Interplay of factors that contribute to ESL at the system level

Thursday 23 July 2015, by Maša Vidmar

Aspects of the education system that concern the risk of ESL include the socio-economic segregation of schools, early tracking and grade retention. In that regard, protective aspects are high quality and accessible ECEC and VET. Well-managed transitions between educational levels that reflect a student’s changing needs in order to ensure the provision of a developmentally appropriate and engaging context are called for.

A relatively recent body of research identifies which factors at the system level contribute to or prevent ESL. The better insights into ESL so far gained show that ESL is a complex issue linked not only to the individual, the family and the school, but also to national policies in the educational, social, health and labour sectors. This article aims to review empirical research regarding ESL system factors with a focus on the education system, but also the labour market.

Evidence shows that the socio-economic segregation of schools and early tracking are linked to a greater risk for ESL. Grade retention is another risk factor, although timing (at the primary or secondary level) and opportunities provided during the retention year play a role. The transition from the primary to secondary level is problematic for following aspects: traditional style of academic teaching, lack of relevance of pathways and an excessively rigid curriculum, disconnection from labour market needs, lack of permeability between pathways (academic, technical, vocational). Together with the disruption in social relationships with adults, changes occur in the secondary school context that are developmentally less appropriate for early adolescents and may lead to disengagement and ESL. Thus, the school needs to adapt to a student’s increasing maturity and changing emotional, cognitive and social needs as they move through the education system. Moreover, access to high quality ECEC and divergent, relevant and high quality VET play the role of protective factors. The need for a coherent and well-balanced education system at all educational levels (i.e. all parts of the education system from preschool to tertiary education fit together well and function synergistically) is also warranted. The education system should be diverse, but not fragmented. Attention to the system blockages and discontinuities across different polices and sectors is needed.


As introduced in the two other articles on ESL factors (factors at the level of the individual, the family and social background, school-level factors), these can be viewed within Bronfenbrenner’s ecological theory of human development. This theory posits that five environmental systems interact at multiple levels with the individual: the microsystem (i.e. people and institutions in the immediate surroundings – family, peers, school), the mesosystem (i.e. relationships between the microsystems, e.g. home-school relationships), the exosystem (i.e. an indirect influence on the individual, e.g. health, social, media, neighbours), the macrosystem (i.e. cultural influences – attitudes, ideologies) and the chronosystem (i.e. events and changes over time) (Figure 1). In the two abovementioned articles that are related, we focused on the model’s inner circles (i.e. the individual and microsystem), whereas in this article we expand the review of ESL risk and protective factors to the outer circle, i.e. the more distant environmental influences located in the exosystem.

JPEG - 80.3 kb
Figure 1. Bronfenbrenner’s model of human development (adapted from Pinterest). The black rectangle constitutes the focus of the article.

Studies on ESL factors and processes have evolved over time. The first wave of studies mostly examined factors within the individual and their family (e.g. Ensminger & Slusarcick, 1992), with this being followed by a group of studies that also examined school-level factors (e.g. Lee & Burkham, 2003). The most recent wave of studies recognises that we still only have an incomplete picture of ESL (e.g. De Witte, Nicaise, Lavrijsen, Van Landeghem, Lamote, & Damme, 2013). Another level has thus been added to the research – the system level (e.g. in the extensive review of ESL factors by Rumberger & Lim 2008 this level was overlooked). The idea is to understand which national-level policies in different sectors contribute to or prevent ESL, and to highlight the need for coordinated policies. Namely, as insights into ESL increased, it became clear that ESL is a complex issue linked not only to the individual, family and school, but also to national policies. It is not only linked to policies in the education sector, but also those in the social, health and labour sectors, making the issue of ESL exceed the scope of school-level and education-sector policies and thus necessitating cross-sectorial cooperation.

The most recent research on ESL factors at the system level examines how different characteristics of various sectors counter or exacerbate the issue of ESL. This article aims to identify and describe certain features of education systems and their relations to ESL, including early childhood education and care (ECEC) and vocational education and training (VET) (see European Commission/EACEA/Eurydice/Cedefop, 2014). Features and measures emerging in other sectors (e.g. the labour market – see Audas & Willms, 2002; de Witte et al., 2013; Tumio & Taylor, 2013) together with inter-sectorial measures are also very important for tackling ESL and should not be overlooked. However, in accordance with the design of the TITA project, the article focuses on features of the education system.


The scientific review was conducted by examining related findings in relevant publications. These were found using computerised searches in the Arizona State University Library search engine (with 650 databases, including, PsycINFO, Academic Search Premier (EBSCOhost), ERIC (Proquest), JSTOR Arts and Sciences, ProQuest, SAGE Premier, Science Direct) and also in other online resources (e.g. ResearchGate, institutional or project webpages). We used the following keywords: “early school leaving”, “drop out” AND “system-level”, “ESL policies”, “policy interventions”, “education system” “early childhood education system and care”, “grade retention”, “vocational education and training” etc. In the next step, we examined references cited in the articles (i.e., “backward search” procedures). Original scientific articles and monographs as well as reports by or for the European Commission and the OECD are considered.

Education-system-related factors

Education systems take many different structural forms which can support educational achievement or create obstacles to it (EC, 2014b). In 2011, the EC identified several prevention, intervention and compensation system-level policies addressing ESL. Prevention measures are, e.g. good quality ECEC, prolongation of compulsory education, desegregation policies, systematic language support, and increasing the permeability or flexibility of educational pathways. Intervention measures include early warning systems, enhancing the involvement of parents, networking with actors outside school, teacher education, student-focused strategies, financial support. Compensation measures are e.g. second-chance programmes or recognition of prior learning. In a literature review (EC, 2014b), the following education system factors were identified: educational differentiation, length of compulsory schools, school segregation, investing in education, availability and quality of ECEC, curriculum, working conditions (e.g. salaries), teacher-student ratio etc. (also see Bridgeland, DiIulio, & Morison, 2006). In the present article, we review the following ESL education system factors: grade retention, socio-economic segregation of schools, early tracking, ECEC, transition to (upper) secondary education and VET, all of which are also highlighted in the European Commission/EACEA/Eurydice/Cedefop report (2014) and other review articles (e.g. EC, 2014b; Jimerson, Anderson, & Whipple, 2002).

Grade retention

Grade retention means a student is held back a year (i.e. repeats a year) due to their lack of progress, with this year giving them an opportunity to acquire the knowledge they need to continue schooling (European Commission/EACEA/Eurydice/Cedefop, 2014). The EACEA/Eurydice study (2011) showed that, even though grade retention is possible in most countries with similar basic regulation (restrictions), practices vary greatly between countries; in some countries, the rates are very low (< 3%) while in others they are very high (around 30%, e.g. Spain, France, Luxembourg). Similar grade retention rates were found 3 years later (European Commission/EACEA/Eurydice/Cedefop, 2014). It seems that grade retention is more common in countries where the view that this may benefit the student still prevails among teaching staff, the education community and parents (EACEA/Eurydice, 2011).

However, this assumption has been challenged in many studies. Janosz, LeBlanc, Boulerice, and Tremblay (1997) followed two samples of adolescents and found that grade retention significantly increased the risks of ESL in both samples. Roderick (1994) and Alexander, Entwisle and Horsey (1997) obtained similar results. Cairns, Cairns and Neckerman (1989) established a clear relationship between ESL and grade retention (i.e. students who were 1 to 3 years older than their peers in seventh grade were more likely to leave school within the next 4 years). This effect was more pronounced for Caucasian students than their African-American counterparts (ibid.). Jimerson’s (2001) metaanalysis of 20 studies showed no clear benefits of retaining students in grades for their academic, social-emotional or behavioural outcomes. When comparing retained students with a matched control group, 16 studies (80%) concluded that grade retention is ineffective (either nonsignificant or negative effects were found). The author also emphasised that, despite some positive short-term effects of grade retention for certain outcomes, the longitudinal studies show a strikingly close link between grade retention and ESL. In fact, another review study examined 17 studies and determined grade retention was one of the most powerful predictors of ESL (Jimerson et al. 2002). Moreover, grade retention was perceived to be the most stressful life event by sixth-grade students when asked to rate the stressfulness of 20 different life events (Anderson, Jimerson, & Whipple, 2005). A comparison of students who were retained and then subsequently became ESLers with those who were retained and then managed to finish high school showed that early socio-emotional and behavioural characteristics as well as maternal education and academic achievement distinguish the two groups of retained students (Jimerson, Ferguson, Whipple, Anderson, & Dalton, 2002). Early attention to a student’s social-emotional competencies is thus warranted. A more recent study showed that grade retention in eighth grade significantly increased the probability of dropping out of high school. However, retention in sixth grade did not affect the likelihood of completing high school and two-thirds of these students were able to catch up with the original cohort within 2 years (Jacob & Lefgren, 2009). Interestingly, authors have found no impact of grade retention in eighth grade for those who were sent to ’transition centres’ rather than remaining in school. Their findings offer some important insights into grade-retention policies regarding the timing of the retention as well as opportunities provided during the retention year.

Socio-economic segregation of schools

The segregation of schools refers to the accumulation of students from disadvantaged backgrounds in certain schools (European Commission, 2011). The socio-economic segregation of schools in the education system means there are large disparities between schools in their mean socio-economic status (SES), i.e. in the composition of the student body at a particular school in terms of SES. Socio-economic segregation occurs for different reasons. It can be due to the selection made within the education system (e.g. students from a disadvantaged background are over-represented in some types of schools) or may result from the tendency of specific social groups to live in certain areas (European Commission/EACEA/Eurydice/Cedefop, 2014).

Socio-economic segregation is the main factor underpinning segregation in European school systems (European Commission, 2011). It has been shown to be problematic in many ways. For example, the socio-economic segregation of schools is closely linked to the issue of (in)equity in education as those secondary school systems in which the schools are strongly segregated on average have lower achievement and greater variability in their PISA achievement; social background is a greater obstacle to educational success than in systems with less segregation among schools (OECD, 2008). Further, there is more problem behaviour in schools that have a concentration of socio-economic disadvantage (David, 2011). Another important issue is that the risk of ESL is considerably higher in schools with a low mean SES (Rumberger & Lim, 2008; Traag & van der Velden, 2008). Low school SES is often interwoven with other factors like a higher share of students with a minority/migrant background (e.g. OECD, 2008). Palardy (2013) showed there is considerable variation in high school graduation rates among schools within each socio-economic composition (SEC) category of schools (low, medium and high SEC schools). In some low SEC schools, all students complete high school while in others the figure is a low as 50%. However, despite this variation within each category, the graduation rates for high SEC schools were substantially on average higher than for low SEC schools. It should be noted that some studies (e.g. Lee & Burkam, 2003; Rumberger & Palardy, 2005) showed that, after accounting for a variety of other school characteristics (e.g. school practices), mean SES no longer had a direct effect on ESL rates.

Some countries apply active desegregation policies aimed at lowering segregation – they want to change the social composition of low SES schools, for example in Bulgaria and Hungary these seek to reduce the segregation of Roma students (European Commission, 2011). The range of mechanisms available for desegregation include, for example: controlling the entry conditions to individual schools, and positive discrimination measures in support of disadvantaged schools (grants, technical assistance, and other additional resources to schools) (ibid.; OECD, 2012). Attempts to increase the social heterogeneity of schools should always be accompanied by measures to ensure school quality (European Commission, 2011).

Early tracking

Tracking refers to the degree of educational differentiation in the education system – early tracking characterises systems with a high level of education differentiation that forces students to follow clearly defined paths early on in their education (EP, 2014b). Students are tracked into different educational pathways based on their achievement – a common practice in many European countries (European Commission/EACEA/Eurydice/Cedefop, 2014).

As with grade retention, the assumption is that this is beneficial for students’ learning – it is assumed that students learn better in a homogenous classroom at the level that is appropriate for their abilities. Yet this becomes problem when there is a mismatch between the track and a student’s potential, aspirations and interests (Hattie, 2009). Studies show tracking negatively impacts students assigned to a lower track and reinforces pre-existing inequities (a student’s SES matters more for their academic performance; de Witte et al., 2013; Gamoran & Mare, 1989; OECD, 2012). Course-level tracking is one factor explaining why the academic achievement gap associated with SES widens during secondary schooling (Caro, Cortina, & Eccles, 2015). Moreover, it seems that certain student groups (e.g. with a migrant background) tend to be more often tracked in the least academic tracks, possibly preventing them from fully developing their potential (European Commission/EACEA/Eurydice/Cedefop, 2014). The role of the teachers’ (low) expectations is also important in early tracking systems (ibid.).

Early tracking holds clear implications for ESL. Placement in the college track (i.e. academic) increases the chances of high school graduation (after controlling for selection factors; Gamoran & Mare, 1989). The type of vocational emphasis in school chosen when students are aged around 12 to 13 years plays a role in ESL as the selection of a school may limit easy access to the preferred intermediate vocational track and increase the risk of ESL (Beekhoven, & Dekkers, 2005).

School systems employing early tracking should postpone its use until upper secondary education while strengthening comprehensive schooling; moreover, opportunities to change tracks and provide high curricular standards for students in the different tracks lessen early tracking’s detrimental effects (OECD, 2012). Thus, permeable educational pathways are an important system-level protective factor against ESL (EC, 2011; European Commission/EACEA/Eurydice/Cedefop, 2014).

Early childhood education and care

The importance of high-quality ECEC is strongly emphasised in various EU documents. It is recognised that it brings a range of short- and long-term benefits for both individuals and society (European Commission/EACEA/Eurydice/Eurostat, 2014). For example, PISA 2015 showed that students who attended ECEC in their childhood outperform those who did not, even after accounting for SES (however, somewhere between two- to three-quarters of the advantage is reduced after accounting for students’ and schools’ SES; OECD, 2017). Generally, across European countries participation in ECEC is high during the year or two before primary education – on average, 93% children attend ECEC before they start primary education (European Commission/EACEA/Eurydice/Eurostat, 2014).

Tracing ESL back to (non-)participation in high-quality ECEC is a challenging task. The quality of early caregiving has been shown to be linked to ESL (Jimerson, Egeland, Sroufe, & Carlson, 2000). In a study carried out by the Public Policy and Management Institute (EC, 2014a), based on a literature review and case studies it was concluded that “in the profiles of pupils who are defined as ESLers and underachievers we see the absence of the very dispositions and skills that are laid down in high-quality early childhood education and care settings” (p. 76). This indicates that high-quality ECEC is a protective factor against ESL, particularly for migrant and minority students (EC, 2012; OECD, 2017).

Transition to (upper) secondary education

A well-managed transition process from the primary to secondary level, from the lower to upper secondary level, and from school to work are all factors helping to cut the ESL (European Commission/EACEA/Eurydice/Cedefop, 2014). However, research indicates this is now always the case, and in particular the following drawbacks of transition to upper secondary level are mentioned: traditional style of academic teaching, lack of relevance of pathways and excessively rigid curriculum, disconnection from labour market needs, lack of permeability between pathways (academic, technical, vocational) (ibid.). Smoothing the transition from primary to post-primary level was also one of the key themes raised in interviews with high-risk ESL groups (e.g. people with special educational needs, travellers) and parents of ESLers (Houses of the Oireachtas, 2010).

Other researchers have also problematized the transition to secondary education and the developmental inappropriateness of the secondary school context (e.g. Eccles, 2004). In her model, Eccles argued that the emotional, cognitive and social needs and personal goals of individuals change as they mature. The transition from primary to secondary level needs to reflect these changes and provide a developmentally appropriate context that will continue to motivate and engage students; she called this stage-environment fit. The author argued that the extent to which this does not happen leads to the first psychological and later physical disengagement from school, which ultimately is what ESL represents. She emphasised the following discrepancies that are especially problematic in early adolescence (coinciding with the transition to the secondary level in many countries): at this age, environmental changes bring a greater emphasis on competition, social comparison, opportunities for close student-teacher relationships are disrupted and academic work is not in line with the increasing cognitive sophistication, diverse life experiences and identity and autonomy needs of adolescents – with all of these potentially decreasing motivation and increasing disengagement. On a similar note, it was recognised that adolescents’ growing need for autonomy and participation in decision-making as well as their continuing need for strong social support and trusting relationships with adults is not matched in the transition to secondary education, leading them to turn away from school and adults in the school (Eccles, Lord, Roeser, & Jozefowicz, 1997). Even though it may appear that these issues take place at the school level (rather than system level), the system can in fact provide strong support to help make the transitions smoother (e.g. by allowing or ensuring continuity in student-teacher relationships between levels).

Vocational education and training (VET)

ESL is more common in vocational routes (partly due to the overlap with lower social background and academic weakness; NESEE, 2010). Despite this, VET is recognised for holding the potential to attract, retain and reintegrate young people by bringing them back into education and training – it can provide a re-entry point or alternative pathway (European Commission/EACEA/Eurydice/Cedefop, 2014). Understanding ESL from VET can help develop this potential of VET by addressing e.g. the lack of relevance of the curriculum, lack of flexibility, inappropriate pedagogy and young people’s labour market aspirations (NESSE, 2010).

The available data show that countries with a relatively weak VET system tend to have higher levels of ESL (ibid.; NESEE, 2010). Countries have adopted many different approaches concerning ESL from VET, but a feature they all share is their awareness of the need to allow VET students to create their own, individual paths. The report identified a student-centred, individualised approach to learning (e.g. through guidance, mentoring, individual learning plans or case management) and a competence-based approach as key aspects (ibid.; also see Rogers, 2016). Moreover, VET must be high in both quality and status, and lead to recognisable qualifications as well as match opportunities available in the job market (NESSE, 2010).

Well-balanced education systems

At a more general, holistic level, researchers have identified so-called well-balanced (EP, 2014a) or consistent education systems (Fthenakis, 2014) as vital characteristics of the education system as a whole. This refers to the characteristic that all parts of the education system from ISCED 0 to ISCED 6/7 fit together well and function in synergy. Well-balanced education systems have four dimensions: efficiency (every part of the system reinforces the results of the other parts), equity (conditions for the success of one sub-group are not allowed to damage the prospects of another sub-group), cohesion (co-responsibility of stakeholders across the education system) and representativeness (the diversity of pupils is reflected by the diversity of staff and policymakers) (EP, 2014a). The consistency of education systems (Fthenakis, 2014) refers to consistency in the theoretical basis, pedagogical principles and values, educational goals and organisation of the learning process. In EP (2014a), this is called pedagogical and professional continuity (quality of programmes and staff), curriculum continuity and structure continuity (transition between levels). According to these authors, the education system should be diverse (in terms of types of educational tracks, teaching styles etc.), but not fragmented.


Education policies may be conducive to learning for all learners, including those who struggle at school as a result of different factors, and thereby play a role in preventing or reducing ESL (e.g. high-quality ECEC). On the contrary, some features of the education system have a negative impact on ESL (e.g. grade retention). However, other sectors also play a role, e.g. the labour market and a favourable socio-economic environment (economic growth, the fight against poverty, effective integration strategies for newly arrived immigrants; see de Witte et al., 2013). For example, the labour market can act as ‘pull’ and/or ‘push’ factors in the ESL process; a high level of employment opportunities (including regional or seasonal jobs) acts as a ‘pull’ factor and can stimulate young people to leave school early in order to become financially independent or help their family; conversely, high unemployment rates can have similar or reverse effects. When students observe that unemployment does not depend on one’s qualification they may become more likely to leave school. Yet, if they notice that adults holding qualifications experience fewer problems in the labour market, they might be more likely to stay at school (de Witte et al., 2013).

The evidence reviewed in this article shows that the socio-economic segregation of schools and early tracking are linked to an increased risk for ESL. Grade retention is another risk factor, although timing (at the primary or secondary level) and opportunities provided during the retention year play a role.

The transition from the primary to secondary level is problematic for the following aspects: traditional style of academic teaching, lack of relevance of pathways and too rigid curriculum, disconnection from labour market needs, lack of permeability between pathways (academic, technical, vocational). Together with the disruption in social relationships with adults, the secondary school context changes in a way that is developmentally less appropriate for early adolescents and may lead to disengagement and ESL. Thus, the school needs to keep pace with students’ growing maturity and changing emotional, cognitive and social needs as they move through the education system in order to ensure they are providing an engaging context. Moreover, access to high-quality ECEC and a wide range of relevant and high-quality VET play the role of protective factors. The need for a coherent and well-balanced education system (i.e. all parts of the education system from preschool to tertiary education fit together well and function in synergy) across all educational levels is also warranted. The education system should be diverse, yet not fragmented. Having said this, the results do not imply that it is easy to formulate policies seeking to bring about suitable changes in this area.

To conclude, several points are important when discussing system-level (policy) ESL factors:

  • the need for divergent policy proposals reflecting different perceptions of the reasons for ESL (disaffection, non-participation and social exclusion) (Petrušauskaitė, 2010);
  • the need for evaluation in terms of the effectiveness and success of the many diverse ESL policies and measures on the national level recently implemented in light of the Europe 2020 headline target to bring ESL rates down to 10%; thus helping to make system-related conclusions at the supranational level (Meierkord & Mascherini, n.d.);
  • even though the factors highlighted in the article are important and measures derived from them should be carefully planned, attention to both the measures used in other (non-educational) sectors and to coordinated inter-sectorial actions is justified; namely, system blockages, diametric splits and displacement across different polices and sectors can seriously undermine efforts made (Downes, 2013). As indicated by the EC (2011), “all policies relevant to children and young people should contribute to the strategy against ESL. This concerns especially social policies and support services, employment, youth and integration policies. Every new policy or measure aimed at children, young people, parents or professionals working with children and young people, irrespective of whether related to the formal education system or not, should thus be tested against its contribution to reducing ESL” (pp. 14-15).

, , & (1997). From first grade forward: Early foundations of high school dropout. Sociology of Education, 70(2), 87–107.

, , & (2005). Student ratings of stressful experiences at home and school: Loss of a parent and grade retention as superlative stressors. Journal of Applied School Psychology, 21(1), 1–20.

, & (2002). Engagement and dropping out of school: A life course perspective. Working paper for the Applied Research Branch, Strategic Policy, Human Resources Development. Retrieved from

, & (2005). Early school leaving in the lower vocational track: triangulation of qualitative and quantitative data. Adolescence, 40(157), 197–213.

, , & (2006). The silent epidemic: Perspectives of high school dropouts. Civic Enterprises, LLC, Washington.

, , & (1989). Early school dropout: Configurations and determinants. Child Development, 60(6), 1437–1452.

, , & (2015). Socioeconomic background, education, and labor force outcomes: Evidence from a regional US sample. British Journal of Sociology of Education, 36(6), 934–957.

(2011). The ‘collateral impact’ of pupil behaviour and geographically concentrated socio-economic disadvantage. British Journal of Sociology of Education, 31(3), 261–276.

, , , , , & (2013). The impact of institutional context, education and labour market policies on early school leaving: A comparative analysis of EU countries. European Journal of Education, 48(3), 331–345.

, , , & (1997). The association of school transitions in early adolescence with developmental trajectories through high school. In J. Schulenberg, J. Maggs, & K. Hurrelmann (Eds.), Health risks and developmental transitions during adolescence (pp. 283–320). New York, NY: Cambridge University Press.

, & (1992). Paths to high school graduation or dropout: A Longitudinal study of a first-grade cohort. Sociology of Education, 65(2), 95–113.

. (2011). Commission staff working paper. Reducing early school leaving. Accompanying document to the proposal for a council recommendation on policies to reduce early school leaving. Retrieved from

. (2012). Study on educational support for newly arrived migrant children: Final report. Luxembourg: Publications Office of the European Union.

. (2014a). Study on the effective use of early childhood education and care in preventing early school leaving: Final report. (N° EAC/17/2012). Luxembourg: Publications Office of the European Union.

. (2014b). Study on the effective use of early childhood education and care in preventing early school leaving: Annex 1. Literature review (N° EAC/17/2012). Luxembourg: Publications Office of the European Union.

. (2014). Tackling early leaving from education and training in Europe: Strategies, policies and measures. Eurydice and Cedefop Report. Luxembourg: Publications Office of the European Union.

. (2014). Key data on early childhood education and care in Europe. Eurydice and Eurostat Report. Luxembourg: Publications Office of the European Union.

(2014). Changing the paradigm, changing the system: For more justice in preschool education. Key note at 24th EECERA Conference, September 7–10, Crete, Greece.

, & (1989). Secondary school tracking and educational inequality: Compensation, reinforcement, or neutrality? The American Journal of Sociology, 94(5), 1146.

(2009). Visible learning: A synthesis of 800+ meta-analyses on achievement. Oxford, UK: Routledge.

. (2010). Staying in education: A new way forward school and out-of-school factors protecting against early school leaving. Joint Committee on Education and Skills, First Report. Retrieved from

, & (2009). The effect of grade retention on high school completion. American Economic Journal: Applied Economics, 1(3), 33–58.

, , , & (1997). Disentangling the weight of school dropout predictors: A test on two longitudinal samples. Journal of Youth and Adolescence, 26(6), 733–762.

(2001). Meta-analysis of grade retention research: Implications for practice in the 21st century. School Psychology Review, 30(3), 420–437.

, , & (2002). Winning the battle and losing the war: Examining the relation between grade retention and dropping out of high school. Psychology in the Schools, 39(4), 441–457.

, , , , & (2002). Exploring the association between grade retention and dropout: A longitudinal study examining socio-emotional, behavioral, and achievement characteristics of retained students. The California School Psychologist, 7(1), 51–62.

, , , & (2000). A prospective longitudinal study of high school dropouts examining multiple predictors across development. Journal of School Psychology, 38(6), 525–549.

, & (2003). Dropping out of high school: the role of school organization and structure. American Educational Research Journal, 40(2), 353–393.

, & (n.d.) Preventing and re-integrating early school leavers – a meta evaluation of policies implemented in 7 European member states. European Foundation for the Improvement of Living and Working Conditions (EUROFOUND), Dublin, Ireland. Retrieved from

. (2010). Early school leaving: Lessons from research for policy makers. European Commission.

. (2008). 10 steps to equity in education. Policy brief. Retrieved from

. (2012). Equity and quality in education: Supporting disadvantaged students and schools. Paris: OECD. Retrieved from

. (2017). Attendance in early childhood education and care programmes and academic proficiencies at age 15. Lessons from PISA. retrieved from

(2010). Dropping out of school – an issue of disaffection, non-participation or social exclusion? Analysing school policies towards Roma school children in Lithuania. Ethnicity Studies, 2010(1–2).

(1994). Grade retention and school dropout: investigating the association. American Educational Research Journal, 31(4), 729–59.

(2016). Disengagement from education. London: Institute of Education Press.

, & (2008). Why students drop out of school: a review of 25 years of research. California Dropout Research Project Report #5. Retrieved from

, & (2005). Test scores, dropout rates, and transfer rates as alternative indicators of high school performance. American Educational Research Journal, 41, 3–42.

, & (2008). Early school-leaving in the Netherlands. The role of student-, family- and school factors for early school-leaving in lower secondary education. Maastricht: Research Centre for Education and the Labour Market.

, & (2015) The impact of local labour market conditions on school leaving decisions. Working paper series from Institute for Social and Economic Research, 2015(14), 1–38. Retrieved from!/file/F2_2_Tumino.pdf