The Impact of Immigrants on the Educational Achievement of Natives: Evidence From PISA (Programme for International Student Assessment) Panel Estimates

• In Event: “Using the World as an Educational Laboratory”: Merging Educational Effectiveness Research and International Comparative Studies

Abstract

Objective
Immigrants are an important educational issue in many industrialized countries, especially because of demographic trends toward higher shares of immigrants in the education system. We estimate the effects of immigrants on the educational achievement of natives. We also explore whether the effects of immigrants on natives’ achievement depend on the institutions of the educational system, such as school autonomy and central exit exams.

Theoretical Framework
Existing research has shown that immigrants perform significantly worse than natives in reading and in mathematics in almost all immigration countries. This suggests that immigrants might exert negative peer effects on their native classmates. Furthermore, the educational institutions of a country might mediate the impact of immigrants on natives’ achievement since immigrants’ performance has been shown to be related to educational institutions, such as central exams, pre-primary education, and time in school (Schneeweis 2011).

Method
We employ panel estimation methods with country-fixed effects to estimate the effects of immigrants on the educational achievement of natives (see Hanushek et al. (2012) as an example of this methodology). We aggregate the immigrant share at the country level to remove any bias from within-country sorting of individuals. Because we exploit variation in immigrant shares within countries over time by using country fixed effects, we control for unobserved time-invariant country heterogeneities. To investigate whether the impact of immigrants on natives’ performance depends on educational institutions, we add to the model an interaction term between immigrant share and the country-specific institutions. This interaction is possible only with cross-country data because educational institutions typically vary only across countries. In all models, we control extensively for measures of student characteristics, family background, and school characteristics. Furthermore, we also include time-varying characteristics of the country, such as GDP per capita and expenditures per student.

Data
The student-level data come from the Programme for International Student Assessment (PISA) tests which contain internationally comparable measures on the educational achievement of 15-year-old students in reading, mathematics, and science. We use the four assessments in 2000, 2003, 2006, and 2009 and keep all countries that have been participating for at least three times, which leaves us with over one million students in 42 countries.

Results
We find a small negative impact of immigrants on natives’ achievement. This result is robust to different specifications. Concerning the influence of educational institutions, we find that the negative impact is smaller if schools have more autonomy.

Significance of the Study
Our approach overcomes several potential issues of migrant selection that typically prevents estimating the impact of immigrants on natives’ educational achievement. Our findings suggest that the negative impact of migrants is limited and that educational institutions seem to mediate this impact.

References
Hanushek, E.A., S. Link, L. Woessmann (2012): Does School Autonomy Make Sense Everywhere? Panel Estimates from PISA, Journal of Development Economics, forthcoming.
Schneeweis, N. (2011): Educational Institutions and the Integration of Migrants, Journal of Population Economics, 24(4), 1281-1308.

Authors

• Eric A. Hanushek, Stanford University

• Susanne Link, Ifo Institute for Economic Research

• Marc Piopiunik, Ifo Institute for Economic Research

• Ludger Woessmann, Ifo Institute for Economic Research