Documentos de trabajo (working papers) · 2016
Isolating causality between gender and corruption: An IV approach
We address the persistent reverse causality problem in estimating the causal effect of female labor force participation (FLFP) on corruption. Employing plow usage as an instrumental variable, an increase in FLFP by one standard deviation is suggested to improve the Corruption Perceptions Index (CPI, ranging from zero to ten) by 0.52 points. This effect is stronger than a one standard deviation change of education levels, government size, or ethnic fractionalization.