segment of the population grew significantly less than households higher up the income distribution. In the UK, Hungary and Italy, people in the middle of the income distribution (often workers) were the most affected by the crisis. In a few countries of continental Europe (Germany, Belgium, the Netherlands, Austria and Poland) the trend was relatively even across the distribution. In Latvia, Greece and Portugal, the impact was progressive, with the top incomes proportionally more affected than those of the rest of the population.

Figure 4.6 — Changes in households' disposable income in different parts of the income distribution (% change in national currencies, 2007-2009); Member States grouped according to the degree of progressivity of the change

Source: EU-SILC

Consumer surveys[56] provide a more recent picture (up to spring 2012) of the trends in the way households themselves see their financial situation. The financial stress index monitors the percentage of people declaring that they had to draw on their savings or to go into debt to keep up current expenses. Trends up to spring 2012 show that most households saw a marked deterioration in their financial situation in most EU countries. All income groups were affected.

However, in some countries, those on low incomes felt the impact most, especially in Estonia, Greece, Spain, Italy, Cyprus, Latvia, Hungary, Malta, Portugal and Slovakia. This could herald significant rises in poverty and material deprivation levels in these countries and confirm the increase in subjective poverty that surveys recorded in the first phase of the crisis. According to data from EU-SILC, between 2008 and 2011, the proportion of people reporting that their household was only just making ends meet rose sharply (by more than 4 percentage points) in the Baltic States, Cyprus, Greece, Hungary and Ireland.. This includes countries where the poor may have lost proportionally less than the rich (see above), but nevertheless experience a stronger deterioration of their actual living conditions given that essential consumption items (food, housing) represent a larger share of their budgets.

Figure 4.7 — Reported financial distress in households by income quartile in the EU and in selected Member States (2000-2012)[57]


Source: Joint harmonised EU consumer surveys & DG EMPL calculations

5. The evolution of poverty and social exclusion 5.1. The multiple facets of poverty and social exclusion

This section deals with trends in poverty and social exclusion in the EU. EU-SILC data confirm the stagnation of relative poverty between 2005 and 2011, but at the same time show that living standards improved in EU-12 countries before the crisis, as measured by severe material deprivation rates.[58] Material deprivation rates eased in EU-12 countries in parallel with the relative rise in GDP per capita in these countries. The crisis has brought this overall improving trend to a halt and material deprivation started increasing again in many countries, especially those most affected by the crisis.[59]

The impact of the crisis is most visible in the percentage of jobless households[60]. This has started to rise again since 2008, especially in the EU-15 (see next section). Related and significant income losses are also reflected in the increase in material deprivation in some countries (up to 2010-11).

Figure 5.1 — Population at risk of poverty and social exclusion in the EU and in selected Member States


Source: Eurostat, EU-SILC 2011, 2010 data for Ireland.

Figure 5.2 — Developments in the at-risk-of-poverty or social exclusion (AROPE) rate and its components in the EU, 2005-2011

Source: Eurostat, EU SILC. Note: 2005, 2006 and 2011 are Eurostat estimates.

Poor health, lack of access to health care, poor or unaffordable housing, poor educational outcomes (especially for early school leavers and NEETs[61] ) or poor access to education, and the deterioration of social participation are important dimensions of social exclusion. The impact of the crisis on these is more difficult to capture, but is likely to have long-term detrimental impacts on the formation and quality of human capital. However, a recent paper by the Commission’s Social Situation Observatory (SSO)[62] finds that the financial crisis has had a statistically significant negative effect on health, in the sense of a significant increase in the number of people reporting poor health after the crisis (i.e., between 2006 and 2009 in Greece and Portugal).

Improving access to healthcare is clearly an important aspect of preventing and tackling social exclusion. The EU_SILC indicator of ‘unmet need for medical care' (due to barriers in access to the health care system[63]) improved in most Member States between 2008 and 2010. While slight deterioration occurred in Finland, France and Malta, Latvia and Poland stand out as the only two countries with a marked worsening. The significant income gradient has been maintained across the EU in the reference period of 2008-2010. More recent information stemming from the annual Social Climate Eurobarometer[64] surveys carried out between 2009 and 2012 show a clear downward trend in the 'judgement of the current situation in the health care provision'. The index for Greece went down from –3.1 to –6.3 (on the scale from –10 to +10). Important negative changes in the 'judgement of the current situation in the health care provision' have also been observed in Italy (from –0.1 to –1.8), Latvia (from –1.8 to –2.8) and Portugal (from –1.0 to –1.9). Generally, the countries scoring poorly are the ones heavily affected by the crisis. Yet, the perceptions became only slightly worse in Spain and remained stable in Ireland. Improvements were observed in Germany, France, Belgium and the Netherlands. Poland represents a special case of a country without a GDP decrease and with a sharp drop in the judgement of health care provision to very low levels (from –1.7 in 2009 to –4.0 in 2012). See the Staff Working Document on Investing in Health for concrete policy responses.

Figure 5.3 — Self-reported unmet need for medical care[65], by income quintiles in selected Member States, 2008-2010

 Source EU SILC 2008 2010

5.2. The gender dimension of poverty and exclusion

Overall, women face a higher risk of poverty and exclusion than men. The crisis has not widened this gap so far, since prime age men have been most directly hit by the deterioration of labour market conditions. However, some categories of women face significantly higher relative risks of poverty — such as lone mothers, inactive women of working age (27%), and elderly women over 75 (20.7%). Profiling of populations exposed to the risk of poverty show that women are more often represented in groups facing higher risks of persistent poverty, linked in particular to inactivity and care responsibilities, which have long-term impacts on future pension entitlements. While inactivity rates have not increased so far as a consequence of the crisis, retrenchments or freezes in social spending may hamper female participation and aggravate the situation of the most vulnerable women, depending on which expenditure items are affected (e.g. reducing spending on childcare facilities will hamper the labour market participation of mothers).

Figure 5.4 — Male and female at-risk-of-poverty-and-social exclusion rates, and the underlying three components, 2008 and 2011[66]

Source: Eurostat, EU-SILC.

The greater exposure of women to the risk of poverty and exclusion partly results from the combination of three gender gaps. Women's activity rate is still 12.7 percentage points below that of men (64.9 % against 77.6 % for the 15-64 age bracket in 2011), reflecting persistent gender divisions in household and care responsibilities. A higher proportion of women works part-time, which means that on average women work 17.0 % less hours than men (33.7 hours per week against 40.6 in 2011). Finally, the gender pay gap of 17 % (in average hourly gross wage) is partly due to women earning lower pay for work of equal value, and partly due to women being concentrated in jobs that pay less. As a result, women's annual gross labour market earnings can be estimated to be 42 % below those of men on average,[67] resulting in lower GDP, lower social security contributions and higher risk of poverty in old age (see ESDE 2012). Older widows are particularly at risk, as discussed in section 6.4.

A large share of women entered the labour market as second earners and thereby improved the income situation of households already at work. However, in many countries the women who are furthest away from the labour market (lone mothers, the low-skilled, etc.) still face major barriers to finding a job due to lack of childcare or care for other dependants, or lack of measures to improve the work/life balance. They also often have difficulties in finding a job that pays because of involuntary part-time work, and the combined effect of high marginal effective tax rates and high costs of childcare. Work disincentives for secondary earners due to joint taxation systems in some Member States can also exacerbate this.

5.3. The depth and duration of poverty

The crisis has increased or is likely to increase the duration and depth of poverty. These deteriorations in the persistence and/or depth of poverty constitute a serious challenge for Member States, requiring more specific actions to prevent the situation deteriorating further and to tackle the deepest roots of poverty.

While relative poverty remained stable in the EU and in most countries, the poverty threshold declined dramatically in a number of countries, reflecting the general decrease of household incomes. Between 2008 and 2011, it fell by 17.4 % in Latvia, 12.5 % in Lithuania, 10.7 % in Ireland, 8 % in Spain, 7.7 % in the UK, 7 % in Greece and 6.7 % in Estonia. This shows that while their relative situation is not deteriorating, people living on low incomes are facing a serious reduction of their resources that were already considered insufficient to maintain a decent living standard before the crisis.

Figure 5.5 — Developments in the at-risk-of-poverty indicator from 2008 to 2011 and the underlying poverty threshold


Source: Eurostat, EU-SILC.

Note: Countries sorted by change in deflated poverty threshold.

In many countries, the poor are getting poorer as the gap between the median income of the poor and the 60 % threshold is increasing (the poverty gap was 23.3 % in 2011, up from 21.7 % in 2008). Between 2008 and 2011, the poverty gap increased for all but a few Member States and with especially strong rises (around 3 pps or more) in high-poverty countries such as the Baltic States, Slovakia, Italy and Spain, as well as in some countries with a low incidence of poverty (Austria, Denmark).

Figure 5.6 — Deepening of the risk of poverty: change in the at-risk-of poverty gap 2008-2011[68]

Source: Eurostat, EU SILC

In 2010, 8.5 % of the working-age individuals were at persistent risk of poverty, meaning that they had been at-risk-of-poverty in at least three out of the previous four years (including the last one). Persistent poverty is high (10 % or more) in Italy, Greece, Portugal, Bulgaria, Romania, Poland and Ireland. Young adults, inactive or unemployed women, lone mothers, or older working-age adults out of the labour market are among those facing higher risks of persistent poverty. Typical profiles vary across countries, suggesting that these people face specific structural and institutional barriers in different countries.

Figure 5.7 — Persistent risk of poverty compared to risk of poverty, 2010 data[69]

Source: Eurostat, EU SILC, ilc_li21

5.4           The most severe forms of poverty and social exclusion: focus on homelessness, the Roma and migrants

Homelessness often results from a conjunction of adverse personal, structural and institutional circumstances. The crisis has had a strong impact on both the gravity and the extent of homelessness. There is a growing share of migrants, young people and unemployed or low-income people among the homeless. Women, families and elderly people are also increasingly seeking shelter.

The scale of homelessness is very difficult to fully assess, as homeless people do not show up in traditional data collection processes such as household surveys. However, according to the 2010 Eurobarometer[70] survey more than 3 million Europeans were at the time afraid of becoming homeless and nearly three out of four Europeans (73 %) thought that homelessness had risen in their country over the last three years up to 2010.

Data[71] collected by municipal, local or national organisations show a general trend towards a growing number of people becoming homeless because of economic hardship. Some countries, e.g. the UK, Ireland and Estonia, despite suffering from recession and the collapse of the housing bubble, managed to contain the spread of homelessness through effective assistance schemes. For a more detailed discussion of the scope and severity of homelessness across the EU and its policy implications, see the Staff Working Document on homelessness as part of the Social Investment Package.

Figure 5.8 — Overview of recent trends in the extent of homelessness over the past 1-5 years

Trend reported || Member State || Total

Increase || AT, CZ, FR, DE, ES, EL, HU, IE, IT, LT, PT, PL, SE, SL, , , UK (England + Wales) || 15

Decrease || FI, NL + North Rhine-Westphalia, Scotland || 2

No trend identified || RO, LU, BE || 3

Stable || DK || 1

Total number of Member States examined: 21

Source: FEANTSA Country Report 2012

Ethnic minorities, including the Roma, are not identified in the EU-SILC. However, 2011 survey data provided by the Fundamental Rights Agency (FRA), the United Nations Development Programme and the World Bank show that Roma experience more severe poverty and social exclusion than other groups in the societies in which they live.

This is due to a complex mix of factors, primarily low educational attainment, low levels of employment, significantly worse health status, and poor housing and living conditions. Moreover, poverty and social exclusion experienced by Roma is often intensified by discriminatory treatment and prejudice.

In all 11 EU Member States covered, survey results show that more than 8 out of 10 Roma surveyed are at risk of poverty. The differences between Roma and non-Roma are most marked in France and Italy, where the proportion of Roma living in households at risk of poverty is more than twice as high as for non-Roma living in the same geographic area. The results are also significant in the countries with the largest Roma communities, such as Hungary, Romania, Slovakia and Bulgaria.

Figure 5.9 — Persons living in households at risk of poverty (%)

Source: FRA Roma Pilot Survey (2011), UNDP/World Bank/European Commission Regional Roma survey 2011

Note: Non-Roma here refers to non-Roma living geographically close to areas with a concentrated Roma population.

Migrants and mobile EU people in the European Union face considerable obstacles getting jobs and are more often unemployed than native-born populations, especially in the case of non-EU born migrants. The linguistic, cultural and labour market acclimatisation of migrants can be a long process, hampered yet further by other barriers such as discrimination.

The risk of poverty or exclusion among the migrant population remains much higher than among the EU population overall. For people aged 18+ born outside the EU-27, it stood at 37.8% in 2011, compared to 20.8% for those born in the country and 22.2% for those born in another EU country. Between 2008 and 2010, people born outside the EU have seen the sharpest rises in the risk of poverty or exclusion in Spain (a 3 pps increase), while in France and Italy, citizens from other EU Member States have been worst affected (see ESDE 2012).

5.5           The economic returns on social investment and the costs of unemployment, poverty and social exclusion

Poverty and social exclusion are linked to adverse health,[72] lower literacy, poor school performance for children, higher stress levels for families, and more crime and social unrest. Efforts to illustrate or quantify the costs of unemployment, poverty and social exclusion generally refer to higher public health care costs, increased policing and crime costs, foregone economic activity, lost wages and productivity, lost tax revenues and the intergenerational costs that flow from the likelihood that a significant number of children from disadvantaged families will remain disadvantaged over their lifetimes. Accordingly, a comprehensive approach to the costs of unemployment, poverty and social exclusion would require a very broad range of effects to be taken into account.

A first step in this direction can be to focus on the opportunity costs (accounting for the direct economic loss in labour market incomes) and the related additional public resources (benefits and taxes) rather than on the overall general costs of poverty (for instance health and crime outcomes, intergenerational transmission of poverty, or life course impacts on children experiencing poverty).

From this perspective, the gain that some people would experience from having better access to the labour market can be reflected in their higher labour market incomes. The assessment can reflect the general economic impact (higher individual value added translating into higher wages and higher GDP), distinguishing between both the private side (higher net incomes) and the public side (fewer benefits and higher tax revenue).[73] An assessment of the gains associated with the return of people of active age to employment can be derived from EU wide surveys such as SILC. This can be obtained by directly comparing average labour market outcomes (and related benefits and tax levels) of poor and non-poor people having the same characteristics (such as education, age, sex, household type) or by relying on matching methods, thus enabling estimates that better reflect the heterogeneity of situations, as well as actual assessments of the overall impact on poverty rates (linked to possible changes in the median incomes).[74]

Figure 5.10 — Estimates of the opportunity cost for the poor and unemployed aged 25-59 (size of the population and gains in % GDP)[75]

Size as % the population                                                    Gains as a % of GDP

Source: EU-SILC 2010 — SSO calculations.

Note : 'size' refers to the number of people affected (i.e., poor unemployed people who are coming back to work), or in other words the shock expressed as a % of total population or total poor population. 'Gains' refers to the opportunity costs of poverty (or gains of moving out of poverty), associated with the corresponding shock (as a % of GDP).

For instance, the impact of poor unemployed people getting jobs is estimated to account for around 1% of GDP in some Member States and to 0.5% of GDP or more in over half of the Member States (see Figure 5.10). Relying on micro-simulation tools, such as Euromod would enable more accurate estimates to be made, since actual changes in benefits and tax levels could be accounted for after changes in attachment to the labour market, through simulations of employment shocks (for instance among the working poor, or the unemployed or inactive poor).

Estimations of the return from measures to encourage people into jobs could be derived from an assessment of the economic gains if some categories of poor people get jobs, by comparing estimates of the cost and efficiency of various measures such as active labour market policies.

Broader approaches to estimating the economic returns on social investment have been the focus of various studies. For instance, a Commission study[76] from 2002 estimated the return on investment in human capital at the microeconomic and macroeconomic levels. The study suggests that an additional year of schooling increases wages at the individual level by around 6.5 % across European countries. Similarly, a year of training is estimated to increase wages by as much as 5 %. At the macro level, the study suggests that an extra year at the intermediate level of education — all other things being equal — increases aggregate productivity by about 5 % immediately and by a further 5 % in the long term.

A December 2011 Eurofound study estimates the economic cost of the group of young people that are not in employment, education or training (NEETs, who represented 13% of the 15-24 in 2010) in 21 EU countries. The yearly total cost for these countries amounts to approximately €100 billion, which corresponds to 1% of their aggregated GDP. It can be split into €94 billion in foregone earnings and €7 billion in excess transfers. Authors note that the estimation is restricted to the current cost only (in 2008) and may be an underestimate of the real cost of NEETs, as additional costs for health, criminal justice and unpaid taxes on foregone earnings are not included in the definition used.

Recent evidence gathered by Bell and Blanchflower (2011) highlights the long-term scarring effects of youth unemployment in terms of both career and earnings prospects for the individuals, and wasted human capital of a generation affected by the crisis. Their results show that entering the labour market during a recession often leads to substantially lower lifetime earnings for graduates and increased risks of ending up in lower-level occupations. They emphasise the delayed negative impacts of unemployment when young on well-being, health status and job satisfaction, pointing out that short-run government savings may be at the cost of increased future expenditures associated with the negative effects of youth unemployment and reduced well-being.

A UK study[77] estimates that by reducing youth unemployment, up to £10 billion per year could be saved. It takes account of spending on unemployment benefits (-€1.2 bn per year), lost productivity (-€6.9 bn per year), the cost of crime (-€1.2 bn) and the cost of educational underachievement (-€22bn for those aged 17-24). In the UK, the Joseph Rowntree foundation[78] estimated that child poverty costs at least 2% of the GDP, £25 billion a year, including £17 billion that could accrue to the Exchequer if child poverty were eradicated.

6. Risks over the life course and opportunities for intervention

The risk of social and economic disadvantage occurs at different stages in the life course but often builds upon earlier experiences and is compounded over time.

Numerous studies show that children growing up in poverty have lower education achievement scores, and that the gap between their scores and those of students from higher-income backgrounds widens over time.[79] Levels of educational attainment for students from low-income backgrounds are also lower, resulting in a higher risk of unemployment and lower future earnings potential later in life. Unemployment and low earnings during prime years continue to have an impact on a person's situation in old age, as low pension contributions often affect the level of pension available upon retirement.

6.1           Children

Children are generally more at risk of poverty or social exclusion than the overall population, with a rate of 27.1% as against 24.2% for the population as a whole in the EU in 2011.[80] Only in a minority of EU countries are children less at risk than the overall population.

Children growing up in poverty and social exclusion are less likely than their better-off peers to do well at school, to enjoy good health or to realise their full socio-economic potential later in life. This is reflected in the Recommendation on Investing in Children, as part of the Social Investment Package.

While it is difficult to estimate the exact degree of transmission of inequality and disadvantage across generations and how this changes over time,[81] recent evidence suggests that intergenerational inequality could be higher than analysts believed a decade ago.[82] A recent UK study showed that children of low-income backgrounds have lower achievement rates in school, and that differences in family parental resources widen during the school years. As a result, there are  wide gaps between of different income groups as regards achievement at school. Poor performance at school also reduces future earning potential in the long run.[83]

Another study showed that in England, for instance, the highest-performing 15-year-olds from poor backgrounds lag, by around two years of schooling Pisa literary test scores, behind the highest-performing pupils from privileged backgrounds.[84] Evidence also shows that growing up in a disadvantaged environment can have a long-lasting negative impact on adult health, and that upward social mobility has very little effect on this.[85] Yet, as highlighted in the figure below, the degree to which disadvantage is transmitted across generations varies significantly across countries.

Figure 6.1 — Link between individual and parental earnings varies across various countries[86]

Source: d'Addio (2007)[87]

The crisis has also worsened children's situations in most of the Member States, mainly because of the sharp rise in unemployment that has hit adults of working age. Single-parent households have been worst hit. They face a much higher risk of poverty and social exclusion (above 50 %) compared to other household types. However, even families with two adults and two children were exposed to greater risks of poverty or social exclusion.

Figure 6.2 — At-risk-of-poverty or social exclusion rate in the EU (%), children (0-17), EU-27, 2008 and 2011[88]

  Source: Eurostat, EU-SILC  

Child poverty results from the conjunction of parental wages that are too low and inadequate support to households. Social expenditure targeting children can play a supporting role to compensate for the cost of raising a child. However, its impact varies significantly across the EU, resulting in uneven outcomes in poverty reduction and redistributive impact. The role of parental employment and of the tax and benefit systems in preventing child poverty and supporting families with children is analysed in the September 2012 issue of the Employment and Social Situation Quarterly Review, and in the SPC report.[89]

The section below draws largely on these reports.

Parental employment is the main safeguard against child poverty, but sometimes it is not enough. In 2010, 10.7% of the EU working population living in a household with dependent children had an income below the national poverty risk threshold, against 8.5 % of the total working population. In most countries, the family with only one breadwinner is a model that no longer protects its members from poverty. The risk of poverty for individuals in households with low to medium work intensity (typically represented by the one-breadwinner family model) ranges between 15 % and 50 %. Families in which both members of a couple have a job are less likely to be at risk of poverty.

The extent to which second earners, often women, take a job is partly determined by the expected income gain, after deducting potential additional taxes, the loss of benefits and the cost of childcare. An OECD study (OECD, 2011[90]) shows that most women entering the labour market as second earners have weak financial incentives to do so. Before including the cost of childcare, many mothers of young children working full-time are unable to increase their family income by even 50 %. Net childcare costs are also a critical factor for parents' employment decisions. Compared to a situation in which no childcare is bought, the financial reward from employment is substantially reduced by childcare costs. At low earnings levels, single parents find that childcare costs reduce the returns by as much as 40 %; for second earners, this can be up to 50 %.

Child and family benefits and, indeed, social protection as a whole have a significant impact on reducing child poverty, by over 40% on average in the EU. However, the impact varies greatly across Member States.

Differences in efficiency can be partly explained by differences in the structure of benefits (e.g. by function) or their design (e.g. combination of universal versus categorical or means-tested, combination of cash or in-kind benefits). Differences in the demographic characteristics of households and in the pre-benefits risk of poverty also matter.

Figure 6.3 — Poverty reduction effect of family and child benefits for children (0-17)


Source: EU-SILC 2010, UDB, DG EMPL calculations; IE and CY not available.

As illustrated above, affordable childcare, along with appropriate tax and benefit incentives, is very important in enabling parents, and especially mothers, to get a job. For children, quality childcare and other services are essential to their well-being and to help them develop the social, cognitive and emotional skills that can enable them to be successful. Many Member States are lagging behind the so-called Barcelona targets for childcare set in 2002.[91] Besides, evidence[92] shows that children from disadvantaged backgrounds, who would benefit most from quality early childhood education and care, are far less likely to benefit from such services. Households on low incomes face barriers in getting access to childcare. These can include eligibility criteria, for instance, if employed parents get priority; lack of services, especially in disadvantaged or rural areas; or services may simply be too expensive. Only one out of two Roma children attend pre-school or kindergarten in BG, EL, HU, IT, PT, SK, CZ, FR, PL, RO, ES on average.[93]

Health status plays a determining role in defining children’s future life chances. There are still many obstacles to good healthcare for children. National and international studies point to the existence and persistence of social inequalities in children’s health outcomes and access to prevention in various fields across the EU. The 2009 EU-SILC module on material deprivation found that the rate of children with an unmet need to consult a general practitioner or dentist is much higher among those at risk of poverty or materially deprived. It also found that 11 % of the children at risk of poverty did not eat fresh fruit or vegetables once a day because the household could not afford it, against 3 % for the rest of the population.[94]

Differences in educational performance and in early-school-leaving rates can also be observed by gender. In education, the share of low-achieving boys is close to twice the share of low-achieving girls. In almost all EU Member States a lower proportion of boys reaches upper secondary graduation compared to girls and on average early school leaving is more than 30 % higher amongst boys than amongst girls.[95] Research on gender and education points to the significance of gender stereotyping in shaping educational outcomes.[96]

Most indicators of housing quality underline that children (and consequently families) are at a greater disadvantage than the rest of the population. Almost a quarter of children (and 40 % of children at risk of poverty) live in overcrowded[97] accommodation, compared to 18 % of the EU population as a whole. Households with children face a slightly higher risk of being overburdened by housing costs, especially in the southern countries. Children at risk of poverty are more likely to be living in unhealthy and unsafe housing conditions.[98]

Almost 1 million children are estimated to be living in alternative care[99] in the EU.[100] The high number of children from a disadvantaged background in alternative care[101] and the reasons they are there are often associated with poverty and social exclusion. This[102] suggests that more support to families at risk would enable more parents in difficult situations to care for their children themselves. If removal from the family is considered as being in the child’s best interest, then it is essential that children are placed in a supportive, secure environment that helps them develop to their full potential.[103] The negative consequences of large residential institutions on children’s health and psychosocial development are well known.[104]

6.2           Youth

The unemployment rates for young people are generally 2.5 times higher than that for the population of working age as a whole. They have been hit particularly hard by the economic crisis. The lack of jobs has been felt particularly severely by young people aged 15 to 24. This has just exacerbated their already weak position relative to other age groups. The unemployment rate for young people reached 23.7 % in November 2012.

Figure 6.4 — Youth unemployment and year-on-year changes, July 2012

Source: European Commission

The fall in employment for young people was most pronounced in permanent and full-time jobs. The relative resilience or expansion of temporary and part-time jobs was not enough to make up for the falls recorded in permanent and full-time employment. More than 40 % of young employees in the EU have temporary jobs, and that proportion has grown during the downturn. In the first quarter of 2012, the percentage stood at 40.6 %, against 13.1 % for the population of working age as a whole.

The deterioration in the employment situation for the young has also led to a sharp rise in the number of young people who are not in employment, education or training (NEET). This can be seen as a specific measure of youth social exclusion, reflecting their lack of contact with the labour market and education and has implications not just in the present, but also for their future inclusion in society. Risk of NEET status is disproportionately linked to low educational attainment and early school leaving. Many of those concerned also lack soft skills, vocational training, and work experience to navigate the transition into the labour market when leaving school.[105] This often indicates earlier social exclusion, during childhood years.

In the first quarter of 2012, 13.2 % of young people (7.5 million) fell into the NEET category, up sharply (by around 1 million) on the 10.7 % registered four years earlier. This raises concerns about the potential consequences and implications of NEET status. The risk of NEET disaffection has been investigated by a Eurofound study[106] exploring the consequences of the social and political marginalisation of youth; it estimated that the cost of NEETs adds up to 1 % of GDP in the 21 EU Member States studied. The analysis also revealed that NEETs have low trust in institutions and a low level of political and social participation. Empirical evidence confirms that NEETs are at a higher risk of disaffection and more likely to withdraw from society. In response to this the European Commission presented a Communication on a Youth Employment Package with concrete proposals and measures aimed at combating youth unemployment, including a proposal for a Council Recommendation on Youth Guarantees.[107]

Figure 6.5 — NEET rates across EU Member States (2008-2011)

While NEET status is more linked to low educational attainment and early school leaving, the labour market situation of young people who have recently graduated from education and training has also deteriorated. In fact, the employment rate of individuals aged 20-34 that recently graduated from at least upper-secondary education has fallen by almost 5 percentage points since 2008, to 77.2 % in 2011.[108] This figure highlights the particular difficulty that young people have transitioning from school to employment.

6.3           Working-age populations

Employment rates increased in all EU countries before the crisis. In the same period, the numbers of working poor remained unchanged and the number of people living in jobless households was only reduced marginally, while overall poverty rates were not significantly reduced. The jobs created often did not reach the most excluded, or did not provide decent living standards. These trends were mainly due to a growing gap between job-rich and job-poor individuals and households, and to disparities in earnings and working conditions among workers.

Education and skills level is a major factor in employment. In 2011, the long-term unemployment rate was more than four times higher for those with lower edu­cation levels (7.9 %) than it was for the highly educated (1.9 %) and more than twice as high as it was for those with a medium education level (3.7 %).[109]

Figure 6.6 —  Long-term unemployment rate by educational level, as a percentage of the active population, 2011[110]

Source: DG EMPL calculations based on Eurostat, EU-LFS

Having a job however does not always avert the risk of poverty, however. The working poor account for a third of adults of working age at risk of poverty. In 2010, 8 % of those in employment were living under the poverty threshold.

In-work poverty significantly increased in a third of EU countries between 2006 and 2010. They included Germany (+2pps), and the Netherlands and Denmark, where labour market reforms (e.g. wage moderation) and reforms of the tax and benefits system contributed to bringing more people into jobs, but did not always provide them with a living wage. Since 2000, there has been a growing trend towards temporary work, part-time work (including situations where that is all the employer offers, rather than being the choice of the employee) and sometimes stagnating wages. Taken together, these factors have increased the number of people on low yearly earnings. These trends particularly affected women and the young. It is also important to note that in many countries, such jobs are not stepping stones towards better ones  —  they are so-called ‘dead-end jobs’(see ESDE 2011 — chapter 4).

In-work poverty is linked to aspects of poor-quality jobs, such as low pay, low skills, precarious employment and under-employment. In-work poverty is also related to low work intensity in the household, i.e. situations where there are too few adults working in the household, or not working enough to earn a living (working too few hours or for only part of the year). Among these, single-parent households where the parent is not working full-time, and single-earner families face the highest risks of poverty.[111]

Figure 6.7 —  In-work poverty: at-risk-of-poverty rate of persons employed, change since 2008[112]

Source: Eurostat – EU-SILC.

The last decade has also seen the persistence of groups that remain outside or on the fringes of the labour market, often facing multiple barriers to getting a job. Low skills, care responsibilities, age, a migrant background, and other factors of discrimination are among these. Those worst-off are in households in which nobody works. In 2010 in the EU-27, 9.9 % of children and adults of working age were living in jobless households (i.e., in households with zero or very low work intensity), against 9 % in 2008. The crisis has already started to increase the number of families having to rely entirely on social benefits.

Figure 6.8  — Developments in the share of people living in jobless/very low work intensity households across EU Member States, 2008-2011

Experience from past crises shows that the numbers of those affected by long-term unemployment or inactivity tend to persist long after recovery has set in.[113] The main reason for this is the loss of human capital and skills caused by long periods spent out of work. In 2011, nearly 10 million unemployed Europeans (accounting for 4.2 % of the active population) had been out of a job for more than 12 months. This is an increase of 3.7 million or 60.8 % in comparison to 2008. Trends in long-term unemployment are becoming increasingly diverse among Member States. Between 2008 and 2011, the long-term unemployment rate rose in almost all Member States, with a particularly strong rise (more than 5 pps) in Greece, Spain, Ireland and the Baltic countries.

Since 2008, the probability of someone unemployed finding a job has decreased in most countries, for both the short- and the long-term unemployed. This trend has been particularly marked in Spain (from 50 % to around 30 %) and Greece (from 25 % to 15 %). On the other hand, the rate has remained stable in the Netherlands and has improved in the Czech Republic and Estonia. The risk of those who have a job falling back into unemployment has also risen. In Spain, Greece and Cyprus (but also in Sweden and Finland) there is a high risk of short spells of unemployment recurring, because of the large number of temporary contracts.

Figure 6.9 — Transition rate from unemployment: situation one year after unemployment, by duration in unemployment, in 2010-11, aged 25-49, EU-13

Long-term unemployed ||

Short-term unemployed ||

Source: Eurostat, EU-LFS, DG EMPL ad-hoc calculations.

The lingering effects of the crisis and the rise in long-term unemployment have sharply increased the risks of long-term exclusion. Evidence[114] shows that, before the crisis (2005-2009), the risks of entering and exiting poverty varied greatly across Member States. There are three main groups of countries:

Group 1: Austria, France and the UK. Rates of entry into and exit from poverty are high, although in some of these countries, a significant share of those at risk of poverty form a ‘core group’ that does not take part in the churning.

Group 2: Baltic States, Bulgaria, Greece, Italy, Malta and Spain. There is a high risk of entering into poverty, and low chances of getting out of it, creating a massive poverty trap. The evidence relates to pre-crisis data; the trap has certainly got worse since.

Group 3: Nordic and Benelux countries. Rates of entry into and exit from poverty are low. However, the proportion of people at risk of persistent poverty is high, which may be a sign of social polarisation, with a group of people trapped in poverty.

Figure 6.10 —Share of persistent poor crossed with entry and exit patterns

Source: EUSILC LONGITUDINAL  UDB 2009 version 3of August  2012;  DG EMPL calculations

6.4           Older people

In half the Member States, the oldest generations (aged over 65) face a lower risk of poverty than the population as a whole. But the risk of poverty is relatively high for the elderly in Cyprus, Bulgaria, Greece, the United Kingdom, Slovenia, Spain, Belgium and Portugal. However, the at-risk-of-poverty rate does not take into account housing costs,[115] and might, in some cases, overestimate the extent of poverty among the elderly as they often own their own housing, so do not have mortgage repayments or rent to pay.

The oldest tend to live on lower incomes and those aged 75 and over tend to be at greater risk of poverty. This reflects the lower levels of payments from pension systems developed in the 1950s and 1960s. It can also be attributed to lower accrued pension entitlements and a lower number of years worked, especially among women.

The gap between men and women facing poverty varies with age. It is clearly worse for people over 65 than it is for younger generations. Differences in life expectancy mean a rise in the number of widows and therefore single women. Because they have worked a lower number of years than men, older women often receive lower pensions, though in many Member States, survivor’s pensions do give widows some protection from poverty.

Figure 6.11 — The risk of poverty for elderly people by gender,[116] EU-27, 2011

Source: Eurostat, EU-SILC, 2011

Men’s pension incomes are usually higher than women’s. The gender pension gap is due to differences in employment rates and employment conditions during working life, e.g. the gender pay gap, and an unequal distribution of roles when it comes to looking after others, but can also be due to the design of pension schemes and trends in pension reforms.

More numerous among older cohorts, women are also more exposed to the risk of poverty, but they may receive better replacement rates (i.e., pension as a percentage of pre-retirement labour income) and better returns on their pension contributions since they are the main beneficiaries of minimum, guaranteed and survivor's pensions.

Figure 6.12 — Gender differences towards the risk of poverty in older age groups, 2011[117]

Source: Eurostat, EU-SILC, 2011

[1]      European Commission (2012) Draft joint employment report,.

[2]      European Commission (2012) Employment and social developments in Europe 2012.

[3]      European Commission Communication - Rethinking Education: Investing in skills for better socio-economic outcomes. COM(2012) 669 final

[4]      Musgrave, R. A. (1959) The Theory of Public Finance: A Study in Public Economy.

[5]      Vandenbroucke, F. and Vleminckx, K. (2011) ‘Disappointing poverty trends: is the social investment state to blame? An exercise in soul searching for policy makers’, CSB Working paper, No 11/01.

[6]         Esping Andersen, G. (1992) ‘The making of a social democratic welfare state’, in M. Misgeld and L. Amark (eds), Creating social democracy, a century of the Social Democratic Labor Party in Sweden, The Pennsylvania State University Press.

[7]      Van Kersbergen, K. and Hemerijck, A. (2012) ‘Two decades of change in Europe: the emergence of the social investment state’, Journal of Social Policy, Vol. 41, Issue 03.  See also Hemerijck (2012). Changing welfare states, Oxford.

[8]      Morel, N., Palier, B. and Palme, J. (2012) 'Towards a social investment welfare State? Ideas, policies and challenges' The Policy Press, University of Bristol.

[9]      Such as unemployment, old age and survivors pensions, healthcare, sickness, disability, family, housing and social assistance.

[10]    OECD (2012) Employment Outlook

[11]    European Commission Communication - Rethinking Education: Investing in skills for better socio-economic outcomes. COM(2012) 669 final

[12]    European Commission Communication - Efficiency and equity in European education and training systems, COM(2006) 481 final

[13]    European Commission Staff  Working Document – Investing in Health SWD(2013) 43

[14]    The apparent fertility increase in most EU Member States since around 2000 may be simply due to women having children progressively later in their lives, as they extended their education; rebalancing in the number of children per woman also contributed to the effect. According to this view, fertility has stabilised (already) in the EU — after decreasing from its high values in the 1960s— at a level slightly below 2 children per woman. 

[15]    Dependency comes in different forms, including 'demographic dependency', i.e., the number of people outside the working age (20-64 in the EU) divided by the number of those of working age; and 'economic dependency', i.e., the ratio between people not in employment over people in employment.

[16]    European Commission (2012) Ageing report.

[17]    Source: Eurostat and DG EMPL. The projections are based on the assumption of constant employment rates per each age, sex, origin (EU-27 or non-EU-27) and education level (ISCED 0-2; 3-4; and 5-6); education attainment rates are assumed to continue increasing in line with the trend of the past five years.

[18]    See European Commission-OECD Health at a glance: Europe 2012

[19]    See European Commission White Paper -An Agenda for Adequate, Safe and Sustainable Pensions - COM(2012) 55 final

[20]    In 2000-2010 EU-27 employment grew by almost 1% per year, which was half the real GDP increase.

[21]    Including only people aged 15 and over. Source: Eurostat, Labour Force survey, special query.

[22]    See European Commission (2010) Demography report, pages 91-92 (

[23]    See European Commission (2011) Report on the Functioning of the Transitional Arrangements on Free Movement of  Workers from Bulgaria and Romania

[24]    And possibly more, according to preliminary indications from censuses around 2011.

[25]    See "Population and social conditions" in Eurostat Statistics in focus — 1/2010

[26]    Persons have a foreign background if they themselves or at least one of their parents were born abroad.

[27]    See Council Declaration on the European Year for Active Ageing and Solidarity

between Generations (2012): The Way Forward

[28]    See also G. Lanzieri, ‘Fewer, older and multicultural? A projection of the populations of the European Union Member States by foreign/national background’. Paper for the European Population Conference, Vienna, 1-4 September 2010

[29]    Eurostat, demo_gind

[30]    Eurostat, lfsa_egan

[31]    See European Commission (2010), Demography report "Older, more numerous and diverse Europeans" chapter I.5 and II.2.

[32]    See OECD (2012) International Migration Outlook, page 127.

[33]    See Hablicsek, L (2004)  'Demographics of population ageing in Hungary' Discussion Paper No.207  Project on Intergenerational Equity, Institute of Economic Research, Hitotsubashi University

[34]    Vlachantoni, A. (2010), The demographic characteristics and economic activity patterns of carers over 50: evidence from the English Longitudinal Study of Ageing, Population Trends Nr. 141, pp. 54-96.

[35]    See the special Eurobarometer EBs337  Geographical and labour market mobility published in June 2010 and EBs346 New Europeans published in April 2011 at

[36]    See European Commission (2010) Demography report

[37]    See

[38]    Eurostat, Survey on ICT usage in households and by individuals

[39]    Council of European Professional Informatics Societies (2007) 'Thinking Ahead on e-Skills for the ICT Industry in Europe'

[40]    European Commission (2012)  Digital Agenda Scoreboard, Chapter 1: A vibrant digital single market

[41]    JRC (2013) Information and Communication Technologies (ICTs) for disadvantaged youth: Opportunities and Challenges. Evidence from literature and practice. Policy Report. Authors: C Centeno, J Cullen, S Kluzer, A Hache (forthcoming 2013), JRC Scientific and Technical Reports Series, EUR NN NNN EN, at

[42]    JRC (2012) The 2012 Predict Report An Analysis of ICT R&D in the EU and Beyond

[43]    JRC (2013) Literature Review on Employability, Inclusion and ICT, Report 2: ICT and Employability. Authors: de Hoyos M., Green A. E., Barnes S-A., Behle H., Baldauf B., and Owen D. Editors: Centeno C., Stewart J. JRC Technical Report Series EUR NN NNN EN, Institute for Prospective Technological Studies, Joint Research Centre, European Commission (forthcoming 2013), at

[44]    Op. cit.

[45]    Report for the European Commission “Anticipating the Evolution of the Supply and Demand of e-Skills in Europe (2010-2015)” Empirica and IDC Europe, December 2009. Updated forecast presented at the European e-Skills Conference held in Brusselson 13 December 2011.

[46]    See European Commission Communication -Towards a Job Rich Recovery, COM(2012) 173 final

[47]    IDC White Paper (2009) Post Crisis: e-Skills Are Needed to Drive Europe's Innovation Society

[48]    For example, one econometric study of the Italian labour market, monitoring for age and education and following individuals over time, found that low educated workers aged between 35 and 49 with no digital skills have a 5% higher risk of being unemployed than those with digital skills; and highly-educated 50-64 workers with no digital skills have a 20% higher risk of being unemployed than those with digital skills. The dataset used provides four different waves of data (2000, 2002, 2004, and 2006). For the illustration of the dataset see Main Report (Codagnone et al 2009), point 8.2.

[49]      JRC (2013) Information and Communication Technologies (ICTs) for disadvantaged youth: Opportunities and Challenges.Evidence from literature and practice. Policy Report. Authors: C Centeno, J Cullen, S Kluzer, A Hache (forthcoming 2013), JRC Scientific and Technical Reports Series, EUR NN NNN EN, at:

[50]    Unit labour cost (ULC) growth = growth compensation per employee adjusted for productivity growth Note: IE: data missing, LV: structural break in data

[51]    'Automatic stabilisers' refers to mechanisms that stabilise economic growth, notably through the stabilisation of households' incomes (through higher social benefits such as unemployment benefits, and lower taxation in the event of negative shock) so as to cushion consumption and savings and thus the overall economic cycle.

[52]    Data available only until 2010

[53]    See European Commission (2011) Employment and social developments in Europe 2011, chapters 1 and 2

[54]    See European Commission (2011)  Employment and social developments in Europe 2011, chapter 1

[55]    European Commission (2011) Employment and social developments in Europe 2011, chapter 4

[56]    Consumer surveys carried out under the joint harmonised EU programme of business and consumer surveys

[57]    Time series smoothed by applying Hodrick-Prescott filter

[58]    "Material deprivation" covers indicators relating to economic strain, durables, housing and environment of the dwelling.

[59]    Eurostat, EU-SILC, ilc_mddd11

[60]    Note: Jobless households (JLH) refers to people living in households with very low work intensity (LWI) by age and sex (population aged 0 to 59 years). In this document both the abbreviations JLH and LWI are used to cover the same indicator.  

[61]    The term 'NEET' refers to young people who were not in employment, education or training

[62]    Economic Recession and Health Outcomes, Research Note, Sotiris Vandoros, Mauricio Avendano-Pabon, Philipp Hessel and Tiziana Leone (to be checked with D4 if already published)

[63]    The barriers considered are "too expensive, "too long waiting lists" or "too far to travel"

[64]    Special Eurobarometers "Social Climate" EBs 315 June 2009, EBs349 June 2010, EBs370 June 2011, EBs391 June 2012, at:

[65]    This indicator refers to people who declare that they did not seek health care even if they needed it because it was too expensive, the waiting list was too long or because it was too far to travel (reasons linked to the organisation of the health care system).

[66]    Reference periods for income and activity status for IE and UK  differ from the other countries (where the data refers to the previous year).

[67]    All data refer to EU-27.

[68]    Figures refer to cut-off point: 60% of median equivalised income, white bars indicate declines in the poverty gap and red ones increases.

[69]    Data for IE, FR are from 2007 (latest available)

[70]    Special Eurobarometer "Poverty and Social Exclusion" EBs355, September 2010, see:

[71]    This section draws on a review of existing data collections by the SSO ( and a special focus from the European Commission's June 2012 Employment and social situation quarterly review

[72]    European Commission Staff  Working Document – ' Investing in Health' SWD(2013) 43

[73]    It can be noted that such an approach remains however partial. From an economic point of view, it notably disregards the multiplier impact of additional private consumption or additional public investment, as well as the impact of possible overall lower taxes or public debt. It also doesn't reflect the potential change in the take-up of some in-kind benefits – such as child care or health – and neither changes in the payments of employers’ contributions or indirect taxes such as VAT.

[74]    This type of approach may also be used for other types of social protection and social inclusion policies, for instance, providing estimates on the return of child care facilities, or on changes in marginal tax rates for different types of situation (for instance for pension take-up or second earners).

[75]    The size of the population poor and unemployed population at the time of the survey (2010) and of the population  unemployed for at least one month over the reference income year (generally 2009) ranged around 1% (or above) of the overall population in BG, CZ, DE, EE, IE, EL, ES, LV, LT, HU, PT, SI and SK. It corresponded to more than 10% of the overall poor population in CZ, LV and LT. Estimates of the gains are obtained by comparing average labour market incomes, benefit levels (excluding pensions) and (direct) taxes across 24 categories (accounting for sex, education in three levels and household composition in four levels  — single with or without children, two or more adults with or without children) of poor unemployed and non-poor employed people (with overall incomes between the poverty threshold and the median income) and are grouped together in three categories : higher household disposable incomes, lower benefits and higher taxes. In one country (LT), non-poor employed people show on average slightly higher average benefits levels than poor unemployed people.

[76]    Ciccone, A. and de la Fuenta, A. (2002) 'Human capital and a knowledge-based economy' at:

[77]    Prince's Trust (2010) The cost of exclusion: counting the cost of youth disadvantage in the UK.

[78]    FEANTSA (2008)

[79]    See, for instance, Sparkes, J. and Glennester, H. (2002) “Preventing Social Exclusion; Education’s Contribution”

[80]    This indicator is the headline indicator for monitoring the Europe 2020 strategy poverty target. It reflects the share of the population living in a household which is either at risk of poverty, or severely materially deprived, or with very low work intensity.

[81]    This is due in particular to the lack of comparable longitudinal data.

[82]    OECD (2009) Doing Better for Children, 2009.

[83]    UK National Equality Panel (2010) An Anatomy of Economic Inequality in the UK

[84]    Studied compared students from the top and bottom HISEI quintiles.  Jerrim, J ' The Socio-Economic Gradient in Teenagers’ Reading Skills: How Does England Compare with Other Countries?' Institute of Fiscal Studies, Vol. 33, Is. 2

[85]    Poulton, R., Caspi, A., Milne, B.J., et al. (2002), 'Association between children's experience of socioeconomic disadvantage and adult health: a life-course study', Lancet, Vol. 360, pp. 1640-1645

[86]    The height of each bar measures the extent to which sons’ earnings levels reflect those of their fathers. The estimates are the best point estimate of the intergenerational earnings elasticity resulting from an extensive meta-analysis carried out by Corak (2006) and supplemented with additional countries from d’Addio (2007). The choice of empirical estimates in this meta-analysis is motivated by the fact that they are based on studies that are similar in their estimation technique, sample and variable definitions. The higher the value, the greater the persistence of earnings across generations, thus the lower is the intergenerational earnings mobility.

[87]    Quoted in OECD (2010) Economic Policy Reforms Going for Growth

[88]    2010 values instead of 2011 for CY, IE, IT and UK; EU-27 is based on Eurostat estimate for 2011.

[89]    Social Protection Committee (2012) 'Tackling child poverty and social exclusion, promoting child well-being', Social Protection Committee, see table p. 27

[90]    OECD (2011) Can parents afford to work? An update. The following text largely quotes the results of this study.

[91]    The Barcelona targets will be used as an indicator in the European Semester

[92]    Ghysels, J. and Van Lancker, W. (2011) 'The unequal benefits of activation: an analysis of the social distribution of family policy among families with young children' Journal of European Social Policy, Vol. 21 No. 5 pp. 472-485

[93]    European Commission Staff Working Document — Progress towards the common European objectives in education and training-Indicators and benchmarks-2010/2011, SEC(2011)526

[94]    Social Protection Committee (2012) 'Tackling child poverty and social exclusion, promoting child well-being', see table p. 27

[95]    European Commission Staff Working Document —Education and Training Monitor 2012,  (SWD(2012) 373 final, 20.11.2012

[96]    Eurydice (2009) Gender differences and Educational Outcomes:  Study on the Measures Taken and the Current Situation in Europe

[97]    The dwelling is considered overcrowded if one the criteria mentioned below is not fulfilled: - 1 room for the household; plus one extra room for— each couple; —each single person aged 18+; — for two single people of the same sex between 12 and 17 years of age; — for each single person of different sex between 12 and 17 years of age; —for two people under 12 years of age.

[98]    'Tackling child poverty and social exclusion, promoting child well-being', Social Protection Committee, 2012, see table p. 27

[99]    'Alternative care' refers to care provided to children deprived of parental care, including:

(i)    informal care: any private arrangement provided in a family environment, whereby the child is looked after on an ongoing or indefinite basis by relatives or by others in their individual capacity, at the initiative of the child, his/her parents or other person without this arrangement having been ordered by an administrative or judicial authority or a duly accredited body;

(ii)   formal care: all care provided in a family environment which has been ordered by a competent administrative body or judicial authority, and all care provided in a residential environment, including in private facilities, whether or not as a result of administrative or judicial measures

[100] Eurochild (2010) National surveys on children in alternative care

[101] Recent studies have also confirmed the overrepresentation of children of Roma origin in institutions across several EU countries, 'Romani Children in Institutional Care', European Roma Rights Center and Bulgaria Helsinki Committee, 2011.

[102] In particular inadequate housing, single parenthood, lack of access to welfare, unemployment, lack of access to day-care and specialised services for children with disabilities, children's health condition, stigma and discrimination

[103] 'Guidelines for alternative care of children', UN framework, 2009

[104] K. Browne (2009) 'The Risk of Harm to Young Children in Institutional Care', pp. 9 – 17;  Office of the United Nations High Commissioner for Human Rights (2011) 'Forgotten Europeans, Forgotten Rights – The Human Rights of Persons Placed in Institutions'

[105] Bynner, J. and Parsons, S. (2002) 'Social Exclusion and the Transition from School to Work: The Case of Young People Not in Education, Employment or Training', Journal of Vocational Behavior, 60, pp. 289-309.

[106] Eurofound (2012) 'Young people and NEETS in Europe: First findings', European Foundation for the Improvement of Living and Working Conditions

[107] See:

[108] The European Council approved in May 2012 a benchmark highlighting the importance of the youth transition phase from education to employment. In particular, the benchmark states that “By 2020, the share of employed graduates, 20 to 34 years old having left education and training (at levels ISCED 3-6) no more than three years before the reference year, should be at least 82%”.

[109] The classification of educational levels is based on ISCED: low level of education means ‘at most lower secondary’ (ISCED 0-2), a medium level of education means ‘upper secondary and post secondary (non tertiary)’ (ISCED 3-4), and a high level of education means ‘tertiary education’ (ISCED 5-6).

[110] MT: data not publishable due to the sample size being too small.

[111] An in-depth analysis of the drivers of in-work poverty and the policy instruments that can best be mobilised to tackle it is available in the European Commission Staff Working Document  'Follow-up on the implementation by the Member States of the 2008 European Commission recommendation on active inclusion of people excluded from the labour market – Towards a social investment approach'  SWD(2013) 39

[112] The income reference period is a fixed 12-month period (such as the previous calendar or tax year) for all countries except the United Kingdom for which the income reference period is the current year of the survey and Ireland for which the survey is continuous and income is collected for the 12 months prior to the survey. 2010 values instead of 2011 for IE ; EU27 is based on Eurostat estimate for 2011.

[113] Social Protection Committee (2009) Growth Jobs and Progress in the EU: A contribution to the evaluation of the social dimension of the Lisbon Strategy. See:

[114] European Commission (2012) Employment and social developments in Europe 2012

[115] Whether or not to include housing costs in the definition of income underpinning the risk of poverty rate has sparked much debate in past years and will probably continue to do so in the future. The conclusion of the SPC indicator subgroup was that such costs should not be included. Indeed, imputing rents is a difficult exercise, especially at the European level. Real estate prices are so heterogeneous across geographical zones that they could induce more bias than correcting it.

[116] About a fifth of people aged 65 or older have pension incomes just below or just above the poverty risk threshold, consequently relatively small increases or decreases in their pensions can lead to important variations in the poverty rates of the elderly.

[117] Data for IE, IT, CY, UK refer to 2010

7..... Sizes, compositions and effects of budgets for social policies: social investment, protection, and economic stabilisation. 49

7.1.    The sizes and structure of social budgets. 49

7.1.1      Trends during recent decades. 52

7.2.    Social policies have contributed to economic stabilisation in the crisis. 54

7.2.1      Social protection significantly helped to mitigate the impact of the crisis. 54

7.2.2      Social protection predominantly sustained households’ incomes, though there are signs of weakening. 55

7.3     Adequacy of budgets and the scope for efficiency gains. 57

7.3.1      Early intervention and breaking the transmission of disadvantage. 59

7.3.2      Lifelong learning, training and up-skilling to improve outcomes in adulthood. 62

7.3.3      Fighting poverty and exclusion, the impact of consolidation measures and benefit coverage and take-up. 63

7.3.4      More effective and efficient policies: the case of combating child poverty. 64

7.3.5      The impact of consolidation measures. 66

7.3.6      Coverage and take-up of social benefits. 67

7.3.7      Labour-market friendliness of social protection spending. 71

7.4     The financing of budgets for social policies. 74

7.4.1      Overall shift from social contributions to general taxation. 74

7.4.2      Potential significant distributional impacts. 76

7.4.3      Shifts in social protection financing and coverage of employment and life-course risks  77

8. ... Improving timeliness of data and the measurement of social outcomes  78

8.1 Harmonised EU statistics and indicators in support of the Europe 2020 target to reduce poverty and social exclusion. 79

8.2  Improving timeliness. 82

8.3  Improving the measurements of social outcomes. 85

8.3.1      Non-monetary income components. 85

8.3.2      Improving the measurement of material deprivation. 87

8.3.3      The dynamics of poverty and social exclusion. 88

8.3.4      Capturing the gender dimension of poverty. 88

8.3.5      Capturing the situations of children and youth. 89

8.3.6      Measuring the most extreme forms of poverty and social exclusion. 90

7. Sizes, compositions and effects of budgets for social policies: social investment, protection, and economic stabilisation.

The sizes and compositions of social budgets are very diverse in the EU, reflecting the various national contexts, such as different emphases given to the three functions of investment, protection and stabilisation of social policies. Well-designed adequate and sustainable social policies combining these three functions can indeed lead to efficiency gains and notably better employment and poverty outcomes. In the current crisis, social policies have in general provided strong economic stabilisation, though there are signs of it weakening. There are also significant differences among Member States showing potential for gains in the effectiveness and efficiency of social spending, for instance through investing in early childhood education, the use of activating and enabling policies to tackle poverty and unemployment, and the varying modes of financing of social policies. 

7.1. The sizes and structure of social budgets

Social systems in EU Member States are very diverse. At EU level, social protection expenditure accounts for almost 30% of GDP, though this varies greatly from around 18% in Bulgaria, Estonia, Latvia and Romania, to as much as 32% in Denmark, France and the Netherlands. [118]

While all pre-2004 Member States (except Luxembourg) spend at least 25% of their GDP on social protection, all post-2004 Member States spend less than this.

Figure 7.1 — Social protection expenditures in EU Member States in 2010 (% of GDP)


The make-up of social protection spending also varies greatly. The largest component is old-age pensions, accounting in 2010 for 11 % of EU GDP, or nearly 40 % of EU social protection expenditure. The lowest levels are in Ireland and Luxembourg (only around 6 % of GDP), and the highest in Italy (nearly 15 % of GDP).

Figure 7.2 — Expenditures on social protection benefits by function in 2010 (% of GDP)

Source: ESSPROS.

In the EU as a whole, the second-largest budget item is sickness and healthcare, which accounts for more than 8 % of GDP (or nearly 30 % of social protection expenditure), though it ranges from 4 % of GDP in Latvia, Bulgaria and Romania to 12 % in Ireland and nearly 11 % of GDP in the Netherlands.

Spending on disability amounts to slightly over 2 % of GDP on average in the EU. In Denmark (where the share of social protection expenditure on health care is among the lowest), the share of spending on disability is the highest in the EU (nearly 5 % of GDP), while Ireland (where the share of spending on healthcare is among the highest in the EU) has one of the lowest shares of expenditure on disability (slightly over 1 % of GDP).

Family/child benefits account for a little over 2 % of GDP in the EU on average, ranging from around 1 % of GDP in Poland, Malta and the Netherlands to as high as 4 % in Denmark, Luxembourg and Ireland.

Spending on unemployment benefits is below 2 % of GDP in the EU on average and ranges from under 0.6 % in Poland, Malta and Romania to as much as 4% in Belgium, Spain and Ireland.  

The share of the social investment function also differs among Member States. While the investment function covers different approaches under the same umbrella, as highlighted by Morel, Palier and Palme (2012) or Van Kersbergen and Hemerijck (2012), it can be misleading to allocate different types of social expenditures specifically to one of the three functions of social policies (investment, protection and stabilisation), though some types of expenditures are more directly linked to the investment dimension of social policies.

An estimation of the size of social investment expenditure can, for instance, be made using the methodology proposed by Hemerijck (2012), which combines active labour market policies, childcare, education, research and rehabilitation as a proxy for social investment expenditures. In this approach, the remaining social protection expenditure such as old-age and survivor's benefits, unemployment-related benefits and disability benefits, healthcare and housing are gathered in a second category of remaining social expenditures.

These estimates highlight that social investment policies have been put into practice to a different extent across Member States, with levels of more investment-oriented social expenditures being higher than the EU average (7.5 % of GDP) in some continental Member States (BE, FR, NL and AT) and in IE and CY and reaching more than 10% of GDP in northern Europe (DK, FI and SE), also reflecting higher overall shares of expenditures in GDP. In contrast, they are relatively low in some southern and eastern Member States (BG, RO, CZ, EL and SK) and in LU. The related share of more social investment-oriented social expenditures reaches 21 % in EU on average and is higher than 25 % in northern Member States (DK, FI, SE) and in the Baltic Member States (EE, LV and LT) and CY. It is lower than 20 % in EL, IT and LU.

Figure 7.3 — More social investment oriented vs. remaining social expenditures in 2010 (% of GDP)

Source: Eurostat, DG EMPL calculations.

Note: expenditures are expressed in GDP percentage points (left axis), while the share of more investment-oriented social expenditures is expressed as a percentage (right axis). More investment oriented social policies combine active labour market policies, childcare, education, research and rehabilitation as a proxy for social investment expenditures.

7.1.1    Trends during recent decades

From the mid-1990s until 2007, social protection expenditure in the EU grew in line with rising overall incomes, with the result that its share remained relatively stable. But with the crisis and a sharp drop in GDP, social protection expenditure played its role of cushioning the effects of the crisis, acting as an economic stabiliser. Spending rose significantly as a share of GDP in 2008 and 2009, while starting to decline slightly with the (temporary) recovery of 2010.

Figure 7.4 — Trends in social protection expenditures (1995-2010, % of GDP)

Source: ESSPROS.

On average in Europe, the structure of expenditure by function remained relatively stable (Figure 7.5). In spite of an ageing population, the share of expenditure on old age and survivors remained virtually stable since 1995, while the share of expenditure on family benefits actually increased slightly. The share of expenditure on housing and social exclusion remained very stable throughout. In the period before the crisis, the share spent on healthcare, sickness and disability benefits rose, while during the crisis unemployment benefits, logically enough, accounted for a growing share of spending.

Figure 7.5 — Trends in social protection structure by function 1995-2010 (as a % of total expenditures)


7.2. Social policies have contributed to economic stabilisation in the crisis

Social expenditure is a powerful stabiliser of economic activity as it helps to sustain effective demand during slowdowns and more particularly recessions, primarily through sustaining household incomes (through higher benefits as a response to a decline in wages, and via lower taxes), but also thanks to a lower need for increases in precautionary savings during economic slowdowns.

The cyclicality both of total social protection expenditure and of different types of benefits varies significantly across Member States. Unemployment benefits respond strongly to the cycle, followed by social exclusion, housing and family benefits. Pensions are generally considerably less anti-cyclical, as are sickness and disability benefits.[119]

7.2.1    Social protection significantly helped to mitigate the impact of the crisis

There is strong evidence of the significant role that social spending played in sustaining gross household disposable income during the 2008-2009 recession in most EU countries.[120] However, while social protection played an important role in mitigating the effects of the crisis, this role came up against clear limits.

In the first phase of the crisis, social protection played an important role in sustaining household incomes. In the eurozone, net social benefits and reduced taxes contributed positively to the change in gross household disposable income (GHDI) during 2009 and in the first two quarters of 2010 (Figure 7.6).

However, in the second phase, this effect started weakening. At the end of 2010 and during 2011, the contribution of social benefits to the change in gross household income lessened and started being negative, while in the first quarter of 2012 it was positive again. This may have occurred because of the phasing-out of entitlements, along with some improvement in the economic outlook in some Member States, as well as because of fiscal consolidation measures that reduced the level or duration of benefits, or because eligibility rules excluded some beneficiaries from some schemes.[121]

Figure 7.6 — Contributions of components to the growth of nominal gross disposable income of households (eurozone)


Source: Eurostat/ECB

7.2.2    Social protection predominantly sustained households’ incomes, though there are signs of weakening

Since the beginning of the crisis, social protection expenditure has gone through two distinct phases, 2007-2009 and 2009-2011, as has gross household disposable income (GHDI, see Figure 7.7). In 2007-2009, cash expenditure increased in all countries covered except Hungary, while in-kind expenditures increased in all except Latvia, Hungary and Bulgaria. Despite this, GHDI decreased in 10 of the 26 Member States for which data are available. The largest declines were observed in Hungary, Estonia and Latvia (between 6 % and 15 %). Social expenditure was able to cushion the fall in incomes in almost all Member States.

There are significant differences between countries that experienced similar GDP shocks, not only as regards the scale of change in social protection expenditure, but also in the change in GHDI. For instance, while Italy and Finland had similar GDP shocks and increased their cash social protection benefits in a similar proportion, GHDI increased in Finland while it decreased in Italy. In the Czech Republic, spending on cash benefits increased slightly less than in the Netherlands, but only in the Czech Republic did GHDI rise.

Figure 7.7 —  Change in real GHDI and real social protection benefits, 2007–2009 and 2009–2011 (%)[122]

Source: ESSPROS and National Accounts

Note: Member States are grouped according to the size of the GDP shock in the initial phase of the crisis 2007-2009.

Social protection has been more effective at sustaining households’ income during the crisis than overall tax systems (Figure 7.8). Between 2007 and 2009, the positive effect of changes in social transfers on GHDI was three times stronger than that of taxes on average in the EU, but between 2009 and 2011, the effects of both were close to zero.

Figure 7.8 — Impact of social transfers and taxes on GHDI in 2007–2011[123]

Source: National Accounts, DG EMPL calculations.

Overall in the EU, social protection expenditure was generally rising until 2009, when retrenchments started, especially in countries most in need of fiscal consolidation. Benefits provided in kind (chiefly services) were affected the most. Cutting or freezing the level of services in areas such as healthcare, training, housing or childcare is likely to have a long-term detrimental effect on the employability of workers and on their capacity to participate fully in society. The more fiscally resilient Member States have been better able to preserve such services. This tends to accentuate the large variations that already exist in the effectiveness of social protection systems across the EU. Nevertheless, further research would be needed to better assess the effectiveness of more investment- oriented social expenditures in weathering the effects of the crisis.

7.3       Adequacy of budgets and the scope for efficiency gains

The size, structure and design of social policies all matter for the performance of welfare systems. Well-designed, adequate and sustainable social policies combining the three main functions of social protection are efficient in reaching social and economic outcomes, and can indeed lead to lower poverty and exclusion and better labour market outcomes.[124]

There are nevertheless significant differences among Member States as regards the effectiveness and efficiency of spending on social policies, both in terms of both poverty reduction and labour-market friendliness. This suggests that there is room for efficiency gains. Long-term trends suggest that the countries with the highest welfare spending are not those with the highest public debt.[125]

Building on recent academic work by Anton Hemerijck, it seems that countries with higher budgets for policies that focus more on social investment[126] and indeed higher total budgets for social policies, are associated with better outcomes in terms of poverty and of employment (Figures 7.9 and 7.10). Countries with higher budgets for more social-investment oriented policies fare better in terms of employment and monetary poverty, suggesting that more investment-oriented social policies may be particularly efficient in raising employment levels and reducing poverty levels.

Figure 7.9 — Social investment and employment rates (2010)

SoSource: Eurostat, DG EMPL calculations.

Figure 7.10 — Social investment and at-risk-of-poverty (AROP) rates (2010)


Source: Eurostat, DG EMPL calculations.

Note: at-risk-of-poverty rates refer to the EU-SILC 2011 wave (except for IE 2010), in general reflecting the income situation in 2010.

Indeed, simple regressions suggest that for every additional 1 % of GDP spent on more investment-oriented social policies (as calculated according to the above), the employment rate is around 1.7 point higher, while the link with total social protection expenditures is weaker (0.5 point higher). Conversely, 1 % of GDP spent on more investment-oriented social protection expenditures is associated with an almost 0.6 point lower at-risk-of-poverty rate, while the link to total social expenditures is around 0.2 point.

These very stylised elements suggest that an approach that integrates both social investment and other social protection is needed. For instance, social investment can be particularly effective in improving employability. This in turn creates the prerequisites for further economic and employment growth which will give room for better social policies, therefore reducing risks of poverty in the longer term. Conversely, remaining social expenditures can support people effectively, especially during a crisis.

Further work in terms of analysing the impact of social investment and in particular its cumulative impact on employment and the at-risk-of poverty rate would be useful. This is because social investments are particularly valuable when they are ongoing and consistent. Social investment leads to a gradual accumulation of human capital in terms of literacy and skill acquisition. This in turn leads to high-quality employment being created in growing economic sectors.[127]

7.3.1    Early intervention and breaking the transmission of disadvantage

Intergenerational disadvantage can be explained by a range of factors. As discussed earlier, children in low-income backgrounds earn lower incomes later on in life. Environmental and cultural factors also play important roles in outcomes for children. Most of the factors influencing intergenerational disadvantage are linked to the main actors responsible for investing in children’s upbringing, namely parents, whose socio-economic status has a strong influence on their capacity and resources to invest in their children, governments and other social institutions.

Intergenerational mobility appears to be highly influenced not just by the degree of investment in children, but also by the rate of return on this investment, which is highest in the very early years of childhood. Recent developments in neuroscience have further emphasised the determining influence of investment in pre-school years (especially before the age of three), during which most of the essential cognitive and social skills are formed. These years appear to be those with the highest rate of return on education achievement and overall human capital investment in children, especially through health and education intervention. Benefits for children from disadvantaged backgrounds are even more marked.

Results from the OECD's PISA assessment of students at age 15 show that, for most countries, students who have attended pre-primary education do better than those who have not. This strongly suggests that early education can improve education outcomes and overall skill levels later on in life. This has profound consequences in terms of human capital stock and overall labour force competitiveness.

Figure 7.11 —  Influence of pre-primary education policies on PISA results, 2009

Source: OECD (2012)[128]

There is, however, a large divergence between some EU countries in terms of investment and participation in pre-primary education. For instance, enrolment in education at the age of four is 100 % in France, while only slightly over 50 % in Greece. While there has been a trend among Member States and OECD countries in general to increase enrolment in early education between 2005 and 2010 (such as FI, DK, DE, SI, EE, AT, and PL) there has been a worrying decline in some Member States (EL, CZ, IT).

Figure 7.12 — Enrolment rates at the age of four in education (2005 and 2010)

Source: OECD (2012)[129]

Better investment in children can also contribute to significant savings in the longer term, since the public expenditure needed to correct the consequences of childhood poverty throughout a person's life-span is significantly higher than that necessary to improve their life chances by support provided during childhood.

Figure 7.13 — Return to a euro invested in human capital at different ages

Source: The case for investing in disadvantaged young children, James J. Heckman.

7.3.2    Lifelong learning, training and up-skilling to improve outcomes in adulthood

As shown in Section 6.3, education has a major influence on risk of unemployment.  Apart from initial educational attainment, training, lifelong learning, up-skilling and training measures (either while employed or unemployed/inactive) can boost positive transitions on the labour market. Measures to improve employability are particularly relevant in times of high unemployment, when people may lose jobs in declining occupations and need to be trained for new occupations. This is seen to apply specifically to older workers. In the case of the Netherlands, Van der Heul (2006) found that the effectiveness of training increased for older workers at a time of high unemployment. At the time of the study, 2003, the majority of the unemployed not only found a new job, but even a new job in a different sector. However, Ecorys and IZA (2012) have found that re-training needs to be accompanied by job search assistance in order to be effective. The average adult participation in lifelong learning in the EU is 8.9 %.[130]

Transition rates from short- and long-term unemployment can be analysed separately, depending on whether or not the unemployed person has had access to lifelong learning in the previous year.[131] The results suggest that participation in lifelong learning can increase the frequency of positive transitions (from unemployment or inactivity to employment) and reduce the frequency of negative transitions (staying in unemployment or in inactivity).

In particular, the transition rate out of unemployment to employment is 6 points higher for those having had some lifelong learning opportunities (37 % vs. 31 %), as also mirrored in a lower persistence rate in unemployment (44 % vs. 49 %).

Figure 7.14 — Transition rate to employment for unemployed and inactive persons, depending on participation in lifelong learning, 2010-2011 (EU-13)

Source: Eurostat, EU-LFS, ad-hoc transitions calculations

7.3.3    Fighting poverty and exclusion, the impact of consolidation measures and benefit coverage and take-up

The design of the tax-benefit system is crucial in determining how and to what extent it affects income inequalities and redistributes resources to the poor. Within this, two key factors stand out: the progressivity of taxes and the degree of targeting and conditionality of benefits (which can create disincentive effects if badly designed), while a number of other factors are also at stake (such as for instance the choice of various tax bases and the existence of various tax exemptions).

Social transfers other than pensions reduce poverty risks to various degrees across Member States (ranging from a poverty reduction effect of 50 % or more in some countries to 25 % or less in others). This largely reflects differences in total expenditure, which vary from 7.5 % to 20.5 % of GDP when pensions are excluded (see Figure 7.15), but the composition of expenditure and the quality of interventions also play an important role. The evidence shows there is much variation across Member States in net cash support to low-income households.

Figure 7.15 — Relationship between social protection spending (excluding pensions, relative to GDP) and relative reduction in the share of population (aged 0-64) at risk of poverty (as a percentage) (2010)

Source: ESSPROS and EU-SILC.

Means-testing of social benefits can reduce social spending while more effectively protecting those most vulnerable. However, attention should be paid to potential work disincentives, low benefit take-up and stigmatisation associated with targeting if badly designed.

A recent Euromod paper[132] illustrates the importance of well-designed child and family benefits. It explores the extent to which a country’s effectiveness in reducing child poverty can be attributed to the size of family cash transfers (i.e., benefits and tax instruments alike) or to their design. The results confirm that the impact of the level of expenditure is significant. Nevertheless, effectiveness is highly dependent on the composition of the measures (universal, categorical, income selective) and the design of policies (thresholds, benefit size determination, etc.). The balance between benefits in cash and in kind also matters. Subsidies for childcare reduce the risk of poverty among children, make the overall income distribution less unequal, and are fiscally progressive. These effects are reinforced if a more dynamic perspective is adopted: subsidising childcare helps improve human capital and achieve higher female employment, both leading to greater prosperity and a more equitable income distribution.

7.3.4    More effective and efficient policies: the case of combating child poverty

As illustrated in section 6.1, fighting child poverty requires action on several fronts through policies that need to be mutually supportive to ensure effective and efficient intervention. Figure 7.16 illustrates that point. It provides a summary of the main drivers of child poverty prevailing in different countries.

Drivers are identified through three indicators, reflecting the exclusion of parents from the labour market (children in jobless households), in-work poverty (parents work but do not earn an adequate living), and the effectiveness of welfare support.

Depending on how these factors interact, countries can be grouped according to four major profiles associated with different combinations of intervention, leading to very different outcomes on child poverty. It shows that countries that combine adequate family support with measures to help parents find jobs have the best outcomes.

Group A includes the Nordic countries (Denmark, Norway and Sweden), Austria, Slovenia and Cyprus. Both the rate of risk-of-poverty and the child poverty gap are lower than in the rest of the EU. This can be attributed to sound performance on all fronts: the low proportion of children living in jobless or working poor households, and the relatively high impact of social transfers in reducing child poverty.

Nordic countries achieve these goals despite a high proportion of children living in single-parent households, thanks to good provision of childcare and a wide range of measures to help parents balance working life and family life. The impact of social transfers is relatively low in Cyprus, but strong family structures in which most adults are at work have so far played a protective role.

Group B includes Belgium, the Czech Republic, Germany, France, the Netherlands and to a lesser extent Lithuania and Estonia. These countries achieve relatively good to below-average poverty outcomes. The main matter of concern in these countries is the relatively high number of children living in jobless households, whereas children whose parents are working have very low risks of poverty. Among these countries, Germany and France are limiting the risk of poverty for children through relatively high and effective social transfers.

Group C includes Hungary, Ireland and the UK. The main concern in these countries is the very high number of children living in jobless households. In these countries, social transfers have a strong impact on reducing child poverty, which ensures a relatively low risk of child poverty in jobless households. However, analysis shows that the design of transfers, compounded by the lack of adequate, affordable childcare, create disincentives to work for specific family types, such as single parents, who account for more than half of all jobless households.

Group D comprises of southern European Member States (Greece, Italy, Malta, Portugal and Spain) as well as most of the eastern and Baltic countries (Bulgaria, Latvia, Romania, Poland and Slovakia). These countries face a high risk of child poverty and a high relative poverty gap for children. The in-work poverty risk among families is high. Important factors seem to be: insufficient work intensity and low earnings (in Latvia, Lithuania, Poland, Portugal and Spain). In these countries, the level and effectiveness of social spending are among the lowest in the EU. Family structures and intergenerational solidarity play a role in alleviating the risk of poverty for the most vulnerable children. Living in multi-generational households and/or relying on inter-household transfers, whether in cash or in kind, may partly compensate for the lack of government support to parents in the most vulnerable situations.

Figure 7.16 — Relative outcomes of countries related to the main determinants of monetary child poverty[133]

Drivers… || || …level of child poverty || Countries || Tentative diagnosis

|| Impact of social transfers is high Low share of children in jobless households Low risk of poverty of children whose parents are working || || Low risk of child poverty || DK AT SI FI SE (CY) || Lowest rates of child poverty thanks to a good balance between income support, labour market conditions and services that facilitate labour market participation of both parents.

|| Impact of social transfers is relatively high Relatively high share of children in jobless households Low risk of poverty of children whose parents are working || || Medium risk of child poverty || CZ  NL BE DE FR   (LT EE) || Low to above average rates of child poverty thanks to a good income support, but the share of children living in jobless households is high.

|| Impact of social transfers is high High share of children in jobless households Relatively lower risk of poverty Average level of in-work poverty || || High risk of child poverty (low poverty gap) || IE UK HU || Average child poverty rates. The high impact of social transfers is mitigated by disincentives to work and lack of adequate and affordable child care for some categories of parents (e.g. lone parents)

|| Low impact of social transfers in reducing child poverty. Limited share of children in jobless households Very high risk of poverty of children whose parents are working || || High risk of child poverty (high poverty gap) ||  PL LV RO BG SK PT IT EL ES MT  || Highest rates of child poverty due to insufficient support for families, both in and out of work, in terms of income and services and poor access to quality jobs, especially for second earners.

Source: Eurostat EU-SILC 2010, European Commission (DG EMPL) calculation.

7.3.5    The impact of consolidation measures

In the current context of fiscal consolidation in a number of Member States, the design of consolidation measures can also have an impact on the ability of social systems to deliver adequate, effective and efficient policies.

Evidence based on micro-simulation provides insight into the likely impact of fiscal consolidation measures on the relative situation of the poorest segments of the population. Euromod has recently reviewed consolidation measures taken in nine EU Member States between 2009 and 2012,[134] showing that the impacts on low-income groups were very diverse reflecting changes in personal taxes and VAT as well as cuts in spending on cash benefits and declines in public sector wages.

Many countries (Estonia, Greece, Spain, Latvia, Portugal and the UK) raised income taxes or social contributions. Many also increased VAT (Estonia, Greece, Spain, Lithuania, Latvia, Portugal, Romania and the UK). In terms of taxes, Greece also introduced an emergency property tax. Other measures remained limited to a few Member States only: minimum wages were cut in Greece; housing benefits were cut in the UK; care benefits were cut in Spain and limited in the UK; Lithuania and Latvia introduced cuts in maternity/paternity benefits. Lithuania also lowered social assistance benefits for those who are able to work, and reduced sickness benefits.

In a few countries, the situation of those on low incomes seems to have worsened more as a consequence of consolidation measures than some other segments (Lithuania, Estonia, Portugal). In other countries, fiscal consolidation measures had a more progressive impact (Spain, Romania, Greece, Latvia). Overall, consolidation measures had a significant negative impact on household income, and in some of these countries, the poorest segments of the population saw their incomes cut significantly (by more than 5 % in Greece and Latvia).

The impact of consolidation measures can also differ for various categories of the population, especially for older people and children. Pensioners were more adversely affected in Greece, Latvia, Portugal and Romania. Unemployment benefits were reduced in Greece, Lithuania, Portugal and Romania. Child benefits were reduced in Estonia, Spain, Lithuania, Latvia, Portugal, Romania and the UK.

7.3.6    Coverage and take-up of social benefits

The degree to which Member State social systems provide effective protection to people in need varies greatly, in terms of the coverage, adequacy and duration of unemployment benefits, and other benefits. Worryingly, administrative data show that in several Member States, a growing number of people are no longer covered by benefits. This raises concerns as to the risk of crime and turning to the informal labour market to earn income.

There are very large differences across EU Member States in terms of coverage. Taking the example of unemployment benefits, the (pseudo) coverage rates of unemployment insurance can be assessed by comparing different sources, supplemented by information on the net replacement rates of unemployment benefits. In most Member States, different sources give relatively consistent results for coverage rates, though in some, such as Italy, Greece and Belgium, they vary significantly (with, for instance, differences ranging from 80 to100 pps in coverage rates).[135]

In some Member States, such as Belgium, Austria or France, both coverage and replacement rates are relatively high, while in others, such as Germany or Finland coverage is high, but replacement rates are lower (around 45 %). In southern Member States, coverage of unemployment benefits is rather low (especially according to surveys) and replacement rates are above 50 % only in Portugal, 41 % in Spain and only 24 % in both Italy and Greece.

The Baltic States have both low coverage and low replacement rates. Those of Bulgaria and Romania are slightly higher. Coverage rates in the Czech Republic and Hungary are similar to those of Portugal and Spain, but replacement rates much lower. On the other hand, Slovakia and Poland have both very low coverage and very low replacement rates. Sweden, which has a universal basic income, has a similar coverage rate, but higher replacement rates.

Figure 7.17 - Unemployment insurance pseudo-coverage rates and net replacement rates (2009)  

|| Coverage rate - administrative sources (in %) || Coverage rate - SILC (in %) || Coverage rate – LFS (in %) || Net replacement rate (in %)

Austria || 88 || 74 || 50 || 52

Belgium || 148 || 89 || 67 || 65

Bulgaria || 49 || 20 || 12 || 25

Cyprus || 54 || n.a. || 26 || n.a.

Czech Republic || 54 || 50 || 30 || 20

Denmark || 55 || 85 || 51 || 63

Estonia || 33 || 46 || 36 || 26

Finland || 94 || 89 || 59 || 44

France || 82 || 69 || 40 || 49

Germany || 110 || 85 || 75 || 45

Greece || 115 || 30 || 22 || 24

Hungary || 36 || 59 || 44 || 22

Ireland || 62 || || n.a. || 58

Italy || 103 || 36 || 6 || 24

Latvia || 35 || 41 || 23 || 24

Lithuania || 31 || 18 || 27 || 21

Luxembourg || 33 || 52 || 31 || 29

Malta || 46 || 41 || 25 || 46

Netherlands || 67 || 55 || n.a. || 38

Poland || 27 || 24 || 15 || 22

Portugal || n.a. || 43 || 41 || 55

Romania || 45 || 29 || 15 || 25

Slovakia || 16 || 30 || 10 || 21

Slovenia || 40 || 31 || 34 || 24

Spain || 39 || 57 || 40 || 41

Sweden || 40 || 37 || 31 || 43

United Kingdom || 62 || 33 || n.a. || 29

Source: coverage rates from EC/OECD database on benefit recipients, SILC and LFS; replacement rates from OECD Tax-Benefit Models[136]

The duration of unemployment benefits also varies greatly across the EU. In Malta and Cyprus, the maximum duration of these benefits was five months in 2011. In the Netherlands, it was over 35 months.

Figure 7.18 - Minimum/Maximum duration of unemployment benefits (2011)[137]

Source: MISSOC

Information on the relative size of incomes of people living on social assistance (including cash housing benefits) makes it possible to broaden the scope beyond unemployment benefits only. To compare the income of such households, including these benefits, with the median equivalised income for three household types, see Figure 7.19. In the EU, single parents with two children are on average getting higher social assistance benefits relative to the median income than single people without children or couples with two children (49 % versus 42 % for the latter two household types).

Only in three Member States (Ireland, Denmark and the United Kingdom) do all three model household types receive social protection benefits high enough to take them above the poverty threshold of 60 % of median income. At the other end of the spectrum, there are Member States in which households do not even reach 40 % of the median income (Greece and Spain; Bulgaria and Romania; Slovakia, Hungary and Estonia).

Of the old Member States, France has the lowest relative net income of people living on social assistance (on average for the three family types, it is 40 % of the median income) and among the new Member States, Lithuania has the highest income (53 % of the median income on average).

Figure 7.19 — Net income of people living on social assistance relative to the median equivalised income, in % (including cash housing assistance) (2010)

|| Single person || Lone parent with 2 children || Couple with 2 children

Greece || 0 || 9 || 2

Bulgaria || 14 || 26 || 22

Romania || 17 || 27 || 26

Slovak Republic || 23 || 33 || 30

Spain || 35 || 33 || 25

Hungary || 31 || 39 || 29

Estonia || 32 || 34 || 31

Poland || 31 || 41 || 33

Portugal || 26 || 42 || 43

France || 41 || 42 || 36

Slovenia || 30 || 53 || 44

Czech Republic || 50 || 45 || 42

Belgium || 45 || 55 || 39

Latvia || 36 || 50 || 46

Sweden || 56 || 48 || 43

Austria || 49 || 50 || 46

Malta || 59 || 54 || 44

Luxembourg || 51 || 54 || 51

Lithuania || 27 || 77 || 56

Germany || 47 || 60 || 53

Finland || 57 || 56 || 50

Netherlands || 74 || 64 || 52

United Kingdom || 66 || 76 || 65

Denmark || 71 || 73 || 71

Ireland || 77 || 71 || 70

Source: OECD Tax-Benefit model. Note: countries are sorted based on the average net income of three family types.

Matsaganis et al. (2008) explore the effect of non-take-up of benefits on the effectiveness of spending on social assistance in terms of poverty reduction, using Euromod for five Member States. They conclude that the design of the targeting of these benefits can have a significant negative effect both on reducing the proportion of the population at risk of poverty and on combating the poverty gap.

7.3.7    Labour-market friendliness of social protection spending

The labour-market friendliness of social systems is a key aspect of their effectiveness and efficiency. Social systems should shield people against labour market risks, while encouraging them to stay in jobs or go back to work. Active labour market policies are shown to have a positive influence on employment rates. The same holds for childcare services and the employment rate of women.

As a complement to income transfers and unemployment benefits, activation policies and ‘making work pay’ tend to improve a person’s chances of getting a job. A core element is the implementation of active labour market policies (ALMPs), in which, for example, the jobless are provided with education and training, as well as active ageing policies, where older workers are encouraged to stay working for longer, and provided with opportunities to do so. ALMPs combine social and economic policies as a means to achieve improvements. But the effectiveness of measures does vary and there are complementarities between active and other labour market policies.

Extensive literature is available on the effectiveness of ALMPs (see for instance European Commission 2006)[138]. It is commonly recognised that such policies facilitate a return to work, minimise long-term unemployment and decrease the loss of productive human capital. During the years 2000-2010, a number of Member States (Denmark, Finland, Sweden, Germany and the Netherlands, for example) engaged in reforms to modernise the welfare state. They aimed to reabsorb high levels of long-term unemployment and/or the swelling ranks of those on long-term illness or disability benefits because of the 1990s recession. A recent study found that countries that invested heavily in ALMP before the crisis saw their employment levels less severely impacted during the crisis.[139]

Spending on and participation in ALMPs tends to decrease long-term unemployment (Figure 7.20).[140] There is broad evidence that spending on or participating in ALMPs decreases the duration of unemployment after taking into account the economic cycle (e.g. Nickel and Layard, 1999). For instance, the Nordic and continental countries tended in 2009 and 2010 to have the highest levels of expenditure on ALMPs, coupled with the lowest persistence rates in unemployment, while central and eastern Member States, and some southern European countries such as Italy and Greece, spend little on ALMPs and generally have a high  persistent rate of unemployment.

Figure 7.20 — Persistence of unemployment and ALMPs expenditures[141]

Source: For persistence rate, DG EMPL calculations based on Eurostat, EU-LFS; for expenditures on ALMPs, Eurostat LMP database.

Both active labour market measures and other measures such as income support can play an important role in helping people to get back to work. For instance, it is possible to assess the degree to which registering with the national Public Employment Service and receiving unemployment benefits influences outcomes in finding jobs for the long-term unemployed.

Figure 7.21 — Transitions from unemployment, depending on registration with the PES and receipt of unemployment benefits for EU-9 (2010 to 2011)[142]

Source: Eurostat, EU-LFS, transitions calculations from DG EMPL.

Evidence from longitudinal data from the EU-LFS shows that those who are registered and receiving benefits have a higher chance of returning to employment than unemployed people not receiving benefits, who may or may not be registered with the Public Employment Services (PES). Econometric analysis presented in the paper confirms that, all things being equal, receiving benefits does influence the likelihood of getting a job, and that registering with the PES alone is not sufficient.[143]However, though registering with the PES is often a precondition for receiving benefits, it does not automatically ensure that a person will have access to services or programmes such as training that may be available.

Childcare provision is a key factor in enabling female employment and fostering labour market participation (see section 5). There is a strong correlation between the employment rates of women with young children and the proportion who have access to formal childcare, especially in the first three years of a child’s life (Figure 7.22), while the correlation is weaker for children aged between three and six years old.

Figure 7.22 — Employment rates of women 20-49 with youngest child below six years old and share of children in formal childcare (2010)

Source: EU-SILC and Labour Force Survey, DG EMPL calculations. Note : For a child to be considered as being in formal childcare, at least one hour per week of formal childcare is required.

7.4       The financing of budgets for social policies 7.4.1    Overall shift from social contributions to general taxation

The relative importance of general government taxes, social contributions and other revenue for financing social protection varies greatly among Member States. Denmark and Ireland finance more than 60 % of theirs from general government contributions,  while in Estonia or the Czech Republic, over 70 % of funding comes from employers’ social security contributions (Figure 7.23).

Figure 7.23 — Social protection financing structures (2010)[144]



In recent decades, the source of such financing in the EU has been shifting from social contributions towards government contributions (Figure 7.23). This trend was very apparent until the current crisis.[145] Since 2008, both social contributions and government contributions have risen significantly as a share of GDP (reflecting the sharp decline in GDP). Government contributions to social protection systems have been slightly more dynamic than social contributions both in the years of the crisis (2009 and 2008) and in 2010, which was generally a year of recovery. In short, government contributions have had a bigger role in financing social protection expenditure over the last 15 years as an overall trend, and during the crisis, though to a lesser extent.

Figure 7.24 — Trends in social protection financing structures (1995-2010)

Source: ESSPROS.

Note: Receipts from the Esspros category of 'other receipts' have been added to the ones from the category of government contributions.

These general developments went along with slow convergence in the financing structures among Member States over the last two decades (Figure 7.24). Member States with relatively high government contributions as a share of GDP financing social protection generally saw a decline (DK, SE or FI). Those with lower levels generally saw an increase (FR, RO, IT, PT, HU or DE). The same type of slow convergence can also be observed with social contributions.

Figure 7.25 — Trends in financing of social protection as a share of GDP (1995-2010)[146]

Government contributions                                             Social contributions


7.4.2    Potential significant distributional impacts

Social security contributions can be reduced by increasing income tax rates, by taxing income other than labour, or by increasing indirect taxation. For instance, in 2007, Germany increased VAT by 3 percentage points to finance a reduction in social security contributions (and a reduction in budget deficits). Some extra VAT revenues were earmarked for social protection. In France, the introduction of the CSG in the 1990s enabled taxation of capital income and replacement income to help finance social protection.

The specific case of a revenue-neutral shift from social contributions to VAT may have adverse redistributive effects for the lower income deciles and a favourable effect on the top deciles, since the former benefit less from a reduction in social security contributions and are more exposed to increases in prices. However, this can be counterbalanced by other measures, which can mitigate or even eliminate the regressive impacts of VAT, such as progressive changes to social security contributions or benefits and non-linear tax credits.[147]

7.4.3    Shifts in social protection financing and coverage of employment and life-course risks

A shift from social contributions towards general taxation also represents a shift from employment-based social protection (or insurance-based in a classical Bismarckian terminology) towards potentially more universal-based social protection provision (or of a more Beveridgean type in the classical terminology), since entitlements to social protection can be seen as less linked to earnings-related contributions. Furthermore, although social contributions and general tax revenues are sensitive to the business cycles, some Member States may have more room for a shift from social contributions to general taxes with smaller implications for the social protection entitlement structures.

Social protection expenditures can be grouped into three categories according to their link to individual employment histories and to whether or not they cover a life-cycle risk (see Figure 7.26), regrouping employment-related social protection provision (pensions, employment and disability), life-cycle and non-employment-related provision (health and family) and non-life-cycle and non-employment-related provision (housing and social exclusion).

Figure 7.26 — Social protection financing and expenditures structures (2010).

Source: ESSPROS.

While such a classification has some clear shortcomings,[148] it can help to identify Member States whose relative share of social contribution revenues tends to exceed the relative size of employment-related social protection expenditure. This is the case for Estonia, the Czech Republic and the Netherlands and, to a lesser extent, Slovenia, Germany, France and Lithuania. In these countries, there may thus be more scope to shift social contributions to other tax bases than in other countries.[149]

8.     Improving timeliness of data and the measurement of social outcomes

With the adoption of the Europe 2020 strategy, the European Union has placed the fight against poverty and social exclusion high on the political agenda. The reshaping of policy objectives through the Europe 2020 strategy brought to the fore the weaknesses of the statistics and indicators available to monitor poverty and inequalities. The lack of timely data on income and living conditions is a serious obstacle to the implementation of Europe 2020. The social consequences of the economic and financial crisis have made the lack of timeliness of data on the extent of poverty and social exclusion an even more burning issue — not least in the countries where the crisis has hit the hardest. In the conclusions of the December 2010 EPSCO,[150] ministers of social affairs recognise the importance of this issue and 'invite the Commission to support, in collaboration with the Member States, the timely availability of valid indicators to monitor the social dimension of the Europe 2020 Strategy'.

The setting of the poverty and social exclusion target also helped to highlight the need to improve the measurement of poverty and social exclusion. This is expressed in the conclusions of the June 2010 EPSCO[151] preparing the adoption of the Europe 2020 strategy, which proposed to ‘strengthen the current instruments for measuring progress in the reduction of poverty and deprivation […] and that the mid-term review of the EU headline target in 2015, […] also include a review of the indicators, … taking into account economic developments and improved measurement instruments.’

Implementation of the Europe 2020 strategy has shown some weaknesses hampering the monitoring of progress towards the EU’s social objectives. Taking into account non-monetary income in the definition of resources, including the value of publically provided services, is essential to capture the full impact of the welfare state, including public services, and to correctly identify groups worst affected by poverty.

The definition of material deprivation needs adjusting to reflect trends in living standards in the EU. Measuring and analysing the dynamics of poverty and exclusion will improve the design of policies and assess whether they have a durable impact on poverty reduction. Improving the measurement of the specific situations of women, children and young people, and of very severe forms of poverty and social exclusion (such as homelessness), will help monitor progress in addressing the social issues at the root of poverty and exclusion.

This section aims to identify the statistical and analytical gaps that hamper the monitoring and analysis of poverty and social exclusion. Improving measures and indicators in these areas would help reach a more accurate and timely diagnosis on which to base recommendations for policy intervention in the context of Europe 2020. Under the new programming period, the objectives of the ESF will be linked to those of Europe 2020 on employment and social inclusion and will support policies set out in National Reform Programmes in response to Country Specific Recommendations. More timely and more accurate data, together with efforts to develop poverty maps (see section 8.3.6), will help with programming and allocating the ESF and to monitor overall outcomes.

Three areas of improvement are explored: (i) improving the analysis, monitoring and dissemination of existing information; (ii) improving data collection systems at EU and national level; and (iii) supporting the development of methods and models and enhancing their use in policy making.

8.1 Harmonised EU statistics and indicators in support of the Europe 2020 target to reduce poverty and social exclusion

The EU target defines poverty and social exclusion on the basis of three main indicators: being at risk-of-poverty, being in severe material deprivation and people living in households with zero or very low work intensity (i.e. jobless or quasi-jobless households). It recognises the multi-dimensional nature of poverty and social exclusion, and allows account to be taken of the diversity of situations and priorities that prevail in the EU, in particular after the last waves of enlargement.[152] An in-depth analysis of the indicator underpinning the target is available in the 2011 review of Employment and Social Developments in Europe.

The choice of the three indicators is the result of a negotiation between Member States with very different poverty profiles and different policy priorities. It also reflects the fact that monitoring poverty solely on the basis of the at-risk-of-poverty rate has major drawbacks.

One of the main drawbacks of the at-risk-of-poverty indicator is its ambiguous movement in periods of rapid growth or of crisis. The risk of poverty depends on the poverty threshold, which is determined by the general level of income and its distribution in the whole population. This threshold may change from one year to another as individual incomes change.

This is especially the case during an economic crisis. Wages are usually the first to decrease as the situation on the labour market gets worse — and many people see their market incomes reduced as they become unemployed. But other incomes, such as pensions and social benefits, do not adjust immediately. As a result, the median income, and therefore the poverty threshold, tends to fall. People earning an income slightly below the poverty line may move above it, even though their situation has not changed  or may even deteriorated.

Statistical developments at EU level have supported work on social indicators. The development of social inclusion indicators and the adoption of the Europe 2020 target would not have been possible without significant EU investment in collecting comparable statistics on income and living conditions since the early 1990s. [153]

EU-SILC (Community Statistics on Income and Living Conditions)[154] is now the reference source at EU level for social statistics and has contributed to strengthening EU social policy coordination by underpinning the analysis and the comparison of Member States' performance in the social field. A key objective of EU-SILC is to deliver robust and comparable data on total disposable household income. Income components were defined to follow as closely as possible the international recommendations of the UN ‘Canberra Manual’.[155] The corpus of comparative research and analysis based on EU-SILC is constantly growing and nurtures the policy debate at both EU and national level. EU-SILC has also allowed the development of the comparative micro-simulation model Euromod,[156] which is a powerful tool for assessing the distributional impact of reforms to the tax and benefit systems (e.g., the impact of fiscal consolidation) and analysing the effectiveness of policies.

While EU-SILC helped to strengthen evidence-based policy-making at EU level,[157] the intensive use of EU harmonised statistics has revealed some shortcomings of the survey, data gaps and new needs. The planned revision of the EU-SILC Regulation provides an opportunity to address these issues (see detailed discussion below).

Another important source of harmonised data is the European System of Social Protection Statistics (ESSPROS). The system gathers administrative data on social protection expenditure and receipts in a harmonised framework. It enables the size, structure and functioning of national social protection systems to be compared and analysed in detail.

Finally, the analysis of poverty and social exclusion and its determinants also relies heavily on the Labour Force Survey which provides the key statistics on employment, unemployment and inactivity. Information on the quality of jobs, on barriers to work, and on access to and participation in training is especially relevant for understanding such phenomena as labour market exclusion or in-work poverty. Household information is especially useful for analysing the impact of one person’s labour market status on other family members (e.g., jobless households).

EU indicators enable monitoring and support diagnosis. The common EU indicators[158] are used for multiple purposes in support of EU-level social policy coordination. At EU level, they are the basis of regular reporting on the social situation of Member States in the context of the Joint/SPC Reports on Social Protection and Social Inclusion[159], as well as in the Commission annual review of Employment and Social Developments in Europe.  The indicators are primarily used for descriptive and comparative purposes to show the relative position of Member States vis-à-vis the multiple dimensions of poverty and social exclusion. They also illustrate, as far as possible, the extent and composition of policy intervention (ESSPROS social protection data).

The indicators are also used to monitor the progress of Member States towards the policy objectives, though the lack of time series (in the first years of EU-SILC) and the significant time lag (nearly two years) of SILC and ESSPROS data clearly affect the relevance of the exercise for policy-making, especially in the context of the crisis. The issue of timeliness is discussed below.

Over time, the indicators sub-group of the Social Protection Committee has developed analytical frameworks in which indicators are combined to support a diagnosis of the main determinants of poverty and social exclusion. An example of such a framework, based on benchmarking, was developed and agreed in 2008 to identify the main determinants of child poverty in each country.[160]

The indicators are also used by a broad range of stakeholders at national and EU level, including national administrations, social partners, organisations representing civil society and academia.

In the context of the Europe 2020 strategy, the use of indicators has been improved by adopting EU and national targets (see above) and by developing a Joint Assessment Framework (JAF) to combine indicators and benchmarks to identify the main drivers of poverty and social exclusion in a given country.

Indicator-based diagnosis needs to be supplemented with detailed country-specific information (both quantitative and qualitative) to support the policy advice each country receives in ‘Country Specific Recommendations’. The JAF provides screening to give a country initial guidance on structural challenges and areas that may need attention, thus supporting the identification of key employment and social challenges.

8.2  Improving timeliness

The lack of timely information on trends, and on poverty in particular, is the main data gap hampering evidence-based policy-making. Since the crisis, it has become very clear that policy-makers at EU and national level do not have the tools to monitor the short-term social impact of economic shocks, or the effectiveness of policy responses. We need to capture changes in social conditions at an early stage, and to identify those who are worst affected by the crisis, as well as those who benefit most from a policy change.

The detailed nature of the EU-SILC survey together with the fact that crucial data, such as income or the calendar of activity,[161] refer to the previous year[162] leads to significant delays in data availability of up to two years. Efforts are being made by the European Statistical System to shorten these delays while maintaining good data quality for the standard SILC delivery.

The Commission is currently exploring different ways to speed up the monitoring of social trends at EU level. A number of options have been identified and their feasibility will be assessed by the European Statistical System in 2013:

1. Early estimates of material deprivation (and possibly subjective poverty, monthly income) based on faster treatment of SILC data. Material deprivation measures are already more timely than income-based data, as they refer to the survey year (N-1) rather than to the ‘income year’ N-2. In addition, they could be treated faster and published earlier. Analysis also shows that the ‘economic strain’ dimension of material deprivation is quite responsive to the effects of economic shocks.[163] This property could be reinforced by developing questions on a household’s current situation. A few countries[164] have published early estimates of the poverty rate based on faster treatment of survey data about one year after the end of the income year.

2. Alternative indicators used as early warnings of deterioration in social trends:

- The financial distress indicator derived from the EU harmonised consumer surveys, collected monthly. This is well suited to signal significant changes in the financial situation of households, by broadly defined income groups (self-declared income quintiles). This indicator is very timely (a few months delay) and is currently published by DG EMPL in the European Employment and Social Quarterly Review.

- Monthly current income could be collected in a high frequency survey and used as an indicator per se, providing timely information on trends in incomes and their distribution, for broad age groups. If collected in the LFS, the indicator would be available with a delay of three to six months. The feasibility of such a system is to be assessed by ESTAT with national statistical offices in 2013.

3. Nowcasts of the poverty rate and related measures based on micro-simulation (taking into account policy and economic/labour market changes as far as possible). Nowcasts are estimates that are similar to economic forecasts, and would be available in year N for income year N. (see box on nowcasting with Euromod). The Commission is also testing the possibility of using the monthly current income survey (or the financial distress indicator derived from EU harmonised consumer surveys) to predict trends in poverty. The gain in timeliness would depend on the frequency of the survey used to collect such auxiliary variables.

4. Another type of information that is important to policy-makers concerns the behavioural response of households in reaction to an income shock[165] (due to unemployment, reduced working hours, separation, etc.), and the transmission channels through which household welfare is affected — labour markets, access to credit, government services.[166] This is especially useful in a downturn.

5. Trends in the disbursement of social benefits, drawn from administrative sources, typically available on a monthly or quarterly basis can provide timely information on increased pressure on safety nets. However, such measures are not comparable across countries and there may be major breaks in series when policies or administrative rules change. The SPC is currently monitoring the variables on the number of social benefit recipients/new registrations for selected social benefits.

Information on trends in the number of clients of social services (publicly provided, or through NGOs) — emergency services, shelters, soup kitchens, etc. — could also be collected more systematically and provide useful insight into pressure on social services. Currently, such information is only available on an ad hoc basis through service providers.

Nowcasting with Euromod

The method uses the micro-simulation model Euromod to adjust market incomes with what is known about their development (wages, prices, etc.) and simulate the effects of the current design of the tax and benefit system in 2012 (level of benefit, duration, conditionality, etc.). Further data adjustments are made to account for labour market developments between 2007[167] and 2012 (e.g. increase in unemployment). The method doesn’t take account of demographic and other compositional changes. However, it makes it possible to predict the potential increase in the risk of poverty and other variables (including the poverty threshold) for the total population and specific sub-groups. It can also illustrate the contribution of different factors to the change, e.g. worsening labour market conditions or changes in the tax and benefit system.

The example below presents preliminary results to be interpreted with great caution. If they are confirmed, they would point to an increase in median incomes in LT and LV, probably due to the improvement of labour market conditions. At the same time, an increase in the risk of poverty among children and the elderly would also reflect measures taken to freeze/reduce some benefits (such as child benefits and minimum pensions) in these countries.

Example of nowcasting the development of income distribution up to 2012 on the basis of SILC 2008 data (2007 incomes) — Change in indicator since income year of latest SILC statistics

Source: Euromod preliminary estimates — paper prepared for NetSILC2 conference December 2012. AROP60: at-risk-of-poverty rate (60 % of median)

The financial distress indicator

The Commission collects monthly information on consumer sentiment in the context of the programme of joint harmonised EU business and consumer surveys. These very timely surveys include a question on household financial situations, which has been used to derive a ‘financial distress’ indicator. The indicator focuses on households declaring that they had ‘to draw on their savings or go into debt in order to meet current expenditure’. Breakdowns are provided by household income quartile. These ‘financial distress’ data can provide a timely indication of trends in the share of the population whose households are facing financial difficulties, and how households in the different income quartiles have been affected by the crisis. The indicator shows that people with lower to middle income have seen their financial situation deteriorating faster than the rest of the population. In some countries, the gap is increasing very rapidly.

Furthermore, it can be used to some extent as an advance indicator of more established ‘hard’ indicators of trends in the social situation in many Member States, although the actual hard indicators it can predict depend on the particular Member State in question and there is no indicator/set common to all countries. The financial distress series may also help to signal when rather dramatic changes have occurred, i.e., when there are really noticeable developments in the underlying hard social indicators. Exploring its use as a key variable in a nowcasting model combined with a set of other potentially relevant, timely items seems worthwhile.

Among the possible developments suggested above, some are well advanced (financial distress indicator, nowcasts, SPC data collection, etc.); others are being investigated and would require further investment. The Commission and the Member States would need to identify the most promising avenues and set priorities for further developments accordingly.

It is also important to identify what use can be made of potential higher-frequency data on the basis of national experiences and empirical analysis. Member States’ current practices can be used as a source of inspiration; and time series will be analysed to identify the links between timely indicators and standard poverty and social exclusion measures.

It is crucial to consider a communication strategy on how to use and present these indicators[168] at EU level and in the Member States. The role of each indicator selected will need to be clarified, and the way in which they will relate to current measures of poverty and social exclusion explained. Good practice in the macro-economic field, in which a variety of indicators is used to assess and forecast the short-term developments in the economy, may provide inspiration here.

8.3  Improving the measurements of social outcomes

Being poor is primarily being short of money, hence the current focus on measures of income. However, there is a consensus that people who are deprived of dimensions of life thought to be essential in society — quality education, health and healthcare, employment, housing, access to public benefits, and social contacts — are also poor or excluded. This is why EU-agreed social inclusion indicators (used in the JAF, the SPPM or for the monitoring of the social OMC) go beyond income measures of poverty. However, a lot remains to be done to properly capture important aspects of poverty and social exclusion.

The definition of resources needs to take into account imputed rent and the value of in-kind transfers. The definition of material deprivation needs to be adjusted to trends in living standards in the EU. As the first set of longitudinal data becomes available, new analysis illustrates the importance of capturing the dynamics of poverty and exclusion. Finally, more work is also needed to reflect the gender dimension of poverty and exclusion, as well as the specific stages of the lifecycle, especially childhood and youth.

Awareness-raising campaigns and the impact of the crisis have highlighted very severe forms of poverty and social exclusion that persist within the EU, such as those facing the homeless, the Roma or people living in isolated rural areas. Specific measurement tools are needed to monitor their situation. The role of the EU in developing such tools and their use by policy-makers needs to be reviewed.

8.3.1    Non-monetary income components

In 2001, the report of the Canberra Expert Group on Household Income Statistics[169] identified four areas as the most fruitful for pursuing a fairer and more accurate picture of income distribution. These are (a) better estimates of property income, self-employment income and own-account production, (b) imputed rent for owner occupied housing, (c) social transfers in kind or non-cash government benefits, and (d) capital gains (especially negative).

A number of academic publications have also underlined the importance of integrating non-monetary income components into cash-based income measures. First of all it can improve the comparability of distribution results across different population subgroups. For instance, the importance of non-cash income varies across age groups; this is true of in-kind benefits such as education and healthcare or of imputed rent which affect different population groups depending on the structure of home ownership. It can also improve comparability of results across countries, since cash and non-cash public transfers vary substantially across Europe, as does the extent and structure of home ownership.

'Social transfers in kind' are a type of income: they are goods and services provided by government and non-profit institutions for free or at subsidised cost. Obvious examples include healthcare, education, housing and childcare. They represent an economic advantage to the households that benefit from them, but since they have no directly identifiable monetary value, this advantage is very difficult to quantify. Research (See ESDE 2011) shows that most in-kind benefits have a redistributive role to the benefit of the poorest segment of the population. This is why Member States that invest significantly in benefits in kind wish this impact to be reflected in poverty measures.

Methodological work is currently being conducted by the Commission also taking account of the work done in the context of EU-financed projects (Net-SILC[170] and EC/OECD joint partnership[171]). Currently, there is no agreed or common standard for valuing social transfers in kind for the purpose of understanding the distribution of income. In the National Accounts, social transfers in kind are measured as the sum of costs. This valuation could serve as a starting point, but more information is needed on how total income is distributed across the population. Specific information on actual use of services (including health care) or on barriers to access may help in allocating the value of transfers in kind to different populations. Such information could be collected every five years, possibly in an EU-SILC module. The Commission, together with the European Statistical System, will explore this possibility in the context of the revision of the EU-SILC Regulation.

Imputed rent is meant to take account of the economic advantage that people get from occupying their own house or from living in subsidised housing. Given the very different home ownership structures in different European countries, not taking account of imputed rent can affect the comparability of poverty measures. The valuation of imputed rent is done in the context of national accounts and experts recommend that it be included in the definition of income (Canberra manual).

In the context of EU-SILC, all Member States are asked to estimate the value of imputed rent on the basis of a common methodology (since 2008). In principle, this variable can be added to the current definition of income. The impact on poverty rates and other distributional indicators is significant in all countries: it decreases at-risk-of-poverty rates by 1 percentage point on average in the EU, and by more than 2 percentage points in ES, IE, EL, IT and MT, but it increases them by 1 percentage point in DE and FR). The impact is especially strong when comparing different population groups. For instance, the risk of poverty of old people tends to drop significantly if the economic advantage they may draw from home ownership is added to their current disposable income.

While there is broad consensus on the importance of taking account of imputed rents to analyse and compare systems and the situation of different population groups, a number of issues are raised concerning their use in the measurement of poverty and call for further methodological and harmonisation work within the European Statistical System.

· The economic advantage derived from home ownership or subsidised rents is not entirely ‘liquid’. It is not obvious that imputed rents can entirely be used for consumption or savings. This is especially true of an old lady living in a large family house, or of people living in subsidised rented accommodation, most of whom are unable to afford housing on the private rental market. This also raises the question of whether primary incomes and their distribution would be the same if most of the population had to find housing on the private rental market.

· Imputed rents are not observed, and as such their value depends a lot on the model and assumptions used. It is also argued that while they are useful for analytical purposes, they are not suitable for descriptive statistics.

8.3.2    Improving the measurement of material deprivation

When adopting the targets of the Europe 2020 strategy, the Council asked for a revision of the material deprivation indicator by 2015, in the broader context of the mid-term target review. Currently, severe material deprivation is measured as the enforced lack of at least four items from the following list: ‘cannot afford (1) to pay rent/mortgage or utility bills, (2) to keep home adequately warm, (3) to face unexpected expenses, (4) to eat meat, fish or a protein equivalent every second day, (5) to take a one-week holiday away from home, (6) a car, (7) a washing machine, (8) a colour TV, or (9) a telephone.' This indicator was adopted by the SPC as a complement to the relative poverty indicators based on current income, taking account of non-monetary resources. Because it is based on a single European threshold, this indicator is also a step towards a more ‘absolute’ measure of poverty. It captures the differences in living standards between countries, as well as the impact of growth on those standards in a given country.

The underlying idea is that the cumulative enforced lack of a given number of items reveals a latent trait in people’s living conditions, called material deprivation. However, the indicator adopted in 2009 had scope for improvement. The number of items was too low and some have become obsolete in today’s Europe (e.g. a colour TV, and a washing machine to some extent).

Making use of the 50 material deprivation items from the 2009 wave of EU-SILC, a proposal for a new indicator has been developed by a Eurostat Task Force on material deprivation[172], on the basis of a detailed assessment of the dimensional structure of these 50 items, their suitability, validity, reliability and additivity. The list of items currently envisaged includes: not being able to afford some new clothes, two pairs of shoes, a meal with meat, chicken or fish every second day, to keep the home adequately warm, to pay for arrears (mortgage/rent, utility bills, hire-purchase instalments), to face unexpected expenses, the enforced lack of a personal car if needed, a computer with an internet connection, to replace worn-out furniture, some money for oneself, regular leisure activity, getting together with friends/relatives for a drink/meal monthly, one week’s annual holiday away from home.

This list would provide a solid basis on which to build the revised indicator. The main improvements are: (1) a higher number of items, which makes the indicator more robust and less sensitive to individual items, (2) the improved relevance of the items. This revised indicator would be in line with the Council definition of poverty and social exclusion, which considered as poor ‘the persons whose resources (material, cultural and social) are so limited as to exclude them from the minimum acceptable way of life in the Member State to which they belong’.

8.3.3    The dynamics of poverty and social exclusion

Poverty is not a permanent state and individuals might stay/exit/enter or even re-enter into it again. From a political point of view, it is crucial to reach out to those in persistent poverty, to prevent those who might enter (or re-enter) poverty from doing so, and to help others to escape from it. Evidence shows (ESDE 2012) that very different patterns of the dynamics of poverty prevail in the Member States.

A better understanding of poverty dynamics would help to target those most at need and better prevent the others from entering into persistent poverty. The longitudinal dimension of EU-SILC, which is beginning to be exploited, is a significant source of greater understanding even if some technical issues have until now inhibited its full use.

The modernisation of household statistics is an opportunity to improve the tools for longitudinal analysis (both in EU-SILC and in the LFS). Areas for the improvement of longitudinal data in EU-SILC include: (1) revised design allowing for longitudinal indicators with better precision and more reliable breakdowns; (2) fine-tuning of variables that could help to better identify the trigger events (job loss, family separation, health problem, etc.), and their dates of occurrence and their impact on poverty and social exclusion; (3) Improving follow-up of people by carefully implementing tracing rules and keeping contacts between dates of interview; (4) Description of non-respondents and control for usual bias limiting panel data (attrition, censoring, non-homogeneous non response). Key variables helping to evaluate the impact of social policies and different welfare regimes on those transitions would also be extremely valuable.

8.3.4    Capturing the gender dimension of poverty

Ample empirical evidence and academic research document the relative disadvantages that women face on the labour market (gender pay gap, the “glass ceiling”, etc.) as well as for their integration in society. However, on-going work at EU level on the development of a Gender Equality Index has highlighted that poverty and income inequality indicators only partially reflect these disadvantages. The main reason is that most indicators aiming to measure access to resources (income, material deprivation) are based on the assumption of equal sharing of resources within the household.

Existing information could nevertheless be better used to monitor the specific situation of women. In EU-SILC, a number of variables are collected at the individual level, in particular some of the income components (e.g. earnings), some material deprivation items (including from among the new list of items envisaged) as well as variables measuring access to services, such as an unmet need for medical care (see also section on access to services). These variables could be analysed more systematically in the context of standard poverty analysis, and specific indicators could also be defined and integrated in the regular monitoring framework.

8.3.5    Capturing the situations of children and youth


Major steps have been taken in recent years to improve the way in which the situation of children is captured. In 2008, the SPC report on child poverty[173] reviewed existing data and indicators that could be used to analyse the situation of children and provided a diagnosis of the main determinants of child poverty. The report made 14 recommendations to improve the monitoring of child poverty and well-being in the EU. A number of these have already been followed up at EU level.

The child dimension of existing EU indicators in the field of social inclusion and health has been strengthened through the development of new age breakdowns, as well as by refining the low work intensity indicator.

Data collection on child-specific deprivation through the 2009 EU-SILC thematic module on material deprivation has been substantially improved, and includes 20 child- specific items. These have been used by the Eurostat Task Force (with the support of Net-SILC 2) to produce a child deprivation indicator.[174] The regular collection of the list of child deprivation items is currently being discussed in the context of the SILC revision.

The Commission[175] and the SPC[176] have carried out extensive work to identify a list of indicators best suited to monitor child poverty and well-being. On this basis, the Commission has identified a set of indicators for monitoring children’s situation in the Recommendation on Investing in Children, as part of the Social Investment Package.

Together with the OECD, the Commission has carried out a review of international surveys of children to identify what key indicators could be derived.[177] DG RTD has launched a call for tender[178] to set up an EU-wide longitudinal survey of children.

Other new steps could be taken as regards data and indicator development.

The current EU-wide surveys do not satisfactorily capture the status of children in the most vulnerable situations (for example this cannot be measured by SILC) and they could usefully be complemented with information relating to children outside traditional households (e.g., alternative care), coming from vulnerable or ethnic minority backgrounds, migrant children, children from an ethnic minority background, children with a disability. There should be specific efforts dedicated to exploring possible data sources and methodologies to collect data on these children.

While there are already important indicators on the health status of children, data on one important dimension of health inequalities children face, the social gradient, are not easily available. This could become a priority in future work to enable appropriate assessment and monitoring of policy interventions.

Existing information on participation in childcare is not sufficient to give accurate information on affordability and quality aspects. These are crucial for supporting parents who go out to work, and for child development. Better measures on affordability may be developed to enable improved monitoring, better informed policy-making and better assessment of the long-term impacts of quality early childhood education and care services. The 2014 EU-SILC thematic module on material deprivation could be used for collecting more information on early childhood.

Young people

The at-risk-of-poverty or social exclusion measurement raises issues that are peculiar to the 18 to 24 age group. Indeed, becoming poor is closely linked to the timing of departure from the parental home. This differs a lot across Europe. Since poverty and exclusion are measured at household level, youth poverty rates are higher in countries in which young people have access to their own resources and lower in countries in which achieving autonomy is more difficult. Scope for improvement will be discussed in an ad hoc expert group on youth indicators set up by the Commission (DG EAC).[179]

In April 2010, the Employment Committee and its Indicators Group (DG EMPL) agreed on a definition and methodology for a NEET indicator to be used in the context of the Europe 2020 strategy. Young people aged 15-24 who are not in employment, education or training (NEETs) have a much higher risk of remaining unemployed, of experiencing poverty and/or of being socially excluded in the future than others in society.

8.3.6    Measuring the most extreme forms of poverty and social exclusion

In June 2010, the European Council called for more work on the measurement of severe forms of poverty and social exclusion. Specific methods are needed to capture the most severe forms of poverty. Methods include looking at more severe (lower thresholds, overlaps between deprivation and low income) and persistent forms of the existing measures of poverty. The analysis of the depth and duration of poverty and exclusion, as well as the accumulation of disadvantage reported in the ESDE 2012 could contribute to formulating of the most appropriate policy responses to severe poverty.

However, the main obstacle to measuring ‘extreme poverty’ is that our main sources (surveys, and to a certain extent registers) do not capture those who are most excluded from society. EU-SILC is not the appropriate tool for two main reasons.

(1) EU-SILC only covers private households, which means that those most excluded from society (the homeless, people living in institutions such as prisoners, migrants, elderly, mentally ill) do not answer the SILC questionnaire.

(2) The measurement of very low income through surveys or registers suffers from major quality problems, such as the treatment of negative income from self-employment, the high non-response among people in the most precarious situations, etc.

Further work is needed on developing adequate measures of the different forms of extreme poverty. However, it is unlikely that a single indicator could capture what is at stake. A great deal of methodological work has already been carried out, not least to measure homelessness and housing exclusion, or the specific situation of the Roma. Concerning the Roma and other ethnic minorities, it has to be added that major legal and sociological barriers prevent the collection of statistics on ethnicity in most EU countries.

In the area of homelessness and housing exclusion, the Commission has asked Member States to collect specific information on the number of homeless people in the 2010 round of censuses (results due in 2013-14). The results of this data collection will be evaluated to assess the added value of such an exercise, and whether it should be repeated on a regular basis.

The Commission supports the development of relevant tools/infrastructure at national level through methodological work. The Commission is promoting the use of a harmonised nomenclature of situations of homelessness and housing exclusion (ETHOS), which should facilitate the compilation of data from different sources. It is also advocating the further involvement of national statistical institutes in the collection of data on homelessness, in partnership with service providers. [180]

The Commission is also investing (in cooperation with the World Bank) in the development of poverty maps that aim to identify local areas of multiple and severe disadvantage, including areas where there are large concentrations of Roma. Such tools can support better targeting of policy intervention for Roma (including through the ESF), and other population groups living in areas that face multiple disadvantages.

[118] For more information on expenditure on healthcare see European Commission Staff  Working Document – Investing in Health SWD(2013) 43

[119] See European Commission (2012) Employment and social developments in Europe 2012,  chapter 3 (Welfare systems) for a detailed analysis of the stabilisation function of social policies by country, highlighting the diversity of the impact of the crisis and of welfare responses across the EU.

[120] See European Commission (2012) Employment and social developments in Europe 2012

[121] See European Commission (2012) Employment and social developments in Europe 2012 key features.

[122] For Lithuania, Hungary, Romania and Bulgaria, GHDI data are only available until 2010.  GHDI data for Luxembourg are not available for 2009–2011.  GHDI data for Malta are not available. National currencies deflated by HICP, DG EMPL calculations.

[123] Data are only available until 2010.

Actual change in GHDI: GHDI in the last year of the given period is compared with GHDI in the first year of the given period (change is expressed as a percentage).

Contribution of social transfers to change in GHDI: the change in social transfers between the first and last year of the given period is calculated and its contribution to GHDI change is calculated.

Contribution of taxes to change in GHDI: the change in taxes between the first and last year of the given period is calculated and its contribution to GHDI change is calculated.

Change in GHDI if social transfers and taxes stayed at 2007/2009 value: GHDI is computed as the sum of all its components (including social transfers and taxes) with the value from the last year of the given period and compared with GHDI calculated with the initial year value of social transfers and taxes (change is expressed as a percentage).

Countries are sorted based on the size of the contribution of social transfers to GHDI change in 2007-2009

[124] See European Commission (2012) Employment and social developments in Europe2012 Chapter 3 (Welfare systems) for a detailed analysis of the effectiveness and efficiency of social policies, in the fields of poverty reduction and labour-market friendliness.

[125] See European Commission (2012) Employment and social developments in Europe 2012

[126]  The approach taken here derives from the work of Hemerijck (2012) who attempts to measure budgets for social investment by combining the budgets for those policies that have greatest investment focus. He combines budgets for active labour market policies, child-care, education, research and rehabilitation. The analysis excludes rehabilitation due to lack of data. In this approach, the remaining social expenditures contain social protection benefits such as old-age and survivor's benefits, unemployment-related benefits and disability benefits, healthcare, family benefits in cash and housing benefits.

[127] See Nelson, M and Stephens, J.D. (2011) 'Do social investment policies produce more and better jobs?” in Morel, N; Palier, B; and Palme, J. (eds.) Towards a Social Investment Welfare State? Polity Press.  

[128] From OECD (2012) 'Starting Strong III: A Quality Toolbox for Early Childhood Education and Care' available at: http://dx/

[129] OECD (2012)  'Education at a Glance 2012' table C1

[130] European Commission Communication - Rethinking education: investing in skills for better socio-economic outcomes. {COM(2012) 669 final}.

[131] In the EU-LFS, the indicator on lifelong learning denotes the percentage of persons aged 25 to 64 who received education or training in the four weeks preceding the survey. The information collected relates to all education and training, whether relevant to the respondent’s current or possible future job or not. It includes formal and non-formal education and training. This means general activities in the school/university systems but also courses, seminars, workshops, etc. outside the formal education system, regardless of the topic.

[132]  Salanauskaite, L. and Verbist, G. (2011) "Is the "neighbour's lawn greener? Comparing family support in Lithuania and four other NMS " Gini Discussion paper 25.

[133] Groups are obtained by cluster analysis based on scores related to the following variables: children living in a jobless household, children living in households at work and at-risk-of-poverty and the impact of social transfers on children’s risk of poverty. For each of these variables, the scores reflect both the situation of children in the country versus the rest of the population, and the situation of children in the country versus the rest of Europe.

LU has not been introduced in the classes as it is an outlier. Trends in risk of poverty rate indicate the trend in the risk-of-poverty rate between 2005 and 2010. Countries in brackets are to be considered as on the edge of the cluster.

[134] Avram S., Figari F., Leventi C., Levy H., Navicke J., Matsaganis M., Militaru E., Paulus A., Rastrigina O. and Sutherland H. (2012) , "The distributional effects of fiscal consolidation in nine EU countries", Research note 01/2012, Social Situation Observatory (forthcoming)

[135] In SILC and LFS the coverage rates are always below 100 % because in these surveys the people identified as unemployed are asked about whether they actually receive benefits (some of them might not be eligible, such as young people or those who have lost their entitlements, and some of them might not be claiming benefits). People that continue receiving benefits when they start to work are not taken into account in the surveys for this purpose. This is different in the administrative sources, which also include among the unemployment benefits recipients at work but still receiving benefits, so that coverage rates can thus exceed 100 %.

[136] Assumptions of the OECD tax-benefit model: The net replacement rates summary measure is defined as the average of the net unemployment benefit (here without social assistance and cash housing assistance) replacement rates for two earnings levels (67% and 100% of average wage), three family situations and 60 months of unemployment.

[137] Note: No legal maximum of duration of unemployment benefit in BE .

[138] European Commission (2006) Employment in Europe.

[139] OECD (2012) Employment Outlook

[140] See Chapter 1 of European Commission (2012) Employment and social developments in Europe 2012 for a detailed analysis of the dynamics of long-term unemployment.

[141] Spending on active labour market policies includes categories 2-7 in the LMP database.

[142]    The Figure presents the transitions for those people aged 15-74 unemployed the year before to unemployment (U), employment (E) or inactivity (I), depending on whether the person was registered with the national Public Employment Service (PES), and whether s/he was receiving unemployment benefits. The longitudinal data used here are based on yearly estimates for nine Member States: Estonia, Romania, Cyprus, Greece, Hungary, Italy, Malta, Sweden and Slovakia. As very few people receive benefits without being registered with the PES, the values for this category are not reliable and therefore not shown.

[143] See Chapter 1 of European Commission (2012) Employment and social developments in Europe 2012, in particular Section 4.6 for related detailed econometric analysis, notably controlling for effects of various individual characteristics, such as education or age.

[144] General contributions include the category 'other receipts' which are generally in nature closer to taxation than to social contributions.

[145] Covering social protection in a broad sense, as reflected in the harmonised European system ESPROSS. Other contributions are here taken together with general government contributions since their nature is generally less similar to social contributions.

[146] 1995 or earliest year available (LT and SI 1996, LV 1997, EE and HU 1999, CY 2000, BG 2005, PL and RO 2000

[147] The Mirrlees Review published by the Institute for Fiscal Studies (2011) found that, for instance in the United Kingdom, increasing all means-tested benefit and tax credit rates by 15% would counter the regressive impacts of VAT. Conversely, it found that applying zero or reduced rates of VAT to items on which poorer households spend a relatively large proportion of their budgets is a blunt instrument with which to help the less well-off, because richer households typically gain more in cash terms from these tax breaks than poorer ones.

[148] The classification of the social protection function into three categories is based on an approximation of the earnings-related and life-cycle risk nature of each category. Expenditure classified as earnings-related may include universal benefits (e.g. minimum pensions, some disability benefits), while non-earnings-related groups may include earnings-related benefits (e.g. sickness benefits). For a more accurate analysis, a more detailed breakdown of social protection expenditures would be needed.

[149] Other approaches may also be used as a complement, for instance based on the relative distribution of the risks across the population, on the rationale that less equally distributed risks would probably generally require some higher financing share through general taxation. For instance, life-cycle risks (pensions, health, family) or labour market risks (e.g. unemployment, disability) can be considered as more equally distributed than social exclusion and housing. In this context, Member States in which the relative share of social contributions revenues is larger than the relative size of more equally distributed life-cycle risks may have more room for a financing shift towards general taxes.

[150] Council Conclusions on 'The social dimension in the context of an integrated Europe 2020 strategy' 3053rd Employment, Social Policy, Health and Consumer Affairs Council Meeting 6 December 2010:

[151] The EPSCO conclusions were endorsed by the European Council on 17 June 2010 ( ).

[152] For a more detailed presentation of the new target and the diversity of populations and forms of poverty it represents, see Chapter 3 of European Commission (2011) Employment and social developments in Europe 2011 (

[153] The European Community Household Panel (1994 to 2001) was the first EU wide harmonised data collection exercise to cover all EU countries. It stimulated a wealth of comparative research and analysis that made possible and underpinned the development of the first set of EU social inclusion indicators adopted by the Laeken European Council in 2001. In 2005, the ECHP was replaced by EU-SILC.

[154] EU-SILC Framework Regulation of the European Parliament and the Council (No1177/2003)

[155] United Nations (2001) or The Canberra Group (2001) Final Report on the Expert Group on Household Income Statistics

[156] See:

[157] An illustration of the value added of EU-SILC is the support provided by Eurostat in setting the Europe 2020 target on poverty and social exclusion. Without comparable and trustworthy micro-data available for all Member States, it would not have been possible for Eurostat to simulate the number of ‘poor and excluded’ (and their characteristics) that a given definition of the new indicator would capture. This information was a crucial element in the negotiations that led to the adoption of the poverty and social exclusion target, thereby securing the social dimension of the Europe 2020 strategy.

[158] This refers to portfolios of indicators developed together with Member States to support the monitoring of policy coordination process at EU level (e.g. Indicators of social inclusion and social protection, European Community Health Indicators). See also European Commission Staff  Working Document – Investing in Health SWD(2013) 43

[159] See:

[160] See Social Protection Committee (2008)‘Child poverty and well-being in the EU — current status and way forward’

[161] 'Calendar of activity' refers to the monthly information about the activity status of an individual during a reference year

[162] It should be noted that the reference year of the income to the previous year allows the best possible measurement as the respondent has the fiscal declaration at his disposal

[163] For instance, items such as ‘ability to face unexpected expenses’ or ‘ability to afford a week of holidays away from home’ have been responsive to the crisis while the main indicator was still stable.

[164] A few countries have already produced or are planning to produce early estimates (ES, CZ, PT, RO, AT, LV, NL, SK).

[165] Such a module has been run as a stand-alone survey or as a module in existing surveys (LFS, LITS) in a few EU and neighbouring countries (Bulgaria, Romania, Latvia, Croatia, Serbia) at the request of the World Bank.

[166] Examples of variables that can be envisaged include: the share of people having to reduce their expenses (by type of expenses — food, healthcare education, housing, etc.); the share of people having to draw on their savings or go into debt; the share of people who increase their working hours because their partner has lost their job; the share of people experiencing difficulties in accessing essential services (healthcare, education, housing, banking, etc.).

[167] The EUROMOD model is currently being updated to refer to the latest SILC data available

[168] In a few countries, nowcasts are produced for the sole use of policymakers, but are not published.

[169] The Canberra Group (2001) Final Report on the Expert Group on Household Income Statistics

[170] Net-SILC: a network for the analysis of EU-SILC financed by the European Commission (Eurostat) and bringing together data producers (national statistical offices) and data users.

[171] EC/OECD joint partnership on the impact of publicly provided services on the distribution of resources.

[172] Members of the Eurostat Task Force include representatives of national data producers (National Statistical Offices), of main users from DG EMPL, SPC-ISG and academia.

[173] European Commission (2008) Child poverty and well-being in the EU: current status and way forward.

[174] The UNICEF has also used the child deprivation items collected in the SILC 2009 module to produce a child deprivation index, published in Report Card No 10. See:

[175] Child poverty and child well-being in the European Union, TARKI Social Research Institute Hungary and Applica Belgium, published in January 2010.

[176] Social Protection Committee (2012), SPC Advisory Report to the European Commission on 'Tackling and preventing child poverty, promoting child well-being' 27 June 2012.

[177] OECD (2012) An Evaluation of International Surveys of Children

[178] See 'Towards a European longitudinal childhood and youth survey'

[179] European Commission Staff Working Document – 'EU indicators in the field of youth' SEC(2011) 401.

[180] See European Commission Staff Working Document  'Confronting homelessness in the European Union' SWD(2013) 42