Introduction

Educational policies and practices have the possibility of either reproducing social structures, or of changing them. If a society has substantial and persistent inequalities – whether of the distribution of wealth, or of recognition of rights, or of access to social provision, or of recognition of culture or language – then it is possible, indeed probable, that educational practices will replicate these inequalities. In 2006 a group of seven European UniversitiesFootnote 1 were commissioned by the European Commission to explore why many educational policies that were designed to address issues of social inequality appeared to be failing to achieve this. This chapter derives from that study – Educational Policies that Address Social Inequality (EPASI) (Ross et al. 2009), and analyses some subsequent changes in policy and practice.

Three principal arguments are generally employed – with differing emphases and priorities – for educational policies to be directed at minimising social inequalities. The Commission’s commissioning brief used all three. Firstly, there are social reasons (the Commission’s brief suggested ‘an harmonious education’: European Commission 2006, p. 5). If different social groups feel they are treated less equally than others, in terms of their access to social, economic and cultural rights, societies are likely to be less cohesive, and some groups may be less likely to participate in civic behaviour, believing that they will not be listened to. But education is only one of the potential agents for such change (Wilkinson and Pickett 2018). Secondly, economic justifications may be advanced (‘the importance of key skills for the development of knowledge-based economies’: European Commission 2006, p. 5). If groups fail to meet their potential, then there is a wastage of human capital. But activities that exclusively focus on education to increase economic competitiveness tend to reproduce and frequently increase inequalities. Thirdly, there is the human rights and equity argument (‘provid[ing] everyone with a high-quality education’: European Commission 2006, p. 5). Respect for the rights of all requires recognising, as far as possible, differences between individuals and groups and minimising the differential access to rights that society may impose – not just political and civil rights, but also social, economic, cultural, religious and linguistic rights.

The Commission’s brief for this research was to identify policies that lead to equality of educational outcomes, not mere equality of opportunity: they expected education to have the power and potential to transform social outcomes. It suggested that there were (unspecified) groups ‘at risk’ of underachievement, and that the study should focus on programmes that systematically targeted such groups – not individuals – through the distribution of resources and programmes, at both local and national policy levels, including non-governmental organisations. This was a recognition that teaching and learning took place in a variety of settings, not simply within educational institutions, and that these also required analysis.

This chapter builds on the conclusions of this study, and analyses changes in the subsequent decade, focusing firstly on four areas of inequalities: (socioeconomic disadvantage, ethnicity, migrancy and gender) and the intersectionality between these areas. Secondly, we focus on changes in the policy discourses and practices in three of the states in the study (Greece, Sweden and the UK). We add some discussion on issues related to migrants and second language learning, based on subsequent research. Our re-evaluation of policy changes in these countries and areas directly addresses the relationship between ‘the public good’ and educational policies that prioritise equity over efficiency.

The research strategy

Our strategy in 2006–2009 was to carry out three parallel investigations. We focused on seven groups of those potentially disadvantaged (who might or might not be equally disadvantaged in each country for a range of reasons). We recognised that the conceptualisation of social difference varies between countries.

Seven thematic reports were produced that focused on:

  1. 1.

    Socioeconomic disadvantage: where a significant marker of educational underachievement is family poverty, but economic disadvantage alone does not explain all social disadvantage, and other characteristics intersect with this and must be employed to explain the institutionalisation of disadvantage and discrimination.

  2. 2.

    Minority ethnic disadvantage: groups experiencing racism and other disadvantages include those of long-settled migrant origin, more recent migrants, refugees and asylum seekers. In some countries, identifying ethnic groups is in itself regarded as racist; other states hold that racism can only be challenged by identifying these groups, and then targeting provision and monitoring achievement. Some states identify ethnic minorities that have settled in the country for several generations as ‘immigrant’, even though there may no longer be any meaningful association with the country of origin.

  3. 3.

    Gender: an area of deeply-ingrained cultural attitudes that lead to different social expectations of roles, and hence to discriminatory and disadvantaging practices. Stereotypical behaviours can lead to gendered practices in educational provision and expectation, impacting on attainment, subject choice and future employment. Under the term gender we also consider discriminatory behaviour related to sexual orientation.

  4. 4.

    Indigenous minorities: including Europe’s longstanding indigenous minorities, such as the Roma, Sinti and Sámi.

  5. 5.

    Disability: in that data suggests that this is a disadvantaged group, and the term covers much more than those with impaired physical abilities.

  6. 6.

    Linguistic minorities: where a dominant language marginalises and discriminates against linguistic minorities, even when these are long-standing and widely spoken in particular regions.

  7. 7.

    Religious minorities: where the relationship between religion and education has led to particular structures and expectations about the role and place of religion in state education.

A series of country policy reports were produced; these analysed each state’s specific educational policy discourses: Belgium (Flanders), Cyprus, Czech Republic, Denmark, France, Greece, Ireland, Luxembourg, Malta, the Netherlands, Slovakia, Spain, Sweden and the United Kingdom (UK) (largely England). In each of these we analysed recent educational initiatives, producing nearly 300 short Project Studies on these.

The issues of how and why these various categories of disadvantage were identified and named (and are thus socially constructed) have consequences in terms of how people have a sense of themselves, and this became significant in the organisation of our research strategy. The reification of a category may have the potential to empower a community, but may also be used to shift responsibility onto the group members to solve ‘the problems’ for themselves. Thus, the neoliberal offer of a ‘choice’ can shift the onus for change to a group who may not be in any sense responsible for, or able to address, wider structural and attitudinal causes. Further, intersectionality is a critical factor in understanding the multiple identities and categories that arise: For example, in Sweden, Alireza Behtoui (2006) showed the intersection between ethnic background and social background, the former being used in popular discourse to discount the effect of poor living conditions: social class acquires an ethnic face. Many groups suffer educational disadvantage through multiple aspects – for example, being poor, members of an ethnic and religious minority, and speaking a different language to that of the majority of the population: each may contribute to overall disadvantage in a different way, and it is analytically useful to identify how marginalisation is identified and created through categorisation.

Our focus was on inequalities between groups (rather than individuals) and social structures, rather than personal attributes, but decisions about educational engagement are, in part, the consequence of individual actions. Some of these will be made by educators (such as advising and guiding pupil choices, streaming, etc.) and policy makers (such as determining types of school, subjects and examinations). Others will be made by the individual or their families, some of whom will be aware of their family’s history of educational non-success, fear of failure, and the potential costs of such a decision. Such decisions may be risk-averse, though rationalised in a discourse about further study ‘not being for people “like us”’ (Archer et al. 2003, p. 178). Richard Breen and John Goldthorpe (1997) argued that young people showed ‘relative risk aversion’, and their goal in schooling was to acquire a level of education enabling a class position at least the same as that of their family, avoiding downward mobility. Breen (2001) extended this, arguing that educational career decisions arose from pre-established family decisions about attaining a particular educational threshold and beliefs about the probability of educational success. None of this denies individual agency, but it does recognise the powerful structural and cultural limiting constraints.

Our research questions

Throughout all aspects of the work, we kept the following research questions before us:

  • What educational policies have been used to combat group social inequalities?

  • How have policy makers identified and analysed these inequalities? (Systematically or reactively? Considering all possible causes?)

  • Was the policy initiative focused on the group, or on wider society?

  • How were the subjects of these intervention programs targeted and resourced?

  • Were groups themselves consulted and involved in these policies?

  • Were programmes implemented to sustain the policy? Were changes embedded in professional practice?

  • Were policies national or local in their design? Was there opportunity for local initiative? Did local actors have ownership over policies or programmes?

  • Were policies evaluated and monitored (and if so, how)? Was this systematic and independent? Did it feed back into policy-making?

  • How can the project inform future policy development?

Intersectionality and inequalities

Throughout our analysis there were many instances where there are combinations of factors that are seen as responsible for particular inequalities, where two or more of the various factors have intersected with each other to cause greater, and more complex, inequities, that are multifaceted and more difficult to address. In the analysis above, we have shown examples of socioeconomic disadvantage interacting with ethnicity, and with gender, and with other factors such as minority language use and disability.

Intersectionality provides an analytic frame to address this. The term was originally used by Kimberlé Crenshaw (1989), to describe the intersection of gender and race in the US legal case of DeGraffenreid vs General Motors. Emma DeGraffenreid’s case was that General Motors had factory floor jobs available for Black men, and office jobs available for White women: Black women were thus unemployable. The case was lost, as the court ruled that Black women could not combine their race and gender claims into one: it was asserted that as they could not prove that their experience was the same as what happened to White women or Black men, the discrimination against them could not be considered. As Crenshaw (2015) wrote much later,

as a young law professor, I wanted to define this profound invisibility in relation to the law. Racial and gender discrimination overlapped not only in the workplace but in other arenas of life; equally significant, these burdens were almost completely absent from feminist and anti-racist advocacy. Intersectionality, then, was my attempt to make feminism, anti-racist activism, and anti-discrimination law do what I thought they should — highlight the multiple avenues through which racial and gender oppression were experienced so that the problems would be easier to discuss and understand. (Crenshaw 2015, on line)

Leslie McCall (2005, p. 1171) describes intersectionality as ‘the relationship among multiple dimensions and modalities of social relations and subject formations’, encompassing ‘perspectives that completely reject the separability of analytic and identity categories’ (2005, p. 1171, fn 1). While Crenshaw’s original axes of identity used in intersectionality in 1989 were those of gender and race: ‘intersectionality has broadened to encompass a number of additional social factors — sexual orientation, nationality, class, disability and others’ (Emba 2015).

But intersectionality is not just about identities but about the institutions that use identity to exclude and privilege. The better we understand how identities and power work together from one context to another, the less likely our movements for change are to fracture’ (Crenshaw 2015). Social oppression is not the consequence of these various factors acting independently, but their intersection creates multiple forms of oppression and discrimination (Ritzer 2007). The examples we have highlighted above show the educational institutionalisation of discrimination through multiple intersections of inequality related factors. This suggests that policy programmes to address particular educational inequalities will need to be finely tuned and focused to recognise, acknowledge and address these intersections in their programmes and the evaluation of their outcomes.

Changes in four areas of inequality

We now consider four particular areas of inequality – socioeconomic status, ethnicity, migrancy and gender, looking at the situation when we first reported in 2009 and the changes in the decade that followed.

Inequalities in socioeconomic status

In our thematic report on socioeconomic disadvantage (Cederberg et al. 2009), education was emphasised as a way of compensating for structural factors of socioeconomic disadvantage. Variables such as parents’ educational level, social class, social heritage, gender, ethnicity, living conditions, and the risks of poverty were discussed. For example, we described how, in Sweden, by 2002, migrants from the Middle East and North Africa had an unemployment rate that was four times greater than that of those who were Swedish-born, and their self-sufficiency level was only 30% of the level of those who were of Swedish heritage. But ethnic background is not a homogeneous category, with substantial differences between and within ethnic minority groups. Emphasis on ethnic background might conceal the effects of class, when this could have been foregrounded. Social class acquired an ethnic face, as Bolette Moldenhawer (2001) argues. We concluded that socioeconomic disadvantage and segregation needed to be discussed ‘in relation to diversity, institutional discrimination, and the complex interplay between the educational system, individuals, groups, and the surrounding society’ (Cederberg et al. 2009, p. 13). In 2009 on average 19% of children within the European Union (EU) were defined as poor and 10% of all children lived in households with no one employed. Of all these children, 60% lived within what the European Commission defined as the poverty zone. Children of working-class families are, of course, not automatically poor. However, they may encounter other forms of disadvantage when they come across the upper- and middle-class hegemony characteristic of many educational institutions. Pedagogic discourse is sometimes constructed to consistently favour middle-class groups, neglecting the experiences and communication styles of others, and thus negatively impact on these groups (Bernstein 1993; Skeggs 1997). In official reports and documents diversity was often expressed as both a resource and a problem – but with an emphasis on the latter. There were seldom identifiable examples or explanations of any substance as to why diversity per se could be positive.

Issues of socio-economic inequality arise because of social practices outside the school setting, and the direct control of educational policy, yet nevertheless have a profound effect on the school’s potential to effect changes in teaching and learning. For example, developments in learning that require access to electronic technologies, whether of the availability of hardware or of high-speed internet access, serve to reinforce social inequalities. Schools in less privileged social locations may have informal policies to make less use of such learning, because of the inequities this will highlight across their school population, thus disempowering even those students who do have such access.

Since our 2009 report was completed, all the countries involved have undergone a period of economic upheaval following the collapse of Lehman Brothers bank in late 2008. The consequences of this, unrolled over the following months and years, was the European Debt crisis that has impacted – sometimes dramatically - on the economies, and thus the level of socioeconomic disadvantage, of all the countries in our study over the decade 2009 to 2019. The most significant impact was on Greece, whose large structural deficit and level of international debt left it particularly vulnerable. There were severe cuts in governmental expenditure, and a series of substantial loans made by the EU, the International Monetary Fund and the European Central Bank. Unemployment rose to 28% by 2013, and youth unemployment to 62%. Very little of the loans went to support government expenditure; most was used to refinance existing loans held by private banks. The economy was not declared by the European Commission as restored until 2018. Sweden, outside the Eurozone, fared much better: its floating currency rate gave it a short-term advantage, and structural reforms and constraints, coupled with labour concessions, tax reform, and a low level of public debt allowed for a swift economic recovery. The UK, also outside the Eurozone, began to make a similar recovery, but the May 2010 election brought about a coalition government (Conservative, with Liberal Democratic support) bent on reducing the overall state share of the economy. They claimed that the deficit recovery policies of the previous government had increased borrowing levels, and introduced dramatic cuts in public services to ‘reduce the debt’ they had inherited (UK Government, 2010, pp. 15–16). The planned five-year programme for debt reduction was extended a for a further four years in 2015 (Conservative Party 2015, pp. 7–9).

The consequences of these changes in the various countries were that by 2017 the percentage of children and young people (0–17) who met the EC criteria for being at risk of poverty or social exclusion was 19.4% in Sweden, 27.4% in the UK, and 36.2% in Greece (EU-28 average 24.4%) (Eurostat 2019a). The severe material deprivation rate, as defined by Eurostat, in each of the three countries in 2016 was 0.8% in Sweden, 4.0% in the UK, and 22.5% in Greece (EU-28: 7.5%) (Eurostat 2019b). The OECD, reporting on the socioeconomic divide in Europe in January 2017, concluded that income inequality was at an all-time high; female unemployment was 9.8% greater than it was for men, and their earnings were 12.8% lower; low-skilled youth who were not in employment or education comprised 17% of 15–29-year-olds in the EU and at risk of permanently being ‘left behind’ in the labour market; and significant gaps in educational outcomes depending on parental socioeconomic background remained: a child from an advantaged background scored an average 20% higher in mathematics than one from a disadvantaged background (OECD 2017a). Immigrants tend to have lower outcomes in terms of labour market or incomes than the native-born in most areas, and those who were employed were twice as likely to live below the poverty line, and the youth unemployment rate for immigrant groups was almost 50% higher (OECD 2017b). Young people at risk of leaving school early was a particular issue in the 2010s (Ross and Leathwood 2013).

Inequalities in ethnicity

Educational inequalities in relation to ethnicity are widespread, but this is a complex area. As we noted in the original study (Williams et al. 2009), debates about race and ethnicity are framed differently across Europe, with differences between countries in terms of conceptualisations and definitions as well as different policies on data collection and monitoring. This makes any comparable assessment of the forms and extent of educational disadvantage for minority ethnic groups difficult. This remains the case: Lilla Farkas (2017) observes serious shortcomings in data on racial and ethnic minorities across Europe, with proxies such as immigration status, language, nationality or religion sometimes being used instead of concepts ethnic origin, despite these being the focus of the EU’s Racial Equality Directive (European Council of Ministers 2000). Indeed, as we noted in 2009, in some countries there is a reluctance to name ‘race’ or ethnicity as the basis for any educational or social disadvantage, reflecting a desire to avoid debates about racism. Where minority ethnic disadvantage is evidenced, there was a tendency to explain this (away) by reference to language skills, parental educational background and/or socioeconomic issues. For example, politicians in Sweden cited an inability to speak Swedish as the main cause of minority groups’ educational disadvantage and we argued that, ‘the cross national focus on immigrants’ socioeconomic and linguistic status constructs them as citizens in training who only need to acquire the right outlook and skills to gain the full citizenship afforded to the native population’ (Williams et al. 2009, p. 6).

Nevertheless, in the 2009 study we were able to identify educational disadvantages that were experienced by particular minority ethnic groups in certain contexts. For example, we identified reports referencing that while differences in literacy levels between native and first generation migrant children were widespread across Europe, differences between native and second generation migrants were much less in the UK and Sweden, and somewhat greater in Germany and Austria. In Greece minorities such as Roma, repatriates, immigrants and members of the Muslim minority of Thrace were most likely to be identified as functionally illiterate. In Sweden, more than 40% of first generation migrant students performed below level 2 in maths (having only basic maths skills), compared with a small percentage of those of Swedish heritage. For those countries where data was available, minority groups were less likely to complete compulsory schooling and less likely to reach the standard measure of attainment than ‘native’ students. However, not all ethnic minority groups were found to be educationally disadvantaged and there were also notable differences within groups. Furthermore, some minority ethnic groups outperformed their majority ethnic peers, for example the children of Chinese and Indian heritage in the UK.

Where data are available, there is evidence of progress for some minority ethnic groups over the twelve years since our original research. For example, educational attainment for Black and minority ethnic young people in the UK is improving, with students from almost all minority ethnic groups making faster progress on average than the majority White group (Morris 2015; UK Cabinet Office 2018). But the EU-MIDIS II minorities and discrimination survey (European Union Agency for Fundamental Rights 2017) found that Roma children across Europe still lag behind their non-Roma peers on all education indicators.

Furthermore, educational disadvantage and inequalities extend far beyond that evidenced by performance measures and outcomes data. In 2009 we reported the intensification of negative attitudes particularly towards refugees and new migrants, and social segregation was seen as a matter of concern, exacerbated by ‘White flight’ from multi-ethnic areas as one consequence of the move towards increased parental choice of schooling across Europe. Racist bullying and social exclusion affect many minority ethnic groups across different national contexts, with racist practices in education manifested in low teacher expectations, stereotypes and prejudicial attitudes, and harsher sanctions to those from particular ethnic groups.

If anything, overt racist attitudes and behaviours across Europe seem to have worsened since 2009, as the monitoring reports of the Council of Europe’s commission on racism and intolerance in Greece, the UK and Sweden (European Commission against Racism and Intolerance [ECRI], 2015, 2016, 2018) indicate. Increased racial hatred and violence against immigrants, Roma, Jews and Muslims was noted in Greece; in Sweden, rising incidences of racist and homophobic hate speech directed at migrants, Muslims, Black persons and Roma were reported, with antisemitic hatred remaining a problem; and in the UK, intolerant and xenophobic discourse from some politicians and the tabloid press has added to a climate of fear, with a sharp rise in anti-Muslim violence and the highest level of antisemitic incidents ever recorded in 2014 (ECRI 2015). A study of Eastern European young people living in the UK found that half of the participants reported an increase in racism and being bullied since the Brexit referendum in 2016, with many not reporting such incidents as they thought that neither teachers nor the police would be interested (Sime et al. 2017), and the scale of racism in UK higher education has recently been highlighted (Batty 2019).

Against this backdrop, there have been a number of policy initiatives in member states and across Europe to address ethnic inequalities in education, although much of the focus in recent years has been on first generation migrants rather than on the disadvantages experienced by those minority ethnic groups with long histories in the country. There also remains a reluctance to mention racism as a factor in educational inequalities – something that appears not to have changed since 2009.

Inequalities in education for migrants and those of migrant background

Since our work in 2009, there has been additional analysis of educational equality policies relating to the educational rights of migrants and their families. Shortly before the conclusion of the EPASI project, IPSE was asked by the Migration Policy Group, based in Brussels, to scope the possibility of adding educational criteria to evaluate different States’ policies on the education of migrants, as part of their Migration Policy Index (MIPEX) longitudinal programme.

IPSE (Hollingworth and Ross 2008) devised indicators for analysing whether each state encouraged children of immigrants to achieve and develop in school in the same way as the children of their own nationals, including whether:

  • migrant children (whatever their legal status) have equal access to all levels of education;

  • the specific educational needs of migrant children (and their parents and teachers) were targeted;

  • the new opportunities immigration brings to schools (such as experience of diversity, exposure to new languages and cultures) were used to benefit all pupils; and

  • all pupils and teachers were supported to learn and work together in intercultural education.

Twenty-six indicators were selected across these dimensions, drawn from some 59 conventions, resolutions, recommendations, advice and goals from the Council of Europe, EU, the International Labour Organisation, OECD, the UN and the UNESCO. Each indicator was assessed for each state in 2010 by two national education policy experts. These initial results were then published by the Migration Policy Groups and the British Council (Huddleston et al. 2011) as an overall comparative evaluation of migrant integration. The study has been repeated, with a second analysis published in 2015, and a further analysis is imminent.

Table 13.1 shows the findings for the three states we are considering in this chapter, over the period 2010–2014. The proportion of first and second generation migrants in all three states was very similar (between 13% and 15%). Greece, 21st of 38 States in 2010, did not change in its provision over this period (which concluded before the significant arrival of Syrian and other refugees in Greece in 2015): the country, already in deep in a monetary crisis, did not worsen its policies on migrant education. The UK, on the other hand, slumped very significantly in 2011, with the arrival of a Conservative coalition government determined both to cut overall public expenditure and to ‘create a hostile environment’ for immigrants (Yuval-Davis et al. 2018; York 2018). The overall score fell from ‘slightly favourable’ to ‘half-way favourable’), and the policy dimensions of Targeting Needs and New Opportunities fell very significantly: the UK’s score was only maintained as high as this by its ‘Intercultural Education for all’ score. Sweden, on the other hand, maintained its overall leading State position, and raised its Access rating in 2013.

Table 13.1 MIPEX results for Education Policies for Migrants, 2010–2014: Greece, Sweden and the United Kingdom
Full size table

Inequalities in education and gender

Gender equality in education is central to rights of access, participation, recognition and being valued. In our original report (Spinthourakis et al. 2009) we pointed to the different contexts, funding and implementation models within which educational polices on gender were implemented. All the countries studied demonstrated commitment to the principles of gender equality and non-discrimination, but few programmes were complete, especially in their evaluation, monitoring and dissemination (ibid., p. 16). Without such information the usefulness of these projects is significantly reduced.

Each country had its own history of addressing gender inequalities. Most state policies had a position on gender-related disadvantage, and projects were designed essentially from a structural viewpoint. The major issues, regardless of differences in policy, were career and subject choice and fighting gender stereotypes. Combating stereotypes was found throughout, and in the area of gender and attainment, boys’ underachievement was often an issue in policy discussions. Many felt that this was used to conceal deeper issues of gender inequalities and outcomes (Arthur and Davies 2010). The absence of policies addressing sexuality in most countries suggested that they were not a priority over the decade.

Since 2009 approaches to gender inequality have in some cases remained static, and others have fluctuated, not always in ways that might have been foreseen. While gender equality is a moral, human rights and justice issue, it is now argued (and increasingly researched) in the 2010s that it has a potential economic cost (Klasen and Minasyan 2017).

The language of gender and sexuality is evolving rapidly, and the diversification of terminology allows greater identity self-determination for some, but is objected to by others, and this has the potential to increase social inequalities (Dunne and Hewitt 2018). Gender equality may have more prominence in legislative texts, but this does not always equate to the achievement of equity. Malcolm Brynin et al. (2019) examined intersectionality between gender and ethnicity, finding diversification of inequality but also that there is more inequality between different ethnic groups than between gender groups.

The Pew Research Center (2010) referred to ‘gender equality’ as being ‘universally embraced’, but questioned whether this had been translated into action. The last decade has shown progress in terms of achieving greater general gender equity, as measured by the European Institute for Gender Equality [EIGE] (2019, Table 13.2). EU Commissioner Jourová commented on this, saying ‘[we must still take] positive measures to address inequalities between women and men and by tackling all forms of discrimination in our society.’ (European Commission 2017, p. 5).

Table 13.2 Gender Equality Index scores for selected EU Member States, 2005, 2015, 2017
Full size table

Gender policies in EU states have had a positive effect especially for those states with little or no tradition in this area (Lempesi 2019, p. 42), and the reasons for the variations in change between states are complex. Maria Karamessini and Jill Rubery (2020) suggest the UK’s policies lack coherence, probably due to unwillingness to acknowledge the negative impact austerity has on general gender equality, while Greece had attempted to create coherent gender mainstreaming, but lacked the resources for implementation. Sweden’s efforts and commitment led to the best results in the EU-28.

Explanations of inequality and policies of evasion in different areas of inequality

Responses to educational inequalities have been varied, at both societal and governmental level. The general discourse of meritocracy is prevalent: a particularly insidious argument that implies that those who do not succeed are themselves responsible for any disadvantages they suffer, discounting institutional and structural impediments. ‘Pure meritocracy is incoherent because, without redistribution, one generation’s successful individuals would become the next generation’s embedded caste, hoarding the wealth they had accumulated’ (Giddens and Diamond 2005). Frank Walkey et al. (2013) found that ‘promoting low or even moderate expectations and aspirations for student achievement may actually reinforce lower academic achievement’ (p. 306). Varieties of explanation for inequality include pathological explanations (including discredited assumptions that intelligence is largely genetically determined); transmitted deprivation (e.g. due to perceived lack of parental education or skills); home-based factors (lack of material resources, etc. in the home); school factors (such as lack of resources or low teacher expectations) and the structure of society (e.g. social class and socially differentiated schooling). All of these have implicit and explicit implications for policy. Some anachronistic explanations are still employed at the policy level, as well as in popular discourse, often without challenge.

We found four particular policy responses that act to undermine and evade the achievement of equality of educational outcomes.

The denial of the existence of disadvantaged groups in, for example, countries where populations asserted a ‘national identity’ that was reluctant to recognise that the population was not homogeneous, and some groups that were being characterised as ‘the other’.

Another response was to confuse categories, subsuming all inequalities under the general category of the (socio)economic. Clearly, family poverty does have a strong impact on educational attainment and participation. But there are other inequalities that intersect with and compound poverty. It may be simpler for governments to attribute inequalities to poverty than to acknowledge more complex patterns of discriminatory behaviour towards disadvantaged groups.

Equality policies may sometimes compete with other policy agendas: as Nancy Fraser (1997) observes, affirmative action policies may not address deeper structures of inequalities, and then, inequalities will persist. The development of an audit culture in education, where schools and teachers are rated according to the successful outcomes of their pupils, may lead to unintended or perverse outcomes. If schools or teachers are judged by the proportion of pupils achieving a particular standard, then they will be tempted to concentrate attention and resources on those pupils who are most likely to move through the threshold to achieve the standard, focusing on pupils in a narrow ability band just below the threshold, thereby neglecting others.

Some policies do not address equality of outcome, even though they have this intention. It is common to focus resources and attention on members of a specific underachieving group, without considering the wider social and teacher expectations that may be leading to underachievement, shifting responsibility for success to the individual, assuming that all that is necessary is to provide ‘equality of opportunity’, and to stigmatise individuals who fail to take advantage of such ‘opportunity’. Widespread assumptions about social and economic roles shape curricular options, which may lead to lower levels of resources, a limited curriculum, and low teacher expectations.

We now therefore turn to the different policy discourses in different states, and examine changes in these between our analysis in 2009 and today.

Changes in policy discourses about inequalities in education in three states

We now examine policy responses to inequalities in three of the countries we originally studied – Greece, Sweden and the UK – again, reviewing the situation at the time we first reported in 2009 and the changes in the decade that followed.

Policy responses to inequality in Greece

Greek state educational policies aimed at eradicating social exclusion and educational inequalities have historically been contentious (Kazamias 1967). Educational policies are often announced with a high degree of promise, yet are slow to materialise (OECD 2018, p. 14); those that are implemented tend to either gradually contract or simply end due to a lack of institutionalised funding. Three-quarters of policy initiative funding is from European Structural Funding, but this requires the balance to be contributed from national funds, which are not available to maintain programmes in the longer term (Eurydice 2019).

In the last decade policies have been introduced to deliver ‘equity, quality and efficiency’ through the national educational system to all people, many of whom would otherwise be marginalised. In 2008 we highlighted how educational policy reform in Greece has been characterised as a ‘reformist reform’, rather than being attuned to structural reform: the educational reformism foci ‘have been the modernisation and democratisation of what was believed to be an anachronistic and deficient educational system, one that was extremely centralised and bureaucratised, economically inefficient, socially exclusive and inequitable, and pedagogically authoritarian.’ (Spinthourakis et al. 2008, p. 5).

This earlier report described the Greek state as focusing directly on the issues to deal with educational inequalities. National policy pronouncements and programmes were intended to serve as catalysts of change, to be implemented at the local level. In hindsight, it can be seen that this marked the end of nearly a decade of apparent national prosperity, increased funding on social initiatives and the promise of a brighter future. That national trajectory, prior to the economic crisis, appears to have demarcated the end of one of its most vibrant and socially conscious periods of stability in contemporary Greece history.

The financial crisis shifted state policy towards efficiency, though not ignoring issues of equity (Tsatsaroni 2011, p. 4), but it was clearly aligned with austerity. As an example, the Kallikratis Programme (Greece, Ministry of Interior, 2010) – on the ‘New Architecture of Local Government and Decentralized Administration’ – combined efficiency and equity objectives. Schools were annually reviewed for attendance rates, teacher–student ratios, and ‘functionality’ (distance and difficulties children face travelling to school), and these measures were used to justify school mergers or closures (ibid., pp. 46–47). But vulnerable populations (e.g., the Roma, the Muslim Minority of Thrace) were exempted from this. Other policies allowed the expansion of Intercultural Schools, so all children could be enrolled, not just the culturally and linguistically ‘other’, in order to enhance intercultural education and foster cultural diversity (ibid., p. 49).

Special Education Needs policies were criticised in our 2009 report for reinstating an anachronistic medical orientation, but it nonetheless allowed SEN children to attend general classes, based on referrals from Centres for Diagnosis, Differential Diagnosis, and Support. Class teachers voiced concerns about funding, infrastructure, and training, and overall adequacy of this to serve SEN children (Pappas et al. 2018, pp. 4–5). New policies attempt to mediate these concerns, such as the co-teaching model of Parallel Support, where a special education teacher is assigned to a student and not to a classroom: unaffected by the economic crisis, this has been significantly implemented (Mavropalias and Anastasiou 2016).

These efforts to temper inequality have been undermined (OECD 2018; Andriopoulou et al. 2017): the crisis-related austerity measures often particularly impact on inclusion and equity policies (Mitrakos 2014). The 25% reduction in Greece’s GDP was comparable to being at war: a new group has been identified of ‘the newly excluded’, which includes members of the former middle class, previously perceived as privileged or rich (Zafiropoulou et al. 2017, p. 2). But these policy changes were not only attributable to the austerity measures, but must also be understood in the context of Greece’s increasing diversity, partly resulting from the refugee crises of 2015 and 2019.

Even under these conditions, efforts at amelioration are being undertaken, such as the three-year plan to achieve ‘higher equity’ in educational provision and outcomes (Greece, Ministry of Education, 2017). While long-standing challenges to equity are issues of lack of inclusiveness, geographic isolation and refugee status, even small villages have their own schools (OECD 2018, p. 104). Notwithstanding the Kallikratis Programme’s policy of merging or abolishing schools, 3.5% of primary schools and 6% of secondary schools are classified as geographically ‘difficult to access’ by the Ministry (Roussakis 2017). Another challenge, as yet without a clear policy, is the educational needs of the major refugee movement through Greece, which in 2015 was nearly one million (UNHCR 2017). Refugee facilitators in the camps are NGOs, but the refugees’ schooling is an issue for the Greek Ministry of Education. Reactions have been mixed, and local communities and schools need to be both prepared and willing to have refugee children enrolled in their schools to minimise segregation and foster integration into Greece society (Simopoulos and Alexandridis 2019).

The challenge of supporting policies to diminish social and educational inequality has been limited in scope, but can be seen in the general focus of the State’s policy initiatives (OECD 2018, p. 120). There may also be a need for a range of targeted approaches to be considered to eradicate educational inequalities, rather than such a generalised approach.

Changes in inequality in Sweden

Three Rädda Barnen [Save the Children] reports (Salonen 2018, 2019) investigated the poverty risk for families with children in Sweden. The first showed increasing poverty risk and the second that, though the long-term policy is to reduce child poverty, regional and inter-group differences remained. Many had missed out on the substantial increases in income that characterised Sweden’s economic growth after the 1990 crisis. The 2010 policy analysis of family economics was replicated by Salonen in 2019 to examine the extent to which policies equalised incomes and reduced family economic vulnerability. While families with children had increased income at all income levels up to 2017, differences between groups had increased. Using the EU relative poverty measure of less than 60% of the country’s median income, the proportion of families with children in poverty had risen by 2017 to 16%. Swedish family policy is now less able to counter the growing income gaps amongst families with children, and is less able to combat poverty (ibid., pp. 28–29).

Socioeconomic segregation contributes significantly to the emergence of segregation patterns. During the last ten years media reports on the state of Swedish schools have generated alarming concern. Children in families in financial need generally have lower grades and leave school earlier, and have higher risks of later developmental and psychosocial problems (Socialstyrelsen 2010). Parents’ educational level and income remain the factors that explain most of the grading results. Foreign-born pupils’ grades are also affected by the age of immigration (Skolverket 2018). However, for parents born abroad, parents’ level of education and degree of allowances have also increased in importance. Segregation challenges what used to be a Swedish ideal of the right to equal schooling (Gustafsson 2010). Segregation is relational: developments in different schools and residential areas interact with each other. A major problem has been the lack of evaluation, measurement and follow up on how segregation develops over time, which makes it difficult to know with any certainty if national strategies to reduce segregation actually work. Delmos [the Swedish Agency against Segregation] suggested there needed to be a cross-sectoral follow-up system based on appropriate indicators. The Government’s reform program for reduced segregation in 2017–2025 led to a new 2018 directive to Statistics Sweden to develop a nationwide socioeconomic segregation breakdown for statistical follow-up, carried out in collaboration with Delmos. This revealed signs of increasing segregation in a greater number of locations. Some housing areas now deviate significantly from the national average in terms of unemployment, school results, income, health, turnout and insecurity. The importance of more children taking part in the pre-school programme, and more young people completing secondary education have been identified as important factors for school results and of entering employment.

Skolverket (the National Agency for Education) is required to promote equal access to education and quality environments. The Agency’s report (Skolverket 2018) analysed family background and school results for all pupils aged 15–16 between 1998 and 2016, and found that socioeconomic background had become increasingly important for success in primary school. School segregation had increased and, with this, differences between attainment in different schools: Skolverket (2017) found that such school segregation increases the difficulties schools have in their compensatory and value-based work. Pupils with different backgrounds rarely meet in school today (Sernhede 2014). The consequence is that schools now offer fewer opportunities for children with a foreign background to encounter Swedish society (Sernhede 2011). This has particular significance for foreign-born pupils in families with lower levels of parental education and income. School segregation has increased as a result of residential segregation and freedom of choice reforms, and is increasingly structured around social and ethnic factors (Axelsson 2014; Wigerfelt 2014). Families with stronger resources and Swedish background tend to move away from low status areas and schools. Eva Andersson et al. (2010) showed that schools in vulnerable areas had very few or no pupils who spoke Swedish. Making active choices of school is one strategy for the families to deal with the structurally unequal conditions that apply within the education system. However, Nihad Bunar (2010) points out that in the context of free school choice and market-oriented competition between schools, most families do not choose to leave low-performing schools, as they value personal relationships that contribute to belonging and cultural recognition in the area where they live.

Skolverket (2018) reports that students with the same socioeconomic background receive higher grades if they attend a school with favourable socioeconomic composition compared to if they attend one with less favourable socioeconomic composition. Skolverket (p. 33) also identifies two reasons why the importance of family background has increased. Firstly, schools can no longer support students from poorer backgrounds and secondly, students’ home conditions have become more diverse, making compensatory action more difficult. Foreign-born pupils start school with poorer circumstances. This analysis also shows the significance of socioeconomic background has increased for pupils with Swedish backgrounds, but considerably less than for foreign-born pupils. Parental education level remains the most important factor, but increasingly family income is becoming the strongest driving factor behind the increasing importance of socioeconomic background, for foreign-born pupils and those with Swedish backgrounds.

Policy responses to inequality in England

Inequalities in educational participation, outcomes and experiences have a long history in the UK and persist despite government policy commitments to tackle disadvantage. Yet policies can make a difference, as our original study illustrated. At that time a Labour Government, committed to education and social justice, had been in power for over a decade. It recognised that economic disadvantage was linked with low levels of achievement, and that poverty and inequality had increased dramatically since the (Conservative) Thatcher Government’s election in 1979, with a third of all children living in relative poverty by 1997 (UK 1999). The Government committed to reversing these trends, with initiatives both to raise educational standards and achievement overall, and to target disadvantaged groups. Funding for education was increased, with school spending per pupil rising by over 50% in real terms between 2000/01 and 2010/11 (Belfield et al. 2018). Initiatives specifically designed to tackle disadvantage included ‘Sure Start’, a programme supporting learning, social and emotional development targeted at parents with children under four in disadvantaged areas; the Education Maintenance Allowance supporting students from low-income households with the cost of further education; and Aim Higher which sought to increase participation in higher education by young people from disadvantaged groups. Some positive outcomes for each of these initiatives were reported, with evidence that children from the most deprived backgrounds were beginning to catch up with their more advantaged peers, participation in further education increased (Heath et al. 2013) and more young people were applying to higher education (Passy and Morris 2010).

Other policies focused on inequalities related to gender and ethnicity. In 2007 the Gender Equality Duty made schools and other public sector organisations responsible for tackling gender equality in relation to achievement, career choices and bullying. As we predicted in our original study, a focus on boys’ achievement appeared be prioritised, with schools more likely to cite positive impact for boys than for girls (Bukowski et al. 2011). ‘Aiming High’, a project designed to increase attainment levels for African-Caribbean pupils, was credited with improving achievement for African Caribbean pupils in those schools that participated in the initiative (Tikly et al. 2006).

Our 2009 conclusion was that, despite a government commitment to tackling inequalities and disadvantage and some positive outcomes overall, these were limited by the ongoing reliance on neoliberal market economics and policies of choice, competition and meritocracy that had been introduced by the Conservative government in the 1980s and which continued to reinforce and reconstruct inequalities.

The political and economic context in 2019 is very different. The financial crisis of 2008 and the election of a Conservative-led coalition Government in 2010 committed to implementing what was described as ‘the most drastic budget cuts in living memory’ (Pimlott et al. 2010) has had a significant impact on the public good. Between 2010/11 and 2017/18, there was a real-terms reduction of funding for local authorities of 49% (NAO 2018). A report by the Equality and Human Rights Commission (2018) found that government changes to tax and welfare had resulted in more people in poverty, including children, disabled people, women and some minority ethnic communities. Between 2009/10 and 2017/18, total school spending per pupil fell by 8% in real terms. Some education policies were designed to tackle disadvantage, including extending the free childcare entitlement initiated by the previous Labour Government from 3–4-year-olds to disadvantaged 2-year-olds, but funding for Sure Start fell by almost 50% between 2010/11 and 2016/17 (NAO 2018) leading to the closure of up to 1000 Sure Start children’s centres by 2017 (Smith et al. 2018). A further policy designed to address disadvantage was the introduction of a ‘Pupil Premium’ in 2011, providing additional funding to schools in England to raise the attainment of disadvantaged pupils. Whilst this last initiative does appear to have encouraged schools to focus more on tackling disadvantage, in some cases this extra money was cancelled out by real terms cuts in school funding (NAO 2015). In addition, the Education Maintenance Award was abolished and replaced with significantly less generous funding for bursary awards only for students from the very poorest families, and tuition fees for higher education were raised from £3000 to £9000 a year from 2012. Although projects were designed to encourage more disadvantaged young people to go on to higher education, the emphasis was on the ‘brightest’ young people going to the elite universities (e.g.Thornton et al. 2014), reflecting the discourse of social mobility underlining exceptional individual success rather than any challenge to wider economic, social or cultural inequalities.

So those government policies that aimed to challenge educational inequalities continued to be undermined by an ideological commitment to neoliberal market economics and by the actions of a right-wing government and its commitment to a smaller state. Social deprivation still impacts strongly on pupil achievement and school exclusion. Boys do worse than girls at school and are more likely to be excluded, as are those with education support needs, Black Caribbean children and Gypsy, Roma and Traveller children. Bullying continues to be a problem, with over a third of girls reporting sexist comments online, and disabled children remain marginalised (Equality and Human Rights Commission 2019). High levels of racial and sexual harassment and sexual violence in higher education have been documented (Blue Marble Research 2018; Batty 2019). The Brexit referendum in 2016 produced further challenges. It is estimated that the UK economy would have been about 3% larger by the end of 2018 if Britain had not voted to leave the EU (Mason 2019), with increased xenophobia and a spike in police reported hate crime of almost a third in the run up to and following the referendum (UK Home Office 2018).

Conclusions

We can define three critical starting points for analysis. Firstly, that identifying disadvantaged groups will be difficult and probably imprecise, and involve intersecting factors; secondly, that the causal relationships between action and remedy will be complex and call for multiple and parallel programmes; and thirdly, that activities need to be directed towards both the disadvantaged and the advantaged, so as not to further ‘other’ disadvantaged groups (Kakos et al. 2016).

The recognition of a disadvantaged group has generally come about through the actions of members of those groups themselves and through identifying inequality of outcomes, and from this examining whether these may result from inequalities of opportunities. This has led at different times to the identification and definition of groups that may not previously have been recognised or conscious of themselves as groups. But data collection to demonstrate inequalities is not easy, particularly if there are issues in identifying members of a particular group: some groups may have concerns and fears about being identified. It may be important, therefore, to also use qualitative evidence of inequity.

It also seems critical to understand in all approaches that there will be no simple monocausal relationships between inequalities and programmes. It is very probable that no single programme will remedy all instances of a particular form of inequality; at the same time, almost every programme will successfully address some instances of inequality. To systematically address inequity, with the aspiration of leaving no individuals left behind, multiple programmes of action will be needed, including not only those that take place within formal educational settings, but also projects designed to address disadvantage in the wider community.

In many cases of forms of disadvantage, there is a strong case to work with the non-disadvantaged community as well as the disadvantaged. Tackling underachievement means raising expectations of success, and this involves everyone’s expectations, not just the expectations of the lower achieving groups. The assumptions of all professionals, policy makers, community groups and the public at large should be that all groups will achieve educational success.

Based on our detailed analysis of the project studies, we suggested (Ross et al. 2009, pp. 42–44) twelve general principles for action, that still seem useful in planning educational intervention projects.

  1. 1.

    Involve the disadvantaged community in planning, delivery and evaluation. Where communities are involved in the planning, management and evaluation of programmes, the chances of success seem to be higher. Recognising knowledge and experience, being culturally sympathetic, and empowering communities give them a hold over their futures (Cummins 1996; Henley 2006).

  2. 2.

    A strategic aim should be to raise the attitudes and expectations of everyone: inclusive programmes with elements variously addressed to more than just the underperforming group.

  3. 3.

    Institutionalise programmes so they support all practitioners. Highly differentiated and targeted programmes can lead to potentially isolated specialists, so most practitioners feel that particular pupils are ‘different’ and can only be supported by specialists, further isolating the target group.

  4. 4.

    Changes in educational programmes and policies take time to have an effect. It takes many years to educate a child, and more to change the whole teaching workforce. Programmes and expectations should be planned with this in mind.

  5. 5.

    Work with a range of agencies, at a range of levels, in a range of areas. Multi-agency working is more likely to produce coordinated action that reaches more pupils at risk, and approaches them with a variety of support strategies.

  6. 6.

    No single programme will remedy all instances of a particular inequality, but many programmes will successfully address some instances. Fixing on a single programme as the most cost-effective will leave some pupils outside the range of the programme.

  7. 7.

    Members of the minority groups should be part of the education professions. Few members of disadvantaged groups are represented: changing this will raise the aspirations and ambitions, and convey to the whole population that members of such groups are entitled to the same respect, rights and authority as the general population (Ross 2002).

  8. 8.

    Targets for who will be worked with, and what should be achieved, should be clear: identifying the nature of the difficulties, the areas to be particularly addressed, and the anticipated outcomes, help focus activity.

  9. 9.

    As far as possible, policy should be based on actual measures of achievement, take-up and need, rather than on proxies that are assumed to stand for these items.

  10. 10.

    Greater attention needs to be given, at national and European levels, to the collection of statistics on disadvantaged groups. The degree of precision may be difficult, but pragmatic efforts are better than none. Good qualitative data will identify the existence and extent of inequalities; help determine the distribution of resources; and help evaluate the success of any interventions.

  11. 11.

    Evaluate and learn from success: all projects should have planned, from the earliest stages, mechanisms to evaluate the activity. These should be both internal (continuous and iterative) and external (supportive but critical).

  12. 12.

    Successful programmes should become standard good practice. Mechanisms need to be in place, and resources available to allow this to happen. John Rawls principle was that resources should not be allotted on the basis of economic returns, but ‘according to their worth in enriching the personal and social life of citizens, including the less favoured’ (Rawls 1971, p. 107).

Education alone cannot reduce inequality, and policies in other areas – for example, housing in Sweden, and public sector cuts in the UK – can undermine and frustrate the impact of education policies designed to address inequity. Our arguments and investigations have been directed particularly at the third element of the European Commission’s rationale for the project: that educational policies should seek to minimise inequalities in order to ‘provide everyone with a high-quality education’ (European Commission 2006, p. 5). The research was based on this requirement for human rights and equity argument, which should be to minimise differences between individuals and groups that may result in differential access to rights – not just political and civil rights, but also social, economic, cultural, religious and linguistic rights. The ‘public good’ is, we argue, best served by educational policies – and educational research – directed at these enSds.