Jump to ContentJump to Main Navigation
Child well-being, child poverty and child policy in modern nations (Revised 2nd Edition)What do we know?$

Koen Vleminckx and Timothy M. Smeeding

Print publication date: 2001

Print ISBN-13: 9781861342539

Published to Policy Press Scholarship Online: March 2012

DOI: 10.1332/policypress/9781861342539.001.0001

Show Summary Details

Child well-being in the EU – and enlargement to the East

Child well-being in the EU – and enlargement to the East

Chapter:
(p.99) Four Child well-being in the EU – and enlargement to the East
Source:
Child well-being, child poverty and child policy in modern nations (Revised 2nd Edition)
Author(s):

John Micklewright

Kitty Stewart

Publisher:
Policy Press
DOI:10.1332/policypress/9781861342539.003.0005

Abstract and Keywords

This chapter highlights the differences among the current members of the EU, and contrasts them as a group with the applicants. It considers whether applicants have a comparative advantage over members in any dimension of well-being. The criteria for European Union (EU) membership and the approach that is currently taken by the European Commission to measure the differences in living standards within the Union are discussed first. This is followed by a discussion on the three dimensions of well-being of European children: their education, health, and economic welfare.

Keywords:   comparative advantage, well-being dimension, EU membership criteria, European Commission, living standards

Introduction

The European Union (EU) is expanding – over the next decade as many as 13 new members may be admitted, ten of them transition countries in Central and Eastern Europe (CEE). In this chapter we consider measurable differences in the well-being of children between current club members, the EU member states, and the ten CEE applicants seeking admission1.

Discussion of the suitability of applicants to join any club provides an opportunity to look in a mirror and consider the state of the existing membership. We therefore emphasise the differences among the current members as well as contrasting them as a group with the applicants. And we consider whether applicants have a comparative advantage over members in any dimension of well-being – a possibility that is completely overlooked in both media and academic focus on the relative economic strengths of the two groups of countries. To anticipate one result: Slovenia has an under-5 mortality rate that is below the EU average.

We first discuss (a) criteria for EU membership, emphasising the human rights dimension, and (b) the approach currently taken by the European Commission to measuring differences in living standards within the Union. In both cases we emphasise the need for a much broader view than is typically taken, and one that includes a comprehensive picture of the well-being of children – the 79 million in the current EU15 and the 25 million in the CEE applicants.

We then consider in turn three dimensions of well-being of European children; their economic welfare, their health, and their education. The (p.100) account we give is far from complete, but this rough sketch provides a starting point for the full picture that we argue is needed.

EU membership, human rights and ‘cohesion’

The best-known criteria on which applicants for EU membership are judged, and those which form the corner-stone of negotiations over accession, are the economic criteria. Applicants must have a functioning market economy in place and the capacity to cope with competitive pressures and market forces.

However, the ‘Copenhagen Criteria’, agreed in 1993 to govern EU entry, are in reality much wider than this. In addition to meeting the economic and financial conditions, “membership requires that the candidate country has achieved stability of institutions guaranteeing democracy, the rule of law, human rights and respect for and protection of minorities”. As the former foreign affairs commissioner, Hans van den Broek, put it in an interview on enlargement in December 1998, “we are not merely a Union of economic benefits, but also a Union of shared values”.

The non-economic aspects of the Copenhagen Criteria are taken seriously in relation to applicants. (We return below to the issue of human rights within existing member states.) In June 1997 all ten CEE countries with the exception of Slovakia were agreed to have met the political criteria, but a number of them were judged to have progress still to make with respect to democracy and the protection of human rights and minorities.

In the October 1999 Commission report on progress towards accession, all CEE candidate countries, and Slovakia in particular, were agreed to have made considerable advances in both these areas. But corruption, the judiciary and the treatment of minorities (in particular the Roma) were raised as ongoing concerns. In addition, Romania’s chances of being granted membership were in effect declared conditional on action to improve the situation of the 100,000 children in institutionalised care. As the 1999 report on progress states: “The rights of these children to decent living conditions and basic health care is a human rights issue” (Composite Paper, Reports on Progress Towards Accession by Each of the Candidate Countries, October 1999)2.

(p.101) Human rights and EU enlargement

There is no doubt that conditions in Romania’s orphanages are indeed a human rights issue. Human rights today are internationally recognised as encompassing much more than just political and civic freedoms. Both the Universal Declaration of Human Rights (1948), and, more recently, the UN Convention on the Rights of the Child (CRC) (1989) have been ratified by all EU and CEE countries. Both documents embody a wide range of rights, ‘positive’ as well as ‘negative’, implying the responsibility of the state not only to protect all individuals from violation by others, but also to provide them with certain conditions and opportunities for life. The CRC, for instance, includes articles on the right of all children to education, to ‘the highest attainable standard of health’ and to ‘a standard of living adequate for physical, mental, spiritual, moral and social development’.

These rights receive little attention in the Progress Reports, beyond a confirmation that freedom of education, the right to a minimum subsistence level of income and (in some cases) the right to health are ‘guaranteed’ by national constitutions3. Poor living conditions for children in orphanages are mentioned in the original Opinion on Bulgaria, but only in the case of Romania do they receive extensive and repeated attention4. In addition to the emphasis on children in institutions, the Romanian report points to concerns about the economic well-being of the population, noting the deterioration in living standards caused by economic contraction and increased unemployment, and highlighting the fact that World Bank poverty data indicate that one third of the population live below the subsistence level. No other report refers to material living standards, although it is well known that these have deteriorated across the region5.

Human rights within the existing member states also need to be considered. It is safe to say that in no EU country are the economic and social rights of all children fully respected in practice. The recent attention of the UK government (rightly placed as we will see later) to such issues as child poverty and social exclusion among the young is a reflection of this. Some might argue that the principle of ‘subsidiarity’ within the EU (the requirement that decisions be taken at the closest level to the citizen at which they can be effective), coupled with the existence of an official UN monitoring body for the CRC, implies that the EU itself should have little concern with human rights within member states6. But such a position is inconsistent with the treatment of the applicant countries (p.102) and the Union’s external policy in general, which takes a principled stand on human rights in dealings with other states. And it is also at variance with concern for greater ‘economic and social cohesion’ within the union.

Economic and social cohesion in the EU

The Treaty on Union emphasises that the strengthening of ‘economic and social cohesion’ among the member states is one of the EU’s principal goals, essential to the Union’s ‘overall harmonious development’ (Articles 2 and 130a). In the first of what is to be a series of regular reports on progress towards this goal, the Commission defined cohesion to imply “greater equality in economic and social opportunities”, towards the ultimate end of “rais[ing] living standards and the quality of life” (European Commission 1996, p 15). However, this first report stuck to disappointingly traditional measures of opportunity, largely employment and average income – although it did also take a brief look at poverty7.

We argue elsewhere for a much broader vision of cohesion, one that takes a wider view of development and of individual well-being, more similar, for example, to that of the UNDP’s Human Development Reports and the associated ‘capability’ approach of Amartya Sen (Micklewright and Stewart, 1999, 2000). Such a view, with its attention to a wide range of different dimensions of individual well-being, leads naturally to attention to the position of children within the EU, both those in the current member states and those in the CEE applicants. Convergence in well-being across the Union means moving towards similar opportunities for European children wherever they are born – because children are European citizens too, and because the nature of their childhood helps determine the shape of Europe’s future.

The importance of children’s well-being to society’s future is underlined by considering the socialist inheritance of the CEE countries in terms of human development. The comparative advantage of socialism over capitalism in the field of family policy is often exaggerated, as is the extent of ‘cradle to grave’ support provided under socialism (Ferge, 1991, Atkinson and Micklewright, 1992). Nethertheless, it is broadly the case that policy in socialist countries attached much greater emphasis to some aspects of children’s development than was the case in other countries at similar levels of national income. The results in terms of healthcare and educational outcomes can still be seen today (as we show below), despite the upheavals of the past decade. It is partly because of the past investments (p.103) in children in the CEE region that the EU is now able to contemplate applications from countries that in some cases are hugely adrift in terms of current levels of national income.

Similarly, it is important that attention is paid to current investments in children in CEE countries if the path taken by societies in the region is to result in greater cohesion in the long-term in an expanded Union. If some children miss out on key aspects of their development as a result of the economic and social pressures of transition, it will not be easy for them to recoup those deficits in later life. All children may suffer if public expenditure on education, for example, falls precipitously. And greater inequalities in health and educational opportunity, caused in part by a widening distribution of income, may lead to the exclusion of particular groups. None of this will aid economic and social cohesion in the future EU.

Measuring the well-being of Europe’s children

Implementation of human rights or the strengthening of cohesion clearly involves change to policy. However, an inspection of the policies represented by the legislation and administrative frameworks of EU and CEE countries provides an insufficient guide to what actually happens in practice. The complete canvas, when painted, would involve the depiction of both polices and outcomes, with a full palette of colours employed. But here we attempt only an incomplete black and white sketch in which the policies do not feature explicitly8.

We focus on a handful of outcome indicators of child well-being that reflect (although they do not precisely measure) how far some of the rights enshrined in the CRC are a reality for children in EU members and CEE applicants. They also relate to several dimensions of well-being that should enter a full assessment of cohesion in the Union. We look at the material welfare of children, their health and their educational attainment. In particular we focus on differences in:

  • child poverty;

  • mortality among children and young people; and

  • school enrolment and learning achievement.

We concentrate on data for the most recent year available, but at times we also put the rankings that these provide in the context of recent (p.104) trends. (The trends within the member states are the focus in our earlier work reported in Micklewright and Stewart, 1999, 2000.)

Space and data constraints rule out the analysis of several issues. Youth unemployment is conspicuous by its absence from our list. This is a subject of concern both as a measure of labour market opportunity for the young and because of its association with other problems. Youth unemployment rates in 1997 varied from 7% in Austria to 39% in Spain, with an EU15 figure of 21% (Eurostat, 1999a, p 143; ILO definition, 15–24 year olds). Rates among the CEE applicants in 1996/97 averaged exactly the same – 21% – and ranged from 8% in the Czech Republic to 36% in Bulgaria (Eurostat, 1999a, p 43). Both members and applicants clearly face a serious problem, but one that varies from country to country.

A further restriction is that our investigation is almost entirely at the national level – the 15 member states and the ten CEE applicants. This contrasts with the Commission’s own analysis of cohesion that focuses in particular on the sub-national unit of the region (of which there are nearly 200 in the EU countries (NUTS-2 level regions) and around 90 in the CEE applicants). A major motivation for the Commission’s approach is that a number of EU policies to encourage employment and income growth operate at the regional level (in the sense of eligibility for funding). While we are critical of the Commission’s examination of cohesion as being too narrow in terms of subject matter, we should acknowledge that the combination of analysis at both the national and sub-national level clearly provides for a richer investigation.

This combination would still be possible with a broader approach to cohesion. Child-related indicators such as infant mortality and teenage fertility that are based on population and vital statistics are available at the sub-national level (as we illustrate below). And the use of these other indicators at the regional level would enrich the analysis of incomes and employment. For example, raising a region’s GDP will not prevent social exclusion if a high rate of births to teenage mothers prevents greater employment opportunities being shared by all young women9.

What sort of variation among members and applicants in any indicator might we consider large? Figure 4.1 presents the benchmark of differences in GDP per capita, the indicator that forms the popular approach to representing variation in living standards across the EU and CEE countries. The four lowest values among the EU15 are for the so-called ‘Cohesion Four’, where EU policy aims to promote catch-up growth (the tag emphasises the association in practice within the Commission of the concept of cohesion with income per head). When the figure for each (p.105)

Child well-being in the EU – and enlargement to the East

Figure 4.1: GDP per capita (purchasing power parity) in 1997 (EU 15=100)

Source: Eurostat (1999a, p 232) and Eurostat (1999b, p 67)

country is weighted by population size, the differences within the EU15 have narrowed over time, with GDP per capita in the Cohesion Four rising from two thirds of the EU15 figure in 1983 to three quarters by 1996. With the exception of Luxembourg, all the other member states have GDP per capita within 15% of the average.

The experience of the CEE applicants in the 1990s has been very different, with marked falls in national income during the process of transition away from the socialist system. The economies in almost all cases have now been growing for several years (Bulgaria and Romania are exceptions with a rocky recent period), but there is clearly enormous catch-up growth to be made if typical levels of EU income are to be reached. Only Slovenia had income per head in 1997 as high as the bottom limit of the EU15 range – and only in Poland had GDP grown to exceed its 1989 level (several other Central European countries were nearly there but Bulgaria and the Baltics were still 25–45% down). In Romania, Lithuania, Latvia and Bulgaria, national income per head in 1997 was less than a third of the EU15 average. The differences within the CEE group are much larger than those among the EU members (the weighted coefficient of variation is nearly three times higher).

The effect on the comparison between members and applicants of moving away from a focus on national income per head can be summarised (p.106) by a comparison of the ranking on GDP per capita with that on the UNDP’s Human Development Index (HDI). Besides GDP per capita, the HDI includes life expectancy and a composite educational indicator combining enrolment and literacy. Both groups of countries are on average higher up the (world) ranking on the HDI than they are on GDP per capita alone (UNDP, 1999, Table 1). But while the average difference in the ranking was four places in 1997 for the EU15 it was as much as 13 places for the CEE-10, reflecting in part the inheritance from the socialist period of the investments in children referred to earlier. Allowance for the two other dimensions of human development starts to change the picture.

Child poverty

Imagine calculating the numbers of children in EU members and CEE applicants that live in poor households, defining poverty as income below a line common to all 25 countries – a line fixed everywhere as the same number of Euros. The differences in national income per head shown in Figure 4.1 would be reflected strongly in the ranking of national child poverty rates that would emerge from such a calculation. But Figure 4.1 would be a far from perfect guide to the variation in child poverty since the differences in the graph refer only to average incomes. The differences within each country, and where children come in those national income distributions, would also enter the equation. The ranking among the richer EU countries, in particular, would be far from a mirror reflection of the ranking of GDP per capita (as we will see).

Such a common poverty line, establishing a fixed threshold relevant for every household, whether it be in Sweden or Slovakia, is one way of thinking about poverty in Europe. In terms of solidarity or cohesion within the Union, a common threshold appears attractive, as it captures what convergence of incomes is about: poorer parts of Europe, and children with lower living standards, catching up with the others.

But ‘poverty’ should be a meaningful term within each country, relevant to national sentiment and policy making as well as to international comparison. What a person needs to participate in normal life clearly depends on what those around him or her have. It is this concept of living standards relative to the national norm that is found in the EU’s broad definition of poverty as persons with “resources (material, cultural and social) that are so limited as to exclude them from the minimum acceptable way of life in the Member States in which they live” (Eurostat, (p.107) 1997, p 3; emphasis added). Poverty, in terms of cash incomes, is assessed by comparison with a national yardstick, and not a common EU poverty line as in the calculation outlined above10.

What measure of child poverty would a human rights approach lead one to? The CRC provides for the right to “a standard of living adequate for physical, mental, spiritual, moral and social development” (Article 27), which among other things underlines one reason why child poverty is such a cause for concern – its impact on an individual’s future life. The wording in the CRC suggests a relative measure, given that what is adequate for full participation in society will be higher where average incomes are higher. When the European Commission’s report on Romania’s progress towards accession draws attention to the large numbers of people living below the subsistence level, that level is defined within the national context.

Figure 4.2 shows the proportion of children in EU and CEE countries living in households with incomes below 50% of the national median, a conventional definition of relative poverty. The data are drawn from the Luxembourg Income Study’s collection of harmonised household surveys and the same adjustment for differences in household size and composition have been applied to the data for each country. The majority of member states are present in the graph, but there are data for only four applicant countries. In view of the changes in income inequality that have occurred in the CEE countries during the 1990s (which we return to below) we have indicated for these four the exact year to which the data relate.

The variation in child poverty across the EU countries is enormous, ranging from 3% in Finland to 21% in Italy and the UK. (Both the extent of the differences and the average level have risen somewhat since the mid-1980s; Micklewright and Stewart, 1999.) These figures underline how misleading national income per head may be as a guide to material living standards of children at the lower end of national income distributions. Figure 4.2 showed Finland and UK at the same level of GDP per capita but here they are at opposite ends of the spectrum. Since the UK is a much more unequal country, the child poverty rate there is much higher (the difference in the years to which the data in Figures 4.1 and 4.2 refer is not important in this case).

The four CEE countries also display considerable variation. Hungary and Poland had rates of child poverty in the early 1990s which were above the (unweighted) average for the EU of 10%, but the two parts of former Czechoslovakia had rates – in 1992 – that were below even that of the best performer among the EU member states.

(p.108)

Child well-being in the EU – and enlargement to the East

Figure 4.2: Child poverty rates (%)

Notes: The EU figures typically relate to the early 1990s. The rates displayed are the percentage of children living in households with income below 50% of the national median. Household incomes are equivalised by the factor (adults +0.7*children)^0.85.

Source: Bradbury and Jäntti (1999, Table 3.3)

The position of Slovakia and the Czech Republic in Figure 4.2 reflects the low degree of income inequality that these countries inherited from the socialist period. Former Czechoslovakia was an extreme case but the broad picture found across the CEE countries at the start of the 1990s was one of less dispersion in household incomes than that generally found among EU members.

Figure 4.3 shows the picture of income inequality in the second half of the 1990s, giving the Gini coefficient for the distribution (among individuals) of household per capita incomes in the ten CEE countries (higher values of the Gini indicating higher levels of inequality). The average value for the distributions calculated on the same basis for 11 EU countries in the mid-1980s is also shown (the EU range at this time was from 0.22 in Finland to 0.36 in Ireland). (More recent EU figures on the per capita basis are not available.)

Seven of the applicant countries have levels of income inequality just (p.109)

Child well-being in the EU – and enlargement to the East

Figure 4.3: Income inequality in CEE countries (1997): Gini coefficient of per capita income

Source: CEE countries from UNICEF (1999, Table 10.13), except the Czech Republic (Vecernik, 1999, Table 2), Hungary (Galasi, 1998, Table 1) and Latvia (data supplied by Latvian Statistical Office). Data for the Czech Republic and Slovenia are for 1996; data for Hungary are for March-March 1995/96. The Hungarian figure is from the TARKI survey (also the source drawn on in Figure 4.2); the official budget survey gives a lower Gini coefficient (Flemming and Micklewright, 1999, Figure 11). The EU average (which relates to the mid-1980s) is from Flemming and Micklewright (1999, Table 3) drawing on Atkinson et al (1995, Table 4.10)

or well above the EU average, with only Slovakia, Slovenia and the Czech Republic clearly below. Estonia and Bulgaria are at the top end of the EU member range. It is notable that the richer countries (in per capita GDP terms) tend to have lower levels of inequality, and vice versa, implying that differences among the CEE countries, if poverty were assessed with a common Euro line, would be even larger than the differences in average incomes.

(p.110) The 1990s have seen some significant increases in income inequality in the CEE countries. This means that the impact on living standards of the resumed economic growth in the second half of 1990s has been at least partly offset for children at the bottom of the distribution. If average incomes rise by a third but the distribution of incomes widens by the same amount then the incomes of those at the bottom of the distribution will remain unchanged. The problem has not been restricted to the CEE applicants. The experience of the UK over the 1980s was one in which rising average incomes and increased income inequality went hand in hand.

Were figures for child poverty available on the same basis for all ten CEE applicants for the late 1990s, they would most likely display wide variation, as among the four in Figure 4.2. The best performers would shame countries such as the UK and Italy among the current member states. (In fairness it should be noted that the reduction of child poverty has recently moved centre-stage in the UK government’s domestic policy.) And the worst performers would threaten cohesion if admitted to the Union.

Finally, it is important not to lose sight of what life at the bottom of the income distribution really means for children. The careful statistical measurement of child poverty among both EU members and CEE applicants needs to be accompanied by an understanding of the consequences for children’s daily lives. Analysis of cohesion in the current or future Union would benefit from drawing on such work as Roker (1998) for the UK and World Bank (1999) for Bulgaria, in which poor families and children are interviewed about their experiences. The plight of children in these studies highlights the question of whether the right provided by the CRC to an adequate standard of living given available national resources is being protected sufficiently.

Mortality among children and young people

Indicators of mortality and morbidity, or of health risk, are natural candidates for inclusion in any broad assessment of the EU’s ‘overall harmonious development’ and hence of cohesion between current and future members. (We have already referred to the presence of life expectancy in the UNDP Human Development Index.) A rights-based approach to development also leads easily to the same indicators. For example, the CRC requires countries to take appropriate measures to (p.111) diminish infant and child mortality, besides recognising the rights of children to “the highest attainable standard of health” (Article 24).

A number of indicators of health status among the young would be interesting choices to look at, both for comparison between the EU members and applicants taken as groups, and for comparison within each group. These include several relating to teenagers and other young people that reflect health-related behaviour, the leading cause of ill-health in industrialised countries for those in their teens and early-20s. Smoking and alcohol abuse, drug-taking and unsafe sex are in principle all candidates for inclusion, although in practice the availability of data certainly varies. For example, there were on average two cases of syphilis per 100,000 persons (of all ages) in the EU in 1997, and about the same in Central Europe, but around 85 in Bulgaria and Lithuania and over 120 in Latvia (UNICEF, 1999, p 71). And incidence is much higher among the young.

In this chapter, however, we restrict ourselves to measures of mortality. Sen (1998) gives three reasons for considering mortality in any appraisal of human development: first, there is the intrinsic importance attached to life itself; second, the fact that other ‘capabilities’ are contingent on being alive; and third, that mortality indicators provide rough proxies for a number of other values for which data are less readily available – most obviously morbidity.

We take one indicator relating to young men aged 15–24, deaths through accidents and injuries (including suicide and murder), a serious and under-recognised problem in the transition countries and one for which variation is surprisingly high within the EU. But we start by looking at deaths among the under-5s from all causes, the most widely-used measure of child mortality in developing countries.

Under-5 mortality

Figure 4.4 shows that under-5 mortality rates (U5MR) in several of the CEE applicants are not too out of line with those in the EU. The Czech Republic is within the EU range while Slovenia performs particularly well with a rate that is even below the EU15 average. If member states with a higher rate were to have had the Slovenian U5MR, there would have been 4,000 fewer deaths among children of this age in the EU15 in 1996.

On the other hand, the poorer countries of Bulgaria, Latvia and Romania stand out with U5MRs that are far higher. A child born in these countries has a probability of death before the age of five of 2% or (p.112)

Child well-being in the EU – and enlargement to the East

Figure 4.4: Under-5 mortality rates (1996)

Note: Data are for 1992 for Belgium, 1993 for Ireland and Italy, 1994 for Spain and 1995 for Finland.

Source: WHO Health for all and UNICEF TransMONEE databases

(p.113) more. And if all CEE applicants were to have had the same U5MR as the EU average, there would have been nearly 11,000 fewer deaths among the under-5s in the ten countries.

The rate shown for Romania is similar to that in Greece in the mid-1970s; and those in Latvia and Bulgaria are similar to the rates in Austria, Germany, Italy and Spain at the same time. This suggests that these CEE countries are currently about 20 years behind progress in the EU. That said, it is wrong of course to think of the EU15 as a homogeneous group. The variation today within the EU is such that 10,000 lives would have been saved in 1996 if all member states had had the U5MR of Sweden. Even among the richer countries already in the club there is clear room for improvement.

The case of Slovenia, with GDP at the lower limit of the EU range but with a better performance than the EU average in terms of child survival, illustrates a wider phenomenon. Although higher national income in general implies lower child mortality (the association is less clear cut if attention is restricted to just the EU members), the CEE applicants as a whole perform better than one would expect given their levels of GDP.

To show this we took the top 100 countries in the world ranked by 1995 GDP per capita (the poorest CEE applicant ranked number 100) and ran a regression of U5MR in 1996 on GDP per capita (in purchasing power parity terms) in 1995 (both in natural logs), including three dummy variables for the 15 EU members, the ten CEE applicants and the four other former socialist countries in the region (in the top 100 on GDP). The results imply that the CEE applicants (and the other socialist countries) had levels of U5MR that on average were nearly 60% less than would be predicted on the basis of just GDP per capita, and, when allowing for the differences in national income, 25% less than those of EU members11.

Of course, national income declined sharply in all the CEE countries during the 1990s and in most has not yet recovered to 1989 levels. Did mortality among young children rise as a result? The answer in most cases is ‘no’ and one of the more welcome trends of the transition in terms of human welfare has been the continued decline in U5MR in many CEE applicants (Bulgaria is an exception). Most countries have therefore been living up to their responsibility under the CRC to ‘diminish child mortality’ despite severe economic setbacks.

We have noted that indicators based on population and vital statistics, such as mortality rates, are available for sub-national analysis. For the EU15 we have made some preliminary investigation of the infant mortality rate (IMR) at the level of the (NUTS-2) region and its association with (p.114)

Child well-being in the EU – and enlargement to the East

Figure 4.5: Injury death among 15- to 24-year-old males (1994)

Source: WHO World Health Statistics Annual (various years) and UN Demographic Yearbook

(p.115) regional GDP per capita. (Deaths among infants – children less than 12 months old – account for the bulk of under-5 mortality.) Taking average values for 1994–96 for regions in ten countries (data for the other five were not readily available), we found a correlation between the two variables of only –0.42. In some countries IMR varied little despite significant differences in GDP (eg, Germany) while in others IMR differed notably among regions with similar GDP (eg, Greece). The Greek rates varied from 4.5 deaths per thousand live births in Notio Aigaio (the richest region) to 10.6 in Dytiki Ellada (one of the poorer ones).

We have less information on the differences within the CEE countries but it is clear that as in the member states the size of the variation is not the same in each country. IMR in Hungary in 1997 ranged from 8.8 in Western Transdanubia to 11.7 in Southern Transdanubia. But in Bulgaria the extremes in the same year were 9.7 in the capital, Sofia, and as much as 27.4 in the region of Bourgas bordering on Turkey. (Our information is drawn from the 1997 Hungarian and 1998 Bulgarian statistical yearbooks.)

Injury death among young men

Death from injuries is one of the most important causes of mortality in Europe for children above the age of five. These are all, by definition, preventable deaths, and the same article of the CRC that refers to infant and child mortality explicitly mentions the prevention of accidents.

Figure 4.5 shows mortality rates from all forms of injury across EU members and applicants for males aged 15–24 in 1994. (The rates for females in all countries are much lower, proof that the variation across country cannot be due to some peculiarly national bad luck.) Four categories of cause of death are distinguished: deaths from motor vehicle accidents (‘road deaths’), other injury death excluding those from violence (eg drownings, burns and poisonings), homicide and suicide (we include among suicides those deaths from violent causes undetermined whether deliberate or not – see the discussion in Micklewright and Stewart, 2000, chapter 4).

A word of comment is merited on the differences between the concepts measured in Figures 4.4 and 4.5 and on their orders of magnitude. A figure of 200 deaths per 100,000 persons aged 15–24 implies that each year everyone in this age group has an average probability of 0.2% of dying from injuries. Over the ten years an individual spends in this age range the probability is therefore 2% – a figure that is strikingly high (and how (p.116) much greater must be the burden of non-fatal injuries alongside those that result in death). The rates in Figure 4.4, on the other hand, are already converted into probabilities over a numbers of years, five in this case. If U5MR is equal to 20 per 1,000 this implies that the average probability of death over five years is 2%.

Injury death rates vary enormously, differing by a factor of nearly 7 to 1 between the worst CEE performer, Latvia, and the best EU performer, the Netherlands. In Latvia and Estonia the probability of death from injury over ten years exceeds 2% (assuming an unchanged mortality rate) and in Lithuania it is only just short of this level. The Baltics stand out within the CEE countries, their average rate being some three times higher than that in the other applicants – the latter have rates that are at or above the EU average. But within the EU there is considerable variation too, with Austria three times higher than the Netherlands, and above five of the CEE countries. Several other member states have rates above those of Romania, Hungary, the Czech Republic and Bulgaria.

What types of injury cause these big differences? Homicides represent a negligible amount of injury death except in the Baltics (and to a lesser extent Bulgaria and Romania). The homicide rate alone in Estonia is equal to the total injury death rate in the Netherlands. These figures seem to accord with the general impression of a more violent society that has characterised the 1990s in some other parts of the former Soviet Union. Suicide is also notably high in the Baltics, but here a matching record is found in the EU in Finland. Elsewhere in the CEE region, similar suicide rates exist to those in various other EU members. Road deaths are not at an exceptionally high level in the Baltics by EU standards, although Estonia and Latvia are outside the EU15 range (they vary by a factor of about 3 to 1 within the Union), while in a number of other CEE countries they are below the EU average. But Latvia, Lithuania and Estonia again stand out as having a very poor record on ‘other accidents’ with rates around six times higher than the average across the EU15.

How have these figures changed over time? The broad pattern – all CEE countries above the EU average, with the Baltic states way out of line – does not depend on the year selected, at least for the early 1990s (LSHTM, 1998). Within the EU, Finland has emerged over the last 25 years as an outlier on suicides, although it is also the case that the EU suicide rate rose by 40% from 1970 to 1995 (Micklewright and Stewart, 2000). Some positive news on the Baltics is that injury death declined during the years 1994–97 in all three countries by between 10–35% (p.117)

Child well-being in the EU – and enlargement to the East

Figure 4.6: Education enrolment (1995) and maths achievement (1994/95)

Notes: The correlation between the two series is 0.73. b Maths achievement score for the UK is a weighted average of scores for England and Scotland. The score for Belgium is a straight average of scores for the Flemish and French educational systems.

Source: Enrolment data from UNESCO website (http://unescostat.unesco.org/indicator/indframe.htm, Age-specific Enrolment Ratios). Maths scores from Beaton et al (1996, Tables 1.4 and 1.5)

(UNICEF, 1999, Annex Table 3.6; data for males aged 5–19, excluding suicide)12.

Education – enrolment and achievement

The narrow view of economic and social cohesion within the EU as largely a matter of income and employment opportunities leads one quickly to consider differences in educational outcomes among both member and applicant countries. (The Commission’s own work on cohesion draws attention to some of the differences.) But education has (p.118) many effects beyond that of increasing material living standards. These enrich a person’s life in various ways, giving him or her the capability to achieve a wider set of Sen’s ‘functionings’. For example, Haveman et al (1998) list (in the US context) “occupational prestige, health status, efficiency in consumption, marriage and fertility choices, and offspring quality” (p 346). All these effects of education are clearly relevant to the wider view of cohesion that we argue is needed.

An approach to child well-being based on human rights should also have education as a key element in the list of issues to be considered. The CRC emphasises the right to education, and to the development of the child’s abilities to their fullest potential (Articles 28 and 29). And it underlines that countries should achieve educational rights “on the basis of equal opportunity”13.

How then do educational opportunities vary across the member states and the CEE applicants? The different dimensions of education for children defy easy summary. One approach is to consider inputs, in particular public expenditure as a percentage of GDP. Here the member states certainly exhibit convergence since 1980, although the Cohesion Four still had the lowest spending in 1995 (Micklewright and Stewart, 2000, Figure 5.1). The CEE applicants have seen large falls in real expenditure per pupil on account of lower levels of national income. But broadly speaking they have maintained (or in some cases increased) education’s expenditure share at typically about 5% of GDP, the level on which the EU member states are converging (Hungary and Bulgaria are two exceptions; UNICEF, 1999, Annex Table 7.6). However, moves to decentralise provision and financing of education in several of the former centrally planned economies raise the issue of within-country differences in funding (UNICEF, 1998).

One level of educational provision on which most of the CEE applicants still score well is pre-schooling, reflecting in part the former socialist system’s commitment to child development and in part its need to mobilise female labour supply. Enrolment rates dipped in the early 1990s due to both supply- and demand-side factors but now seem in general to be recovering. The rates vary substantially (the Czech Republic and Hungary led the field in 1997 with rates of over 80% for 3- to 5-year-olds, while Latvia and Lithuania brought up the rear with rates of around 50% and 40% respectively of 3- to 6-year-olds).

There is increasing recognition in all industrialised countries of the value of pre-school investment in children, and the CEE countries’ tradition in this area is an important inheritance. Tradition varies (p.119) substantially in the EU members with countries such as Ireland and the UK recently initiating new policy to promote pre-schooling, or realising the need to do so.

Figure 4.6 shows two measures of national educational performance for older children, age-16 enrolment rates and age-14 achievement in standardised maths tests. Results for the latter are taken from the Third International Maths and Science Study (TIMSS), a rigorously conducted investigation of learning achievement in some 40 countries. The graph is restricted to those countries for which both indicators are available, 12 from the EU and 6 from CEE.

The vertical axis shows the enrolment rates. The EU15 have been converging strongly since the mid-1960s, with the average rate rising (Micklewright and Stewart, 1999, Figure 4). Nevertheless, in Greece and Portugal around 20% and 30% respectively of 16-year-olds were not in education in mid-1995. These are the two poorest EU members in terms of national income per head; but the UK, with GDP per capita at the EU average, is another laggard (and even more so in earlier years).

The enrolment rates in the CEE countries display substantial variation. Romania, Lithuania and Latvia, three of the poorest applicants, have rates similar to that for Portugal – although the fact that Portugal’s measured GDP per capita is some two to three times higher should not be forgotten. The Czech Republic is up with the best of the EU members.

But total enrolment is only part of the story. The issue of what form of education the enrolment is in has come to be especially important in the CEE countries. The educational systems of the planned economies encouraged many children, especially boys, away from a general education at age 14 or 15 and into vocational and technical schools, some of them highly specialised. (Romania is an example of a country where there was extreme specialisation in schools.) It is the broader skills taught in general secondary schools that are in greater demand in the transition to a market economy, and these schools also provide the easier route into tertiary education, which is undergoing a big expansion. In fact the story here in the ten CEE applicants is typically a positive one, with falling enrolments in vocational and technical schooling and rising enrolments in general secondary schools.

Whether young people in CEE and EU countries are all on the same level playing field depends not just on the type of school and the curricula but also on what they actually manage to learn. The horizontal axis of Figure 4.6 gives the TIMSS results for maths at age 14, showing the (p.120) percentage of children who scored at least as high as the median among all children in all 40 participating countries.

Among both members and applicants, countries with higher enrolment rates in general have the further advantage of children with higher maths scores too, emphasising the differences that exist between the educational systems in both sets of countries. The differences in maths achievement are indeed considerable, larger than those for enrolment. In Lithuania and Romania only half as many children reach the international median as in the Czech Republic. Portuguese children are hugely adrift of those in most of the rest of the EU.

It is also clear that the EU members do not dominate the CEE applicants regarding maths achievement. The Czech Republic out-performs all 12 EU members, and Hungary comes higher in the ranking than eight of them. Bulgaria, the poorest country, comes higher than such large EU countries as Germany and the UK. Even the two Baltic countries and Romania are on a par with Greece and Spain, as well as being well ahead of Portugal.

These data on learning achievement would repay more detailed analysis. They could be used to reveal much more about the extent of variation in achievement within countries. Our focus on the percentage of children reaching the international median can be likened to the application across EU members and applicants of a single poverty line, with the results affected by both differences between countries in average scores and the differences within them. One notable result in the published TIMSS report for maths is that the advantage in learning achievement associated with greater parental education seems just as large in the CEE countries (Beaton et al, 1996, p 103). This questions whether ‘equal opportunity’ is indeed the basis for educational progress in the transition countries – an issue of continued relevance in EU members as well.

How has learning achievement changed in the CEE during the transition years? Evidence suggests increasing inequality in what children have learned – in line with increasing income inequality (UNICEF, 1998; Micklewright, 1999). But investigation with comparable cross-country data will have to await the results of the TIMSS follow-up study conducted in 1999. Comparisons of all the EU members with both themselves and Poland, Hungary and the Czech Republic will also be possible for a wider assessment of ‘functional literacy’ among 15-year-olds being conducted by the OECD. Results for 16- to 25-year-olds from a more limited exercise in 1994 showed Poland to be far adrift of the six EU members included in the survey in terms of the ability to absorb the (p.121) correct information from written documents. Ireland and the UK scored well below Sweden and the Netherlands (OECD, 1997, p 154)14.

Conclusions

Many of the CEE states applying for EU membership bring an impressive inheritance with them. A long history of low-income inequality, combined with a tradition of investment in the social sectors in general and in children and the family in particular, has left a lasting mark. On several indicators of child well-being there are CEE countries that compete well with current EU members, despite much lower levels of national income, and even after the upheavals of the last decade.

Fourteen-year-olds in the Czech Republic, for example, performed better at standardised maths tests in 1995 than their peers in any other present or potential EU state, while the level of achievement in both Hungary and Bulgaria was well above the EU average. The limited data available also suggest that the former Czechoslovakia would emerge well from a comparison of child poverty rates with its Western neighbours: child poverty was just 2% at the start of the transition, and even by 1997 both the Czech Republic and Slovakia had income inequality below the 1980s EU average.

Not all applicants have so much to be proud of, but performance in several areas is still impressive given current levels of national income. Romania and the Baltic states did fairly poorly on the standardised maths tests, for instance, but no worse than Greece or Spain and much better than Portugal, all countries with substantially higher per capita GDP. Similarly, under-five mortality in all CEE states except Slovenia is above the EU average – but well below that of other countries with similar levels of national income.

In some ways, the comparison of EU and CEE performance shows up the gaps in EU achievement, or at least that in certain member states (there are of course wide disparities within the EU group). While the current membership ponders whether its eastern neighbours are fit to join the EU club, it should also note that the new members may bring important lessons with them. If admitted tomorrow, in some areas of child welfare some CEE countries would set new high standards for the rest of the Union to follow.

On the other hand, there are also large disparities in performance within the CEE group, and in several applicant countries there are matters for concern. As noted, the standard of living of children in Romania’s child (p.122) institutions has been raised as an issue by the EU’s reports on progress towards accession, but attention should be paid to rising child poverty in much of the CEE region. We also described the disturbingly high level of violent death among teenagers and young adults in the Baltic states. This reveals a side to society not picked up by the standard economic indicators of progress in the Baltics, raising questions about whether these countries are ready to join a ‘union of shared values’.

Our discussion has focused on just a few indicators of child well-being, and the chapter has not attempted to provide a comprehensive picture. We have sought to highlight dimensions of well-being that are not being studied under the current interpretation of the Copenhagen Criteria governing EU entry. Without diminishing the importance of the issues that are at present considered in EU progress reports, such as the treatment of minorities and the realisation of basic civic and political freedoms, we have aimed to draw attention to the fact that human rights extend beyond these boundaries. The accession criteria should not represent double-standards, however. The enlargement process offers a valuable opportunity for existing member states to hold up a mirror on their own performance. Few, if any, should be fully satisfied with what they see.

The indicators examined are also important from the perspective of the analysis of economic and social cohesion within the EU, which ought to be broader than the Commission’s current focus on employment and incomes. While many aspects of well-being tend to correlate roughly with per capita GDP, the impact of other factors, including public services and the distribution of income, ensure that GDP is only ever the beginning of the story – a fact underlined by the superior performance of some CEE countries in some areas. A serious study of economic and social cohesion should cover a wide variety of aspects of life, and there would be no better place to start than with the very disparate opportunities currently faced by the 100 million children of the EU and CEE.

Notes

(1) The CEE applicants are the Czech Republic, Hungary, Poland, Slovenia, Estonia, Slovakia, Latvia, Lithuania, Bulgaria and Romania. (The first five were originally treated as being at the front of the queue.) The three other official applicants are Cyprus, Malta and, since December 1999, Turkey.

(p.123) (2) The Commission concluded that Romania fulfils the Copenhagen political criteria, ‘on the assumption that the authorities continue to give priority to dealing with the crisis in their child care institutions’. The Commission’s initial Opinions on Applications for Membership and subsequent Reports on Progress can be found at http://europa.eu.int/comm/enlargement.

(3) Commenting on the initial Opinions, Nowak (1999) notes that their analysis of human rights “seems rather superficial and relates more to the de jure than the de facto situation” (p 695).

(4) While the conditions in Romanian orphanages may be of particular concern, excessive institutionalisation and the factors associated with it are a region-wide problem in the CEE countries (UNICEF, 1997). Bulgaria, rather than Romania, had the highest rate of infants aged 0–3 in institutions in 1997 (over 1%) and the growth in institutionalisation among young children has been largest in Estonia, with a rise of 75% over 1989–97, followed by Latvia with a rise of two thirds (UNICEF, 1999, Annex Table 8.1).

(5) Other rights relating to children are mentioned in the reports, although again in haphazard fashion. Several reports refer to issues relating to juvenile justice. The 1999 report on Poland raises the need for policies to address violent and abusive treatment against children, while the Slovenian report notes that provisions on the rights of children are included in the Constitution. The original Opinion on Lithuania highlights the absence of appropriate legislation against child prostitution or the sexual abuse of children. Children also come up in discussion of minority rights (eg, with regard to discrimination against Roma and to the question of citizenship for ethnic Russians in the Baltic states).

(6) The periodic reports of EU and CEE countries to the UN Committee on the Rights of the Child, which monitors the CRC, can be found at http://www.unhchr.ch/html/menu2/6/crcs.htm. Issues of human rights within the EU are discussed at length in Alston (1999). (See in particular the introductory chapter by Alston and Weiler.) In extremis, the Treaty on Union permits the suspension of benefits of membership in the case of a member state committing a ‘serious and persistent breach’ of human rights.

(7) The Commission’s Sixth Periodic Report on development of EU regions extends the first cohesion report’s analysis to the CEE countries (European Commission, 1999).

(p.124) (8) The convergence of polices affecting children within the EU member states is considered by a variety of authors, including Ruxton (1996) and Ditch et al (1998).

(9) Of course, greater employment opportunities may help reduce teenage motherhood, but the strength of the local labour market is far from being the only factor involved (the issues are reviewed in Micklewright and Stewart, 2000).

(10) We focus exclusively on cash incomes in this section but the EU definition is a reminder that ‘poverty’ can be the deprivation of things other than money. On the issue of a common EU line, see Atkinson (1998) who explores the idea of a weighted average of common and national thresholds.

(11) The difference between the coefficients on the CEE and EU dummies was significant at the 2.5% level. The data were drawn from the Human Development Report.

(12) The same source suggests that the rate for Slovenia in Figure 4.5 is much higher than for other years in the 1990s.

(13) Hammarberg (1998) discusses in detail the shape of educational policy based on the principles in the CRC.

(14) Results of the TIMSS follow-up and of the new OECD assessment are both expected to be available in 2001 (see http://timss.bc.edu and http://oecd.org/els/pisa).

References

Bibliography references:

Alston, P. (ed) (1999) The EU and human rights, Oxford: Oxford University Press.

Atkinson, A.B. (1998) Poverty in Europe, Oxford: Basil Blackwell.

Atkinson, A.B. and Micklewright, J. (1992) Economic transformation in Eastern Europe and the distribution of income, Cambridge: Cambridge University Press.

Atkinson, A.B., Rainwater, L. and Smeeding, T. (1995) Income distribution in OECD countries, Social Policy Studies No 18, Paris: OECD.

(p.125) Beaton, A., Mullis, I., Martin, M., Gonzalez, E., Kelly, D. and Smith, T. (1996) Mathematics achievement in the middle school years: IEA’s Third International Mathematics and Science Study, Chestnut Hill, MA: Centre for the Study of Testing, Evaluation and Educational Policy, Boston College.

Bradbury, B. and Jäntti, M. (1999) ‘Child poverty across industrialized nations’, Innocenti Occasional Paper no ESP 71, Florence: UNICEF Innocenti Research Centre.

Ditch, J., Barnes, H., Bradshaw, J. and Kilkey, M. (1998) Developments in national family policies in 1996, Brussels: The European Observatory on National Family Policies, Commission of the European Communities.

European Commission (1996) First report on economic and social cohesion 1996, Luxembourg: Office for Official Publications of the European Communities.

European Commission (1999) Sixth periodic report on the social and economic situation and development of the regions of the European Union, Luxembourg: Office for Official Publications of the European Communities.

Eurostat (1997) Statistics in focus (Population and Social Conditions), no 6.

Eurostat (1999a) Eurostat yearbook 98/99, Luxembourg: Office for Official Publications of the European Communities.

Eurostat (1999b) Statistical yearbook on Central European countries 1998, Luxembourg: Office for Official Publications of the European Communities.

Ferge, Z. (1991) ‘Social security systems in the new democracies of Central and Eastern Europe; past legacies and possible futures’, in G.A. Cornia and S. Sipos (eds) Children and the transition to the market economy, Aldershot: Avebury.

Flemming, J. and Micklewright, J. (1999) ‘Income distribution, economic systems and transition’, Innocenti Occasional Paper no ESP 70, Florence: UNICEF Innocenti Research Centre.

Galasi, P. (1998) ‘Income inequality and mobility in Hungary, 1992–96’, Innocenti Occasional Paper no ESP 64, Florence: UNICEF Innocenti Research Centre.

Hammarberg, T. (1998) ‘A school for children with rights’, Innocenti Lecture no 2, Florence: UNICEF Innocenti Research Centre.

(p.126) Haveman, R., Wilson, K. and Wolfe, B. (1998) ‘A structural model of the determinants of educational success’, in S. Jenkins, A. Kapteyn, and B. van Praag (eds) The distribution of welfare and household production: International perspectives, Cambridge: Cambridge University Press.

LSHTM (London School of Hygiene and Tropical Medicine) (1998) ‘Childhood injuries: A priority area for the transition countries of Central and Eastern Europe and the Newly Independent States’, LSHTM: Final Report under contract 97/C/29 for UNICEF Regional Office for CEE/CIS/Baltics, Geneva, available at www.lshtm.ac.uk/centres/ecohost.

Micklewright, J. (1999) ‘Education, inequality and transition’, Economics of Transition, vol 7, no 2, pp 343–76.

Micklewright, J. and Stewart, K. (1999) ‘Is the well-being of children converging in the European Union?’, Economic Journal, vol 109, no 459, pp F692–F714.

Micklewright, J. and Stewart, K. (2000),The welfare of Europe’s children: Are EU member states converging?, Bristol: The Policy Press.

Nowak, M. (1999) ‘Human rights “conditionality” in relation to entry to, and full participation in, the EU’, in P. Alston (ed) The EU and human rights, Oxford: Oxford University Press.

OECD (1997) Literacy skills for the knowledge society, Paris/Ottawa: OECD/Human Resources Development.

Roker, D. (1998) Worth more than this: Young people growing up in family poverty, London: The Children’s Society.

Ruxton, S. (1996) Children in Europe, London: NCH Action for Children.

Sen, A. (1998) ‘Mortality as an indicator of economic success and failure’, Economic Journal, vol 108, no 446, pp 1–26.

UNDP (1999) Human Development Report 1999, New York, NY: Oxford University Press.

UNICEF (1997) Children at risk in Central and Eastern Europe, Regional Monitoring Report no 4, Florence: UNICEF Innocenti Research Centre.

UNICEF (1998) Education for all?, Regional Monitoring Report no 5, Florence: UNICEF Innocenti Research Centre.

UNICEF (1999) Women in transition, Regional Monitoring Report no 6, Florence: UNICEF Innocenti Research Centre.

(p.127) Vecernik, J. (1999) ‘Distribution of household income in the Czech Republic in 1988–1996: readjustment to the market’, Luxembourg Income Study Working Paper no 198.

World Bank (1999) Bulgaria: Consultations with the poor (background paper for WDR 2000) available from http://www.worldbank.org/poverty/wdrpoverty.

Notes:

(1) The CEE applicants are the Czech Republic, Hungary, Poland, Slovenia, Estonia, Slovakia, Latvia, Lithuania, Bulgaria and Romania. (The first five were originally treated as being at the front of the queue.) The three other official applicants are Cyprus, Malta and, since December 1999, Turkey.

(p.123) (2) The Commission concluded that Romania fulfils the Copenhagen political criteria, ‘on the assumption that the authorities continue to give priority to dealing with the crisis in their child care institutions’. The Commission’s initial Opinions on Applications for Membership and subsequent Reports on Progress can be found at http://europa.eu.int/comm/enlargement.

(3) Commenting on the initial Opinions, Nowak (1999) notes that their analysis of human rights “seems rather superficial and relates more to the de jure than the de facto situation” (p 695).

(4) While the conditions in Romanian orphanages may be of particular concern, excessive institutionalisation and the factors associated with it are a region-wide problem in the CEE countries (UNICEF, 1997). Bulgaria, rather than Romania, had the highest rate of infants aged 0–3 in institutions in 1997 (over 1%) and the growth in institutionalisation among young children has been largest in Estonia, with a rise of 75% over 1989–97, followed by Latvia with a rise of two thirds (UNICEF, 1999, Annex Table 8.1).

(5) Other rights relating to children are mentioned in the reports, although again in haphazard fashion. Several reports refer to issues relating to juvenile justice. The 1999 report on Poland raises the need for policies to address violent and abusive treatment against children, while the Slovenian report notes that provisions on the rights of children are included in the Constitution. The original Opinion on Lithuania highlights the absence of appropriate legislation against child prostitution or the sexual abuse of children. Children also come up in discussion of minority rights (eg, with regard to discrimination against Roma and to the question of citizenship for ethnic Russians in the Baltic states).

(6) The periodic reports of EU and CEE countries to the UN Committee on the Rights of the Child, which monitors the CRC, can be found at http://www.unhchr.ch/html/menu2/6/crcs.htm. Issues of human rights within the EU are discussed at length in Alston (1999). (See in particular the introductory chapter by Alston and Weiler.) In extremis, the Treaty on Union permits the suspension of benefits of membership in the case of a member state committing a ‘serious and persistent breach’ of human rights.

(7) The Commission’s Sixth Periodic Report on development of EU regions extends the first cohesion report’s analysis to the CEE countries (European Commission, 1999).

(p.124) (8) The convergence of polices affecting children within the EU member states is considered by a variety of authors, including Ruxton (1996) and Ditch et al (1998).

(9) Of course, greater employment opportunities may help reduce teenage motherhood, but the strength of the local labour market is far from being the only factor involved (the issues are reviewed in Micklewright and Stewart, 2000).

(10) We focus exclusively on cash incomes in this section but the EU definition is a reminder that ‘poverty’ can be the deprivation of things other than money. On the issue of a common EU line, see Atkinson (1998) who explores the idea of a weighted average of common and national thresholds.

(11) The difference between the coefficients on the CEE and EU dummies was significant at the 2.5% level. The data were drawn from the Human Development Report.

(12) The same source suggests that the rate for Slovenia in Figure 4.5 is much higher than for other years in the 1990s.

(13) Hammarberg (1998) discusses in detail the shape of educational policy based on the principles in the CRC.

(14) Results of the TIMSS follow-up and of the new OECD assessment are both expected to be available in 2001 (see http://timss.bc.edu and http://oecd.org/els/pisa).