In almost every educational system in the world, students from higher socioeconomic statuses (SES) outperform their socioeconomically disadvantaged peers. Despite decades of policies targeted towards narrowing the SES- achievement gap, in many countries, the gap is widening (Chmielewski, 2019). By now, it is well established that teacher competence is one of the most important school level inputs when it comes to student outcomes, especially for socioeconomically disadvantaged and low-performing students (Darling-Hammond, 2000; Goe, 2007). A host of research has confirmed the importance of experience on the job and subject matter specialization for mathematics and science student performance, in particular (Rice, 2003; Rockoff, 2004; Goe, 2007). Recently, scholars have turned their attention to the potential role of teacher sorting in inequitable educational outcomes. Many comparative studies within the United States have evidenced the phenomenon of inequitable teacher sorting, where teachers with lower competence tend to be concentrated in socioeconomically disadvantaged schools and areas (Clotfelder et al., 2008; Bacolod, 2007; Akiba et al., 2007). Several nationally and internationally focused studies outside the US have confirmed that in many countries, low-SES students tend to have teachers with lower qualification levels, including fewer years of experience, lower levels of subject-matter specialization, and fewer years of education (Sims and Allen, 2018; Glassow and Jerrim, forthcoming; Glassow, Franck and Yang Hansen, forthcoming; Luschei and Jeong, 2019; Han, 2018). As of yet, however, the phenomenon of teacher sorting in its relation to educational inequity remains relatively under-explored from a cross-national perspective. International large-scale assessments such as the Trends in International Mathematics and Science Study (TIMSS) are uniquely positioned to investigate such questions, as they have a longitudinal design at the country level despite being cross-sectional at the student level (Gustafsson, 2013). We use three TIMSS waves from 2003, 2011 and 2019 eighth grade population and data from 40 educational systems in the current analysis. Following previous approaches such as that of Hanushek et al. (2013), we employ a country- fixed effects approach to account for unobserved heterogeneity by accounting for country mean variation in our dependent variable. The main purpose of the paper is to examine whether degrees of inequitable teacher sorting affect the performance gap between high- and low-SES students, examining moderating effects for a host of factors at the country level, including socioeconomic school segregation, income inequality, GDP per capita, educational tracking, within- school ability grouping, and size of the private sector. Preliminary results show that more inequitable teacher sorting predicts larger socioeconomic achievement gaps in mathematics, particularly in countries with early educational tracking. The results are less straightforward for science, with no relationship between teacher sorting and the socioeconomic performance gap. The results of this study provide the first causal international evidence for the link between teacher sorting and the widening SES-achievement gap. We discuss the methodological strengths and limitations of the study, and position the results in line with their implications for policy and future research.

In recent years, scholars have turned towards tackling unobserved heterogeneity at the country level through longitudinal analyses of international large-scale assessment data (Hanushek et al., 2013; Gustafsson, 2013; Teltemann and Schunk, 2019). This is largely because in order to estimate causal effects, one must account for vastly different economic, cultural, and societal settings which may influence the variable of interest. For example, in relation to the current study, teacher sorting may be correlated with other unobservable aspects of national education systems which might affect student outcomes. While this study attempts to account for possible confounders at the country level as well as include country level moderators, it is not realistic to include all conceivable confounders. Thus, a country fixed effects approach is used, where mean between-country differences in the SES performance are accounted for in the model. We are therefore only interested in the within country variation of our dependent and independent variables, i.e. the change in the SES achievement gap in relation to the change in degrees of teacher sorting over time within countries. In order for the model to be well founded, it is important that the independent variables of interest change over time which is a factor we explore in the study. Last, fixed effects models are not immune to bias from within-subject changes in other unobserved variables over time, which presents one possible limitation to the current paper which we address. TIMSS employs a two stage stratified sampling design, which samples schools according to previously determined strata and their size and whole classrooms to cover a range of nationally representative educational contexts. TIMSS has minimum participation requirements for a country to be included, which calls for a minimum of 150 schools to be sampled per grade, and a minimum of 4000 students. We measure teacher sorting (our main independent variable of interest) as the between school intraclass correlation coefficient in average teacher qualification levels, including teacher education, subject-matter specialization, and teacher experience levels. Our dependent variable is measured as the difference in performance points between high- and low- SES students in both mathematics and science. Following Chmielewski (2019), low- and high- SES groups are determined by 90/10 percentiles for each country-year on the Home Educational Resources scale provided by the IEA. We account for the missing data in socioeconomic status variables through multiple imputation and also consider plausible values. Analyses are conducted in RStudio.

It is notoriously difficult to measure teacher effectiveness, and especially so in international large-scale assessments. Previous international research on the topic of inequitable teacher sorting (also termed ‘the teacher opportunity gap’) and inequity in student outcomes remains inconclusive (Akiba et al., 2007). Additionally, there remains considerable debate as to why SES achievement gaps are widening, and no recent internationally focused studies have investigated the role of teacher sorting. However, new methods, attitudes, and approaches to international large-scale assessment data have been put forth over the past decade which generate new possibilities for investigating this decades-old question. This paper presents some of the first international evidence for the role of inequitable teacher sorting in the socioeconomic performance gap. While longitudinal data allow us to control for unobserved heterogeneity at the country level, the data are far from perfect, with some countries showing high degrees of missingness on the socioeconomic status variables. This is a challenge which we discuss in further depth in the article. We believe, however, that the strengths of the present study significantly outweigh its limitations. Although many policy-makers have shifted from emphasizing equality of outcomes to the importance of equality of opportunity, teacher sorting represents one aspect of educational settings which remains considerably unequal and which may play a key role in perpetuating the difference in performance between diverse socioeconomic groups. Previous research has shown that teachers prefer working conditions over salary incentives (Bacolod, 2007). As such, educational systems should continue to pay attention to the problem of teacher sorting if they wish to minimize the SES performance gap, which is ultimately the problem of improving working conditions in socioeconomically disadvantaged schools.

Akiba, M., LeTendre, G.K., & Scribner, J.P. (2007). Teacher quality, opportunity gap, and national achievement in 46 countries. Educational Researcher, 36, 369–387. Bacolod, M. (2007). Who teaches and where they choose to teach: College graduates of the 1990s. Educational Evaluation and Policy Analysis, 29, 155-168. Chmielewski, A. (2019). The global increase in the socioeconomic achievement gap, 1964-2015. American Sociological Review, 84, 517-544. Clotfelter, C., Glennie, E., Ladd, H., & Vigdor, J. (2008). Would higher salaries keep teachers in high-poverty schools? Evidence from a policy intervention in North Carolina. Journal of Public Economics, 92, 1352–1370. Darling-Hammond, L. (2000). Teacher quality and student achievement: a review of state policy evidence. Educational Policy Analysis Archives, 8, 1-44. Glassow, L.N., and Jerrim, J. (Forthcoming). Glassow, L.N., Franck, E., & Yang Hansen, K. (Forthcoming). Goe, L. (2007). The link between teacher quality and student outcomes: A research synthesis. National Comprehensive Center for Teacher Quality, Washington, DC, USA. Retrieved from: http://www.gtlcenter.org/sites/default/files/docs/LinkBetweenTQand StudentOutcomes.pdf Gustafsson, J.E. (2013). Causal inference in educational effectiveness research: a comparison of three methods to investigate effects of homework on student achievement. School Effectiveness and School Improvement, 24, 275-295. Han, S.W. (2018). School-based teacher hiring and achievement inequality: a comparativeperspective. International Journal of Educational Development, 61, 82-91. Hanushek, E.A., Link, S., & Woessman, L. (2013). Does school autonomy make sense everywhere? Panel estimates from PISA. Journal of Development Economics, 104, 212-232. Luschei, T.F., & Jeong, D.W. (2019). Is teacher sorting a global phenomenon? Cross-national evidence on the nature and correlates of teacher quality opportunity gaps. Educational Researcher, 47, 556-576. Rice, J. (2003). Teacher quality: understanding the effectiveness of teacher attributes. Washington, DC: Economic Policy Institute. Rockoff, J. (2004). The impact of individual teachers on student achievement: Evidence from panel data. American Economic Review, 94, 247–252. Sims, S., & Allen, R. (2018). Do pupils from low-income families get low-quality teachers? Indirect evidence from English schools. Oxford Review of Education, 44, 441-458. Teltemann, J., & Schunk, R. (2016). Education systems, school segregation, and second-generation immigrants’ educational success: evidence from a country-fixed effects approach suing three waves of PISA. International Journal of Comparative Sociology, 57, 401-424.