Abstract
Cross-country evidence on student achievement might be hampered byomitted country characteristics such as language or legal differences.This paper uses cross-state variation in Germany, whose sixteen statesshare the same language and legal system, but pursue different educationpolicies. Education production function models are estimated usingstate-level PISA-E data, where possible pooling three subjects and threewaves to obtain up to 138 test-score observations. The same resultsfound previously across countries hold within Germany: Higher meanstudent performance is associated with central exams, private schooloperation, and socio-economic background, but not with spending, whilehigher equality of opportunity is associated with reduced tracking. Inmodels that pool German states with OECD countries and combine up to 54state and country observations, these institutional determinants do notdiffer significantly between the sample of German states and the sampleof OECD countries, indicating that the existing cross-country evidenceis not substantially biased by unobserved country-specific factors.
Dokumententyp: | Zeitschriftenartikel |
---|---|
Fakultät: | Volkswirtschaft
Volkswirtschaft > Lehrstühle > CESifo-Professur für Empirische Innovationsökonomik |
Themengebiete: | 300 Sozialwissenschaften > 330 Wirtschaft |
Sprache: | Englisch |
Dokumenten ID: | 19681 |
Datum der Veröffentlichung auf Open Access LMU: | 15. Apr. 2014, 08:53 |
Letzte Änderungen: | 29. Apr. 2016, 09:17 |