School tracking and intergenerational social mobility

Postponing school tracking can increase social mobility without significant adverse effects on educational achievement

VATT Institute for Economic Research, Finland, and IZA, Germany

one-pager full article

Elevator pitch

The goal of school tracking (assigning students to different types of school by ability) is to increase educational efficiency by creating more homogeneous groups of students that are easier to teach. However, there are concerns that, if begun too early in the schooling process, tracking may improve educational attainment at the cost of reduced intergenerational social mobility. Recent empirical evidence finds no evidence of an efficiency–equality trade-off when tracking is postponed.

Most European countries have moved school
                        tracking to a later age, 2002

Key findings


There is no empirical evidence of an efficiency–equality trade-off when school tracking is postponed.

Postponing school tracking weakens the association between pupils’ educational achievement and parental background, improving social mobility.

Postponing school tracking leads to greater income mobility across generations.


There is no evidence that postponing school tracking has large effects on average achievement.

Postponing school tracking may increase dropouts among low-achieving students.

Postponing tracking may increase the total cost of the educational system.

Author's main message

The effects of early tracking on educational achievement are very modest, even slightly negative. However, empirical evidence shows that delaying school tracking until a later age is good for social mobility without significantly affecting average educational achievement. Thus, there is no evidence of a trade-off between educational efficiency and equality of educational attainment. But postponing tracking is not a cure-all policy that can be used to improve educational achievement regardless of other factors.


Countries in the Organisation for Economic Co-operation and Development (OECD) differ in how they segregate or “stream” students into different types of school on the basis of ability, a practice known as school tracking (see Different types of tracking). One of the main differences between countries is the age at which tracking begins. While some countries track students as early as age ten, many others postpone tracking to age 16 and beyond.

Because educational tracks differ in curricular content and in the further educational opportunities they provide, many opponents of early tracking have voiced concerns that it could reduce social mobility between generations. They argue that early tracking decisions are more probably determined by parental background than by a student’s innate abilities. In contrast, proponents of early tracking emphasize the greater efficiency that results from teaching students in more homogeneous ability groups. These conflicting arguments suggest that early tracking opens the potential for a trade-off between efficiency of education and equality of educational opportunity.

Discussion of pros and cons

School tracking and ability grouping

While the number of tracks differs across countries, the most common pattern is to separate students into schools that teach more academic material and those that concentrate on vocational subjects that prepare students for more immediate entry into the labor market. Consequently, school tracks differ in the kinds of opportunities they offer for further education. In some countries, being assigned to a vocational track explicitly rules out the possibility of entering university education.

The initial tracking age in European countries in 2002 varied from ages ten to 16, with Austria and Germany applying the earliest tracking. By the 1960s, after several decades of educational reforms, most countries in Western Europe had delayed tracking to some extent. Only German-speaking countries have retained the pre-World War II system of early tracking.

It is important to distinguish school tracking, the topic of this paper, from ability grouping within schools (see Different types of tracking). School tracking explicitly allocates students to different curricula and standards, whereas ability groups usually follow the same educational goals (or are supposed to do so). Furthermore, school tracking is a feature of the whole educational system and can be more easily affected by educational policies, whereas ability grouping is often at the discretion of individual teachers or school administrators.

Measuring intergenerational social mobility

The persistence of economic and occupational status across generations is one of the most thoroughly studied subjects in the social sciences. One reason is that measures of persistence of status have been seen as a way to assess how far societies are from reaching the objective of equality of opportunity. According to this widely accepted principle, an individual’s lifetime status should depend as little as possible on circumstances beyond the person’s control, such as parental background.

Persistence of economic status is measured by estimating the association between children’s lifetime earnings or educational status and those of their parents. When trying to understand the mechanisms underlying the intergenerational persistence of status, researchers often also examine intermediate outcomes such as test scores and educational transitions, such as from secondary to tertiary education. Here, the focus is on studies of the effect of early school tracking on intergenerational mobility that look at the persistence between generations of both early educational achievement and adult outcomes, such as labor market status or completed schooling.

The efficiency–equality trade-off

The efficiency argument

The proponents of early school tracking argue that because students are fundamentally heterogeneous in their abilities, aspirations, and academic interests, assigning them to different types of school results in efficiency gains from both improved academic achievement and better use of resources. Tracking students based on these attributes makes individual schools more homogeneous and makes it easier to tailor the curriculum and teaching practices to student needs. Thus, the argument goes, tracking should improve the quality of teaching and lead to better educational outcomes. To fully reap these benefits, tracking should take place as soon as teaching moves beyond basic skills in reading, writing, and arithmetic.

Early tracking also makes it easier to allocate appropriate resources across the tracks. For example, class sizes can vary depending on the track, and the same applies to the qualifications required of teachers. This flexibility makes it possible to fine-tune resource allocation to address the problems of particular groups. Furthermore, tracking should save resources by allowing schools to concentrate on teaching subjects that are more closely related to the future needs of students.

In emphasizing these efficiency gains, proponents of early tracking argue that tracking should improve average academic achievement. Moreover, as the efficiency argument relies on the better design of teaching that tracking allows, academic achievement should improve in all tracks. Proponents do not see early tracking as an obstacle to social mobility, arguing that the assignment of pupils to tracks should depend on parental background only to the extent that pupils’ academic ability or aspirations depend on background. Indeed, early tracking may enable some high-ability students from disadvantaged backgrounds to overcome the obstacles created by their background and therefore even enhance mobility (see The efficiency and equality debate).

The equality argument

The opponents of an early start to school tracking base their arguments on considerations of both fairness and efficiency. The concern about fairness arises from the arbitrariness of early tracking decisions. Tracking is usually based on test scores, grades, or teacher recommendations. These are all noisy measures of academic ability, and they are likely to be especially noisy in the early years. This measurement problem raises the concern that early tracking decisions are driven by family background. That means that early tracking could be an obstacle to intergenerational social mobility and could aggravate the persistence of both labor market and educational inequalities.

Opponents further argue that the claimed efficiency gains from tracking may fail to materialize. Student educational achievement depends on the quality of the peer group around them as well as the quality of teaching. Tracking may reduce the quality of the peer group for students who end up in less demanding tracks. If low-achieving students are more sensitive to peer effects than high-achieving students, tracking may reduce average achievement by changing the composition of peer groups in ways that work to these students’ disadvantage. The opponents of early tracking also point out that resources tend to be allocated unequally across tracks: Typically, the less demanding tracks receive fewer resources than the more academically orientated tracks.

Because of these concerns related to the fairness and efficiency of early tracking, opponents argue that early tracking condemns many able children from disadvantaged backgrounds to lower educational achievement as well as lower lifetime earnings. Therefore, early tracking should lead to a higher correlation of educational and labor market outcomes across generations and have insignificant and potentially negative effects on average educational achievement.

Empirical evidence on the effect of school tracking on educational achievement and social mobility

The theoretical arguments for and against early tracking suggest that it comes with an efficiency–equality trade-off. Proponents argue that early tracking should increase average educational achievement, while opponents assert that early tracking could increase the inequality of educational achievement and make achievement more dependent on parental background. This section assesses the empirical evidence for both of these claims.

Various strategies have been used to study the effect of early tracking on mobility and average achievement. The key methodological problem is that variation in tracking age needs to be uncorrelated with unobservable determinants of achievement such as the quality of schools or the student population. There are essentially two sources of variation in tracking age, and both create problems for this type of study:

  • The first is the variation in tracking age across countries and sometimes across regions within countries. When this kind of variation is used, the analysis needs to take into account other determinants of achievement at the country or regional level that are correlated with tracking age.

  • Second, many countries have gone through educational reforms that included changing the tracking age. Several studies have used these reforms to try to infer the effect of tracking age. However, this approach assumes that it is really the change in tracking age—and not other changes associated with the reforms—that is driving the effects.

The choice of outcome variables also differs across studies. Some studies focus on short-term outcomes, such as test scores in secondary school. These outcomes have the advantage that they are directly linked to tracking decisions. Furthermore, there are international test scores that provide comparable measures of achievement across countries, such as the Program for International Student Assessment (PISA), Progress in International Reading Literacy Study (PIRLS), and Trends in International Mathematics and Science Study (TIMSS). To get a fuller picture of the effects of tracking, it is also instructive to examine longer-term outcomes, like final educational attainment or labor market outcomes. These measures are more commonly used in studies of social mobility.

Early tracking and social mobility

Some studies that use international test score data, such as PISA and TIMSS, to study the effect of early school tracking on inequality rely on cross-country variation in initial tracking age to identify the effect of early tracking [1], [2]. Although the studies differ in details, their common conclusion is that early tracking increases inequality in test scores. The study that examines how the dispersion of test scores changes across grades finds that the dispersion of test scores increases more in countries where tracking starts between primary and lower-secondary school than in countries where it starts later [1]. The study of the association of test scores from PISA and PIRLS with family background (proxied by number of books in the household) finds that this association is stronger in countries where there is early tracking [2]. Another study that uses test scores from TIMSS finds very similar results [3].

A 2013 study uses test score data to analyze within-country reform [4]. The study looks at the effect of the Finnish comprehensive school reform on cognitive tests taken during the mandatory Finnish military service. This reform raised the initial tracking age from 11 to 16 and was rolled out at different times across regions, so the effect of the change in tracking age on test scores can be estimated while controlling for regional differences in educational attainment. The reform increased test scores among recruits from families with little formal education and had no effect on recruits from highly educated families. This result again suggests that early tracking leads to a stronger association between educational achievement and family background.

Studies of intergenerational mobility typically focus on the cross-generational association between measures of adult status rather than of the educational achievement of adolescents. To get a fuller picture of the effect of early tracking on mobility, it is therefore useful to look at the effect of early tracking on the association between adult outcomes and parental background. Whether early tracking accentuates the association between family background and these kinds of long-term outcomes has also been investigated using the time variation in tracking policies in several countries in 2007 [5]. The study finds a stronger association in countries with early tracking between parental background and long-term outcomes, such as final educational attainment and adult earnings.

Other studies focus on long-term outcomes using educational reforms within countries to estimate the effect of initial tracking age. One study finds that the comprehensive Swedish reform, which postponed tracking by two years, had a positive effect on the earnings and final educational attainment of children from low-education backgrounds [6]. Another estimates the effect of the Finnish comprehensive school reform on the association between the earnings of parents and their children, the most commonly used measure of intergenerational mobility [7]. The reform, which postponed tracking by five years, significantly increased intergenerational income mobility.

Early tracking and efficiency

Quite a few studies focus on the effect of early tracking on inequality. A much smaller number address its effects on efficiency by looking at average educational outcomes, largely because of the econometric problems associated with estimating the effect of tracking using cross-country data. Typically, these studies cannot separately identify the effect of tracking from other country-level characteristics.

One exception is a 2006 study that exploits differences in initial tracking age in different countries to investigate the effect on test score dispersion across grades [1]. This strategy also makes it possible to estimate the effect of early tracking on average test scores. Early tracking has no clear effect on score levels, and the study concludes that the hypothesis that early tracking has no effect on this outcome cannot be rejected.

While the use of cross-country data makes assessing the efficiency argument very difficult, within-country educational reforms allow researchers to estimate the effect of changes in tracking policies on the level of educational attainment. The study of the efficiency effects of the Finnish comprehensive school reform using Finnish military test scores as outcomes finds modest positive effects of the reform for one of the three tests used by the military [4].

Postponing tracking by five years had a small positive average effect on the verbal skills of the Finnish military recruits. Using a similar strategy to look at the effects on long-term outcomes in Sweden, a study finds that the comprehensive school reform had no significant effects on educational attainment or adult earnings [6]. The only study that finds clearly negative effects from postponing tracking shows a higher probability of dropping out among low-achieving students in Sweden when tracking is postponed beyond the end of compulsory schooling [8].

Is there an efficiency–equality trade-off?

The theoretical arguments for and against early school tracking suggest that policymakers face a trade-off between the efficiency of the educational process and the equality of educational opportunity offered to students in different tracks. In theory, making schools more homogeneous by tracking students early can boost average educational achievement, but at the cost of increased intergenerational persistence of educational and labor market outcomes. How much weight should be placed on equality effects and how much on efficiency effects depends on the social preferences of the policymakers. If intergenerational mobility is viewed as very desirable, theory posits that policymakers could choose to postpone tracking even at the cost of average educational achievement.

However, in the light of the empirical results summarized here, it is legitimate to ask whether there is any relevant trade-off to speak of. The evidence shows quite clearly that early tracking leads to a stronger association between child outcomes (whether test scores during adolescence or educational and labor market outcomes during adulthood) and parental background. And there is no empirical evidence that early school tracking improves average achievement. Nor is there evidence that the educational achievement of any subgroup will dramatically benefit from early tracking. Thus, these studies suggest that higher social mobility may be achieved by postponing tracking without imposing any large negative effects on average educational attainment.

Limitations and gaps

The empirical findings on the effects of early tracking on average achievement and intergenerational social mobility come from studies using observable variations in tracking across countries and regions or within countries over time. But these results might confound the effects of tracking with other, unobservable factors. Ideally, randomized experiments would study the effect of initial tracking age on educational achievement. Since school tracking is a characteristic of public school systems and difficult to change without changing the system, such experiments are not very feasible.

There are a few studies based on randomized experiments, but they involve ability grouping within schools rather than school tracking and are not widely generalizable. A study summarizing experiments with ability grouping within schools in American secondary schools before the early 1970s finds little evidence of any effect on student achievement [9]. More recently, an experiment in Kenya in 2011 finds that ability grouping with random teacher assignment and the same curriculum across groups benefits all students—strong evidence for narrowly defined ability grouping—but the results might not apply to developed countries [10]. Further, as school tracking implies different curricula and nonrandom selection of teachers across tracks, it is unclear how informative the evidence from ability grouping experiments is for school tracking policies.

Summary and policy advice

School tracking is a controversial policy topic, since the main arguments for and against it suggest that early tracking is beneficial for average educational achievement but potentially harmful for intergenerational social mobility. That puts educational policymakers in the difficult position of having to make judgments about an efficiency–equality trade-off.

However, recent empirical evidence on the effect of early tracking on average educational achievement suggests these effects are very modest or even slightly negative. At the same time, there is consistent evidence from studies applying different strategies and using data from many countries that early tracking leads to higher persistence of economic and educational status across generations. Thus, there is no empirical evidence to support the theoretical argument that policymakers face an efficiency–equality trade-off.

If increased social mobility is desired, recent empirical research shows that delaying tracking can enhance social mobility without impairing the quality of learning. However, the evidence also shows that moving to late tracking is no panacea for arriving at the high achievement levels that some late-tracking countries have reached.

Postponing tracking should be seen as a policy to address concerns about fairness and not necessarily as a way of improving average educational attainment.


The author thanks two anonymous referees and the IZA World of Labor editors for many helpful suggestions on earlier drafts.

Competing interests

The IZA World of Labor project is committed to the IZA Guiding Principles of Research Integrity. The author declares to have observed these principles.

© Tuomas Pekkarinen

evidence map

School tracking and intergenerational social mobility

Full citation

Full citation

Data source(s)

Data type(s)