{"id":1490,"date":"2025-12-04T10:04:05","date_gmt":"2025-12-04T08:04:05","guid":{"rendered":"https:\/\/2026.inimareng.ee\/aruanne\/%chapter%\/pisa-blindness-why-is-estonia-not-using-the-full-potential-of-its-register-data\/"},"modified":"2026-06-09T08:00:54","modified_gmt":"2026-06-09T06:00:54","slug":"pisa-blindness-why-is-estonia-not-using-the-full-potential-of-its-register-data","status":"publish","type":"article","link":"https:\/\/2026.inimareng.ee\/en\/aruanne\/hariduse-andmetarkus\/pisa-blindness-why-is-estonia-not-using-the-full-potential-of-its-register-data\/","title":{"rendered":"PISA blindness: Why is Estonia not using the full potential of its register data?"},"content":{"rendered":"\n    <div class=\"highlight-box highlight-box-purple p-8 xl:p-12 my-10\">\n                    <div class=\"mb-6 font-bold text-3xl uppercase text-purple\">KEY MESSAGES<\/div>\n        \n        <ul>\n<li><strong>Educational inequality in Estonia has the face of a child who studies in an under-resourced school and comes from a disadvantaged socioeconomic background.<\/strong> School environments vary widely \u2013 some lack teachers, support specialists and learning materials, while others are comparatively well resourced. This raises the question of whether shortages at school are increasingly pushing parents to compensate for these shortcomings with their own resources.<\/li>\n<li><strong>Tallinn is the region where educational stratification is deepening most rapidly.<\/strong> In the capital, the association between pupils\u2019 mathematics results and the average income of their parents has consistently been nearly twice as strong as elsewhere in Estonia.<\/li>\n<li><strong>Although Estonia is known as an e-state, digital development has not been evenly reflected in the education sector.<\/strong> Register data \u2013 whose use has been constrained by limited awareness and increasingly stringent data protection requirements \u2013 could already some time ago have shown that educational inequality was deepening. By contrast, PISA\u2019s socioeconomic status indicator, which rests on a weak scientific foundation, tends to soften or distort the actual picture.<\/li>\n<\/ul>\n    <\/div>\n\n<h2 class=\"mb-6 text-3xl uppercase font-medium text-purple\">\n    INTRODUCTION<\/h2>\n<p class=\"wp-block-paragraph\">According to the OECD definition, educational equity does not mean that all pupils achieve the same outcomes, but that their success should not depend on factors beyond their control, such as socioeconomic background.<a href=\"#references\" id=\"reference-1\" class=\"reference-number\">1<\/a> The influence of social background on learning outcomes varies by age, gender and subject: it tends to be stronger where the impact of school is weaker \u2013 for example, in reading \u2013 whereas school quality plays a greater role in subjects such as mathematics and science. The deepening of educational inequality in Estonia calls for decisive action to improve the quality of schools that are in a weaker position. Data-driven education analytics have significant \u2013 yet largely underused \u2013 potential to support such efforts.<\/p>\n\n<h2 class=\"mb-6 text-3xl uppercase font-medium text-purple\">\n    CRITIQUE OF PISA\u2019S socioECONOMIC STATUS INDICATOR<\/h2>\n<p class=\"wp-block-paragraph\">Until now, Estonia has relied on the PISA assessment to evaluate educational inequality. However, the 2022 results, which indicated increasing stratification \u2013 with 13% of the variance in mathematics performance explained by pupils\u2019 home background \u2013 raise the question of whether educational inequality in Estonia has genuinely intensified in recent years, or whether the change reflects differences in measurement rather than a new development.<a href=\"#references\" id=\"reference-2\" class=\"reference-number\">2<\/a> It should be noted that PISA\u2019s background indicator has changed over time, most notably in the 2022 cycle, and direct comparisons with earlier results are not always appropriate, even though they are often made. We wish to emphasise that this article examines only the adequacy of PISA\u2019s socioeconomic background indicator and does not in any way question the quality of PISA\u2019s academic assessments.<\/p>\n\n<p class=\"wp-block-paragraph\">International research cautions against drawing overly far-reaching conclusions from PISA\u2019s background indicator. It cannot therefore be claimed that pupils\u2019 socioeconomic background has begun to influence results only recently. In the academic literature, uncritical reliance on prestige and titles is described as prestige bias.<a href=\"#references\" id=\"reference-3\" class=\"reference-number\">3<\/a> We tend to place greater trust in individuals and institutions that have already achieved success or recognition in a particular field. This so-called fast track of trust operates especially quickly in the case of international rankings such as PISA. However, the reliability of a study depends not on institutional authority, but above all on methodological rigour and relevance. Unfortunately, PISA\u2019s high prestige has overshadowed a more thorough examination of internal inequalities in Estonia.<a href=\"#references\" id=\"reference-4\" class=\"reference-number\">4<\/a><\/p>\n\n<p class=\"wp-block-paragraph\">In the academic literature, the so-called golden trio for measuring educational inequality consists of parental income, occupation and education. In PISA, pupils\u2019 home background is measured using the index of economic, social and cultural status (ESCS), which combines social, economic and cultural factors into a single composite indicator. Researchers have argued that the ESCS lacks a clear scientific foundation: while it includes parental education and occupation from the classical socioeconomic trio, it substitutes family income with household possessions as a proxy.<a href=\"#references\" id=\"reference-5\" class=\"reference-number\">5<\/a> The core problem of PISA\u2019s socioeconomic background indicator lies in how the data are collected. The measure is based on pupil questionnaires, and evidence from Germany suggests that children from poorer families are more likely not to respond.<a href=\"#references\" id=\"reference-6\" class=\"reference-number\">6<\/a> We therefore examine in more detail the limitations of the three components used by PISA to assess pupils\u2019 socioeconomic background. <\/p>\n\n<p class=\"wp-block-paragraph\">The economic component, \u2018home possessions and resources\u2019, measures material well-being on the basis of consumption patterns \u2013 specifically, items available in the homes of 15-year-olds.<a href=\"#references\" id=\"reference-7\" class=\"reference-number\">7<\/a> Pupils indicate whether certain items are present at home, such as a desk for studying or their own room. The index also includes three country-specific items (see Table 2.2.1) that PISA considers indicative of family wealth in that particular country.<\/p>\n\n<p class=\"wp-block-paragraph\">According to a 2025 survey by Kantar Emor, 56% of Estonia\u2019s population believe that small loans are granted too easily, and nearly one fifth have taken out a consumer loan in the past three years.<a href=\"#references\" id=\"reference-8\" class=\"reference-number\">8<\/a> This raises doubts about the use of possessions as a proxy for wealth. For example, purchasing a PlayStation on instalments does not mean that a family can afford to hire a private mathematics tutor if the child develops learning gaps.<\/p>\n\n<p class=\"wp-block-paragraph\">Traditionally, the number of books at home has been treated as an indicator of a strong intellectual environment associated with higher academic achievement. PISA includes a question on how many books pupils have at home, broken down by type, such as religious books, classical literature, scientific books and technical manuals, and by precise ranges of 1\u20135, 6\u201310 and more than 10 books. It is questionable whether even the average adult could answer this with such precision.<\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Table 2.2.1<\/strong>\n                            <span class=\"text-brown font-medium\">PISA indicators of family wealth (home possessions) in selected countries<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <a data-fslightbox href=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Tabel-2.2.1-scaled.png\">\n                <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Tabel-2.2.1-scaled.png\" alt=\"\" class=\"object-cover\">\n            <\/a>\n        <\/div>\n    <\/div>\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">OECD<a href=\"#references\" id=\"reference-9\" class=\"reference-number\">9<\/a>, PISA<a href=\"#references\" id=\"reference-10\" class=\"reference-number\">10<\/a><\/span>\n            <\/div>\n        \n            <\/div>\n\n<p class=\"wp-block-paragraph\">In PISA, the background questionnaire is generally completed by pupils themselves, while parental questionnaires are voluntary and implemented only in some countries. In a limited number of countries, this has made it possible to compare pupils\u2019 and parents\u2019 responses to questions on socioeconomic background. For items such as books and digital devices, agreement between pupils\u2019 and parents\u2019 responses is below 50%.<a href=\"#references\" id=\"reference-11\" class=\"reference-number\">11<\/a> The OECD itself acknowledges this limitation: \u2018Direct data collection from parents can be impractical in large-scale studies due to its associated costs. [\u2026] Proxy-reporting by pupils about their parents\u2019 occupations and educational attainments is a pragmatic solution, but concerns about the completeness and accuracy of the collected data can be raised.\u2019<a href=\"#references\" id=\"reference-12\" class=\"reference-number\">12<\/a><\/p>\n\n<p class=\"wp-block-paragraph\">The social component, \u2018parents\u2019 occupational status\u2019, is measured on the basis of parents\u2019 occupation and job position. In general, pupils underestimate their parents\u2019 occupational status almost as often as they overestimate it \u2013 by 20% and 21%, respectively.<a href=\"#references\" id=\"reference-13\" class=\"reference-number\">13<\/a> PISA reports also note that the share of missing and inaccurate data has increased over time. Most frequently, pupils fail to report their parents\u2019 occupation \u2013 in the United Kingdom and Germany, this applies to more than 20% of pupils \u2013 and non-response is more common among those from weaker socioeconomic backgrounds.<a href=\"#references\" id=\"reference-14\" class=\"reference-number\">14<\/a><\/p>\n\n<p class=\"wp-block-paragraph\">As for the cultural component, \u2018parents\u2019 level of education\u2019, almost one in three pupils provides information that differs from that reported by the parent. Underestimation (18%) is more common than overestimation (13%). Pupils whose parents in fact have a lower level of education tend to overestimate their parents\u2019 educational attainment.<a href=\"#references\" id=\"reference-15\" class=\"reference-number\">15<\/a><\/p>\n\n<h2 class=\"mb-6 text-3xl uppercase font-medium text-purple\">\n    EDUCATIONAL INEQUALITY IS DEEPENING \u2013 REGISTER DATA WOULD HAVE SHOWN THIS SOME TIME AGO<\/h2>\n<p class=\"wp-block-paragraph\">Across Estonia, the association between parental income and schools\u2019 mathematics results has strengthened over time (Table\u00a02.2.2). In 2017, the correlation with mothers\u2019 and fathers\u2019 income was moderate; by 2021, it had become strong. At the same time, the negative association between pupils\u2019 examination results and the share of social benefits in family income has intensified. The higher the proportion of benefits in household income, the lower the school\u2019s average mathematics examination results tend to be.<\/p>\n\n<p class=\"wp-block-paragraph\">In the capital city, Tallinn, educational stratification has deepened particularly markedly (see Table\u00a02.2.3 and Figure\u00a02.2.1). While the association between a school\u2019s average income of mothers and its average mathematics examination results has remained consistently strong in the capital, the same relationship elsewhere in Estonia has been almost twice as weak and has strengthened only gradually. In 2021, the average income of mothers explained 24% of the variance in lower secondary mathematics examination results across schools in Estonia as a whole; in Tallinn, the figure was 50%, compared with 15% in the rest of Estonia.<\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Table 2.2.2<\/strong>\n                            <span class=\"text-brown font-medium\">Association between a school\u2019s average mathematics examination result and parental economic status, 2017\u20132021<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Tabel-2.2.2.png\" alt=\"\" class=\"object-cover\">\n        <\/div>\n    <\/div>\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">Statistics Estonia data<\/span>\n            <\/div>\n        \n                    <div>\n                <span class=\"text-brown font-semibold\">Note.<\/span>\n                <span class=\"text-sm text-brown\">The table presents Pearson correlation coefficients. In interpreting their strength, r < 0.3 indicates a weak association, 0.3 \u2264 r < 0.5 a moderate association and r \u2265 0.5 a strong association. Correlations marked with (***) are statistically significant at the 0.05 level. The square of the Pearson correlation coefficient corresponds to the coefficient of determination in a regression model with one explanatory variable. The coefficient of determination shows the proportion of variance in the dependent variable that is explained by variation in the independent variable.<\/span>\n            <\/div>\n            <\/div>\n\n<p class=\"wp-block-paragraph\"><strong>Data and methodology.<\/strong><em> Statistics Estonia\u2019s data on the income of parents (mother, father and total household income) of pupils in Estonian general education schools, as well as the share of social benefits in household income for 2017\u20132021, were combined with average lower secondary mathematics examination results obtained from the Ministry of Education and Research (EHIS data). The data were linked by Statistics Estonia; researchers received the results in aggregated form, that is, as school-level averages. Time required: 1\u20132 months for consultations on data protection requirements and six working hours for data linkage (Statistics Estonia).\nCost: 360 euros, covering the linkage of datasets from different registers by Statistics Estonia. Sample: Using Estonian register data, it was possible to cover nearly 350\u00a0basic schools, where between 12,000 and 14,000 pupils per year sat the mathematics examination and completed Grade 9 in 2017\u20132021. In those years, approximately 126,000\u2013133,000 pupils were enrolled in basic education, depending on the academic year (population used to calculate average parental income). By comparison, PISA 2022 covered 6,392 pupils in 196 schools. After the initial analysis, the authors also applied to extend the income data series to 2024, but due to stricter data protection requirements this was no longer permitted.<\/em><\/p>\n\n<p class=\"wp-block-paragraph\">The influence of fathers\u2019 income is even more pronounced in Tallinn: over time, the association between mathematics results and fathers\u2019 income has become very strong in the capital, while elsewhere in Estonia it has remained moderate. This indicates that in Tallinn, school results increasingly reflect parental economic background. A similar pattern is observed in relation to the share of social benefits in household income. <\/p>\n\n    <div class=\"highlight-box highlight-box-purple p-8 xl:p-12 text-2xl xl:text-3xl text-brown font-semibold my-10\">\n        \n        In 2021, the average income of mothers explained 24% of the variance in lower secondary mathematics examination results across schools in Estonia as a whole; in Tallinn, the figure was 50%.\n    <\/div>\n\n<p class=\"wp-block-paragraph\">Whereas in 2017 a school\u2019s average mathematics result was largely clustered between 30 and 40 points, by 2021 Tallinn in particular had seen growth at both ends of the distribution \u2013 more lower-performing schools (below 20 points) and more higher-performing schools (above 40 points). This points to widening inequality between schools in the capital (Figure 2.2.2). In Tallinn, children living only a few kilometres apart may face markedly different starting points in life: in one school the average mathematics result barely exceeds 20 points, while in another nearby it approaches the maximum.<\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Table 2.2.3<\/strong>\n                            <span class=\"text-brown font-medium\">Association between a school\u2019s average mathematics examination result and parental economic status: Tallinn, the rest of Estonia excluding Tallinn, and the rest of Estonia excluding Tallinn and Tartu<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <a data-fslightbox href=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Tabel-2.2.3.png\">\n                <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Tabel-2.2.3.png\" alt=\"\" class=\"object-cover\">\n            <\/a>\n        <\/div>\n    <\/div>\n\n\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">Statistics Estonia data<\/span>\n            <\/div>\n        \n                    <div>\n                <span class=\"text-brown font-semibold\">Note.<\/span>\n                <span class=\"text-sm text-brown\">The table presents Pearson correlation coefficients. In interpreting their strength, r < 0.3 indicates a weak association, 0.3 \u2264 r < 0.5 a moderate association and r \u2265 0.5 a strong association. Correlations marked with (***) are statistically significant at the 0.05 level. The square of the Pearson correlation coefficient corresponds to the coefficient of determination in a regression model with one explanatory variable. The coefficient of determination shows the proportion of variance in the dependent variable explained by variation in the independent variable.<\/span>\n            <\/div>\n            <\/div>\n\n<h2 class=\"mb-6 text-3xl uppercase font-medium text-purple\">\n    WHAT MATTERS MORE \u2013 PARENTS\u2019 EDUCATION, MONEY OR GENES?<\/h2>\n<p class=\"wp-block-paragraph\">The general premise underlying socioeconomic status indices is that parental education, occupation and economic resources are directly associated with children\u2019s learning outcomes. This is often supplemented by the so-called genetic lottery argument \u2013 the assumption that higher-status families also possess more favourable genetic endowments. Let us consider whether the picture is really so clear-cut.<\/p>\n\n<p class=\"wp-block-paragraph\">The influence of parental background, and especially educational attainment, on educational outcomes has been debated extensively since 1966, when James S. Coleman and colleagues published the report <i>Equality of Educational Opportunity<\/i>.<a href=\"#references\" id=\"reference-16\" class=\"reference-number\">16<\/a> That study, along with subsequent research, found that alongside school resources, family background and parental education play a significant role in shaping children\u2019s academic achievement. <\/p>\n\n<p class=\"wp-block-paragraph\">Should this be the end of the debate \u2013 the conclusion that children of more educated parents perform better academically? <\/p>\n\n<p class=\"wp-block-paragraph\">It is nevertheless important to remain critically aware of context when invoking older studies \u2013 national and societal conditions have changed, particularly when comparing Estonia\u2019s education system with others. Arguments often draw on research conducted in the United States in the 1960s\u20131980s, which highlighted the decisive role of parental education.<a href=\"#references\" id=\"reference-17\" class=\"reference-number\">17<\/a> Yet this effect depends heavily on interactions with other factors \u2013 family income, family structure and regional opportunities \u2013 which differ today. For example, a highly educated mother raising a child alone may not necessarily be able to offset learning gaps that arise in a low-performing school facing shortages of qualified teachers and support specialists. In short, parents may at times have to compensate for differences in school quality. This does not imply that parental education is unimportant, but it does not explain the full picture, nor why children of less educated parents also reach higher education. What matters more is parental example \u2013 the extent to which parents engage with their children and encourage learning.<a href=\"#references\" id=\"reference-18\" class=\"reference-number\">18<\/a> Although such support is more often associated with educated parents, it may equally characterise those with lower levels of education.<\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Figure 2.2.1<\/strong>\n                            <span class=\"text-brown font-medium\">Association between school average mathematics examination results (2021) and mothers\u2019 average annual income (2020)<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <a data-fslightbox href=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Joonis-2.2.1.png\">\n                <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Joonis-2.2.1.png\" alt=\"\" class=\"object-cover\">\n            <\/a>\n        <\/div>\n    <\/div>\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">Statistics Estonia data<\/span>\n            <\/div>\n        \n            <\/div>\n\n<p class=\"wp-block-paragraph\">The genetic lottery argument has also been widely examined: do genes determine a child\u2019s academic success? Genes do play a significant role in explaining differences in mathematical ability and general cognitive capacity.<a href=\"#references\" id=\"reference-19\" class=\"reference-number\">19<\/a> However, this does not mean that the environment lacks substantial influence. Nor can it be assumed that only talented or genetically advantaged children are born into wealthier and more educated families. The academic literature recognises the phenomenon of \u2018twice-exceptional\u2019 (2e) children \u2013 those who are both highly gifted, or have a high IQ, and have special educational needs, such as dyslexia, attention deficit hyperactivity disorder or autism. Research shows that pupils from weaker socioeconomic backgrounds are identified as gifted far less often, meaning that twice-exceptional children from poorer families frequently go unrecognised and unsupported. Giftedness and special educational needs may co-occur and often mask one another, making identification and appropriate support more complex. Studies in the United States indicate that pupils from the highest-income families are identified as gifted six times more often than those from the lowest-income families, which suggests that twice-exceptional children from poorer families face a particularly high risk of being overlooked.<a href=\"#references\" id=\"reference-20\" class=\"reference-number\">20<\/a> By neglecting disadvantaged schools and children from weaker socioeconomic backgrounds, substantial talent may be lost in Estonia. <\/p>\n\n<p class=\"wp-block-paragraph\">The expression of genetic potential depends to a large extent on the environment, and educational institutions play a central role in this process \u2013 schools may function either as a force that mitigates inequality or as a mechanism that reinforces it.<a href=\"#references\" id=\"reference-21\" class=\"reference-number\">21<\/a> A school may reinforce inequality if a systemic shortage of qualified teachers and support specialists disproportionately affects schools serving pupils from weaker socioeconomic backgrounds. Relying on the genetic lottery argument effectively downplays the role of school as an environment shaping child development and overlooks children and young people born into poverty who attend schools lacking the skills and resources needed to compensate for their home background. <\/p>\n\n<p class=\"wp-block-paragraph\">A higher socioeconomic background also gives families a compensatory advantage.<a href=\"#references\" id=\"reference-22\" class=\"reference-number\">22<\/a> A cognitively less capable child born into a more advantaged family is more likely to obtain a good education than a gifted child from a weaker socioeconomic background. Parents may compensate for lower academic performance by providing additional support, such as preparatory schooling or private tutoring.<a href=\"#references\" id=\"reference-23\" class=\"reference-number\">23<\/a> They may also help to mitigate obstacles linked to potential genetic risks, including learning or behavioural difficulties.<a href=\"#references\" id=\"reference-24\" class=\"reference-number\">24<\/a><\/p>\n\n<p class=\"wp-block-paragraph\">A Dutch study involving more than 29,000 pairs of twins found that both family and school socioeconomic background shape the expression of genetic risks, placing children from weaker backgrounds at a double disadvantage.<a href=\"#references\" id=\"reference-25\" class=\"reference-number\">25<\/a> Their home environment does not sufficiently offset these risks, and they often attend under-resourced schools. In such cases, a supportive school environment may be decisive. Meta-analyses confirm that a safe school climate, positive relationships with teachers, high expectations and the strengthening of self-confidence can reduce the negative effects of socioeconomic background.<a href=\"#references\" id=\"reference-26\" class=\"reference-number\">26<\/a> Dutch data further indicate that when schools raise expectations and self-belief among pupils from weaker backgrounds, the impact is considerably greater than for children from wealthier families.<a href=\"#references\" id=\"reference-27\" class=\"reference-number\">27<\/a><\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Figure 2.2.2<\/strong>\n                            <span class=\"text-brown font-medium\">Average mathematics examination results in schools in Tallinn and elsewhere in Estonia, 2017 and 2021<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <a data-fslightbox href=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Joonis-2.2.2.png\">\n                <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Joonis-2.2.2.png\" alt=\"\" class=\"object-cover\">\n            <\/a>\n        <\/div>\n    <\/div>\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">Statistics Estonia data<\/span>\n            <\/div>\n        \n            <\/div>\n\n<p class=\"wp-block-paragraph\">Much has been written about pupils\u2019 socioeconomic background as a feature of the home environment, but less attention has been paid to the socioeconomic environment of schools, that is, their resources. While pupils\u2019 socioeconomic background largely reflects parental background, two dimensions can be distinguished at school level: the school\u2019s socioeconomic profile (the aggregate of pupils\u2019 backgrounds) and its internal resources and quality indicators, including the quality of teaching and school leadership. Schools serving pupils from weaker socioeconomic backgrounds often operate with limited resources \u2013 shortages of teachers, support specialists and teaching materials are common.<a href=\"#references\" id=\"reference-28\" class=\"reference-number\">28<\/a> These schools therefore face a double disadvantage. In Estonia, educational inequality has the face of a child who studies in an under-resourced school and comes from a disadvantaged socioeconomic background. For example, a systemic shortage of subject teachers in certain schools reduces pupils\u2019 opportunities to progress, giving an advantage to those whose parents can provide private tutoring or have relevant subject knowledge themselves. The expansion of so-called shadow education \u2013 the private tutoring market \u2013 is often an indicator of educational inequality.<a href=\"#references\" id=\"reference-29\" class=\"reference-number\">29<\/a><\/p>\n\n<p class=\"wp-block-paragraph\">How, then, should we resolve the debate over what is decisive in a child\u2019s development \u2013 the mother\u2019s education, family wealth or genes? Education is a complex system in which no single factor determines success: if one link is weak, others must compensate. In Estonia, it is worth asking to what extent disparities in school resources \u2013 shortages of teachers and support specialists, limited opportunities to purchase high-quality textbooks or equip laboratories \u2013 intensify educational inequality, and whether this increasingly pushes parents to offset school shortcomings through their own spending or through educational mobility, that is, seeking a better school elsewhere.<\/p>\n\n<h2 class=\"mb-6 text-3xl uppercase font-medium text-purple\">\n    SUMMARY: HOW DO WE MOVE FORWARD?<\/h2>\n<p class=\"wp-block-paragraph\"><strong>In Estonia, there is a clear need to differentiate teachers\u2019 salaries by subject, alongside a strategic focus on improving teaching quality in those subjects.<\/strong> Why is it important to consider subject teachers\u2019 alternative employment opportunities? Research from other countries shows that teachers earn on average 23.5% less than other professionals with the same level of education \u2013 a phenomenon known as the teacher pay penalty.<a href=\"#references\" id=\"reference-30\" class=\"reference-number\">30<\/a> This gap is most pronounced for science and mathematics teachers, whose skills are in high demand and often better rewarded outside schools. At the same time, in some subjects teaching offers a relative pay advantage and may represent the most attractive career path.<\/p>\n\n<p class=\"wp-block-paragraph\">Data on those trained as mathematics teachers in Estonia show that in 2024, a qualified mathematics teacher not working in a school earned approximately 1.6 times as much in Harju County (2,838 euros), 1.4 times as much in Tartu (2,573\u00a0euros) and 1.3 times as much elsewhere in Estonia (2,459\u00a0euros) compared with the minimum teacher salary (1,820\u00a0euros in 2024).<a href=\"#references\" id=\"reference-31\" class=\"reference-number\">31<\/a> Research indicates that increasing the salaries of teachers in STEM subjects could reduce their rate of leaving the profession by 18\u201328%.<a href=\"#references\" id=\"reference-32\" class=\"reference-number\">32<\/a> The Basic Schools and Upper Secondary Schools Act provides that knowledge in selected subjects \u2013 primarily mathematics, Estonian and a foreign language \u2013 is assessed at graduation. The state has therefore already assigned greater importance to certain subjects. These subjects play a decisive role in shaping young people\u2019s subsequent educational pathways. <\/p>\n\n    <div class=\"highlight-box highlight-box-purple p-8 xl:p-12 text-2xl xl:text-3xl text-brown font-semibold my-10\">\n        \n        In Estonia, it is worth asking to what extent disparities in school resources \u2013 shortages of teachers and support specialists, limited opportunities to purchase high-quality textbooks and similar constraints \u2013 further deepen educational inequality.\n    <\/div>\n\n<p class=\"wp-block-paragraph\"><strong>Data-driven education policy.<\/strong> Estonia has excellent administrative registers for addressing educational inequality in a targeted way, yet for years \u2013 if not decades \u2013 we have largely relied on conclusions drawn from PISA background questionnaires. This has contributed to a form of PISA blindness in understanding inequality within the education system. Register-based data in Estonia should be used far more decisively to design targeted measures to reduce educational inequality. In analysing inequality, particular attention should be paid to schools\u2019 socioeconomic context \u2013 including teacher qualifications and availability, the quality of school leadership, the presence of support specialists and pupil\u2013teacher ratios. However, the potential of Estonia\u2019s e-state in education is limited by insufficient awareness of existing data \u2013 we tend to commission new and time-consuming surveys rather than use administrative data \u2013 and, even more so, by increasingly stringent data protection requirements.<\/p>\n\n<p class=\"wp-block-paragraph\">Systematic use of register data should form the foundation of education policy, both in monitoring developments and in planning reforms. For example, when designing teacher career and salary models, decisions should be based on a comprehensive data-driven assessment rather than on formal qualifications alone. Otherwise, we may end up in a situation where mathematics is taught by a formally qualified history teacher without subject-specific competence, while an experienced engineer with strong teaching skills who wishes to change career is considered less suitable simply because they lack formal teacher certification. <\/p>\n\n<p class=\"wp-block-paragraph\">It would also be worth considering linking PISA academic test results with Estonia\u2019s administrative register data in order to analyse the background characteristics of participating schools and pupils\u2019 socioeconomic background, and to track developments over time \u2013 for example, what has become of high-performing PISA pupils ten years later, whether school results have improved or declined, and what characterises the schools of top performers. The aim would not primarily be international comparison but a deeper understanding of Estonia\u2019s own education system.<\/p>\n\n    <div class=\"highlight-box highlight-box-purple p-8 xl:p-12 text-2xl xl:text-3xl text-brown font-semibold my-10\">\n        \n        In general, a pupils\u2019s home background matters more where the school\u2019s influence is weaker (e.g. reading), whereas school quality plays a greater role in subjects more directly shaped by the school itself (e.g. mathematics and science).\n    <\/div>\n\n<p class=\"wp-block-paragraph\"><strong>Teacher quality is most decisive for pupils from weaker socioeconomic backgrounds.<\/strong> We must also have the courage to ask what impact a poor teacher or school has on a pupil\u2019s future. Nobel laureate James J. Heckman has emphasised that educational inequality is not only a social issue but also an economic one \u2013 supporting disadvantaged children increases societal productivity and reduces future costs, such as unemployment and crime.<a href=\"#references\" id=\"reference-33\" class=\"reference-number\">33<\/a> Preventing educational inequality is therefore far less costly than addressing its consequences later. As Heckman\u2019s well-known curve on the rate of return to education investment illustrates (see Figure\u00a02.2.3), public investment in the early development of children from low socioeconomic backgrounds yields the greatest returns.<\/p>\n\n    <div class=\"mb-6\">\n                    <strong class=\"text-purple\">Figure 2.2.3<\/strong>\n                            <span class=\"text-brown font-medium\">Rate of return to investment in education across the life course<\/span>\n            <\/div>\n<div>\n            <div class=\"mb-6\">\n            <img decoding=\"async\" src=\"https:\/\/2026.inimareng.ee\/wp-content\/uploads\/2025\/12\/Joonis-2.2.3.png\" alt=\"\" class=\"object-cover\">\n        <\/div>\n    <\/div>\n\n    <div class=\"mb-6 space-y-3\">\n                    <div>\n                <span class=\"text-purple uppercase font-semibold\">Source:<\/span>\n                <span class=\"text-sm text-brown\">Heckman<a href=\"#references\" id=\"reference-34\" class=\"reference-number\">34<\/a><\/span>\n            <\/div>\n        \n            <\/div>\n\n<p class=\"wp-block-paragraph\">Research also shows that the impact of pupils\u2019 social background on learning outcomes varies across countries: it depends on age, gender and even subject. In general, home background matters more where the influence of school is weaker (e.g. reading), whereas school quality plays a stronger role in subjects shaped directly by school practice (e.g. mathematics and science).<a href=\"#references\" id=\"reference-37\" class=\"reference-number\">37<\/a> For this reason, in Estonia greater attention should be paid to the quality of teaching in STEM subjects \u2013 including learning materials, teacher competence and instructional quality, competitive salaries, and class sizes in these subjects. The quality of school leadership must also not be overlooked: even the most capable teachers cannot realise their potential if school management and working conditions are weak.<a href=\"#references\" id=\"reference-36\" class=\"reference-number\">36<\/a><\/p>\n","protected":false},"featured_media":0,"parent":0,"menu_order":0,"template":"","chapter":[3],"class_list":["post-1490","article","type-article","status-publish","hentry","chapter-hariduse-andmetarkus"],"acf":[],"_links":{"self":[{"href":"https:\/\/2026.inimareng.ee\/en\/wp-json\/wp\/v2\/article\/1490","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/2026.inimareng.ee\/en\/wp-json\/wp\/v2\/article"}],"about":[{"href":"https:\/\/2026.inimareng.ee\/en\/wp-json\/wp\/v2\/types\/article"}],"wp:attachment":[{"href":"https:\/\/2026.inimareng.ee\/en\/wp-json\/wp\/v2\/media?parent=1490"}],"wp:term":[{"taxonomy":"chapter","embeddable":true,"href":"https:\/\/2026.inimareng.ee\/en\/wp-json\/wp\/v2\/chapter?post=1490"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}