pay transparency, gender gap, wages, policy-design, model averaging, meta-analysis

Pay transparency laws are a key policy response to persistent gender wage disparities, yet evidence on their effectiveness is mixed. This meta-analysis synthesizes 268 estimates from 12 studies. Across a broad suite of publication bias diagnostics, we find at most weak evidence of selective reporting, while most approaches indicate a small but significant positive effect beyond bias. The pooled mean effect is 0.012 log points, corresponding to an average 1.2% increase in women’s wages relative to men, consistent with a modest narrowing of the gap. Heterogeneity analysis using Bayesian and frequentist model averaging shows that policy design is pivotal. Public disclosure regimes produce larger reductions than internal access or job-ad disclosure, while evidence for pay-secrecy bans is imprecise. Specification choices also matter, with regional and employee controls attenuating effects and sector controls amplifying them. Overall, effective transparency depends on both robust policy design and careful empirical specification.