International gender wage gap: A Meta-analysis
|Author:||Mgr. Tereza Navarová|
|Year:||2022 - winter|
|Leaders:|| prof. PhDr. Tomáš Havránek Ph.D.
|Work type:|| Masters
|Awards and prizes:|
|Abstract:||This thesis uses modern meta-analysis methods to produce a systematic
quantitative review of the literature on the international gender wage gap. After
collecting 661 estimates from 51 peer-reviewed studies, estimates from published
studies are subjected to several tests studying the presence of publication bias.
To test for publication bias, standard errors of estimates are approximated for
studies that do not report standard errors or other confidence measures. Based
on both linear and non-linear formal tests, the presence of significant publication
bias is not detected. Furthermore, Bayesian model averaging is performed to
explain the heterogeneity in the collected data, with robust control performed
using frequentist model averaging. The results suggest that omitting variables
related to human capital leads to the estimation of higher average gender wage
differences. In contrast, authors who focus subsamples of the labour market
(such as college graduates or only fulltime workers) estimate on average smaller
gender wage differences.