A meta-regression analysis on intergenerational transmission of education: publication bias and genuine empirical effect

Nicolas Fleury (1, 2) , Fabrice Gilles (2, 3)

1 Centre Etudes & Prospective
2 EQUIPPE - Economie Quantitative, Intégration, Politiques Publiques et Econométrie
3 TEPP - Travail, Emploi et Politiques Publiques

In this article, we evaluate to what extent parental education impacts the education of their children by using a meta-regression analysis. Since the mi-1970s, there is a large and growing literature that deals with the causal impact on parental education on children’s education. Those studies exhibit a large range of values for the education transmission coefficient. We consider an alternative way to estimate a true effect of parent education, discussing the existing empirical literature by using a meta-regression analysis. Our database is composed of a large set of both published and unpublished papers written over the last 40 years (1974-2014). This database allows us to econometrically evaluate an effect of parents education on their children, irrespective of articles heterogeneity (data sources, included explanatory variables, econometric strategy, type of publication), and of publication bias. We find evidence for both a publication bias and a large transmission coefficient of education.

Mots clés: education; intergenerational transmission; meta-regression analysis.

Domaines: Economies et finances

Nicolas Fleury, Fabrice Gilles. A meta-regression analysis on intergenerational transmission of education: publication bias and genuine empirical effect. 2015. ⟨halshs-01143490⟩