October 2015

No. 9447: Gender and Corruption: A Reassessment

Julia Debski, Michael Jetter

This paper analyzes the relationship between gender and corruption, controlling for country-specific heterogeneity in a panel framework. Using annual observations in a pooled setting (no country-fixed effects) confirms the positive link between the involvement of women in society and the absence of corruption. However, once country-fixed effects are acknowledged, only the share of female employers remains a positive and statistically meaningful correlate. Nevertheless, the derived magnitude is negligible in a global sample. Analyzing potential nonlinearities reveals that this effect is driven by African nations, where a one standard deviation increase in the share of female employers is related to a substantial decrease of corruption by 2.5 index points (scale from zero to ten). Surprisingly, the link between the share of women in the labor force and the absence of corruption becomes negative once country unobservables are accounted for. Taken together, these findings cast doubt on a general, global relationship between gender and corruption.