Abstract

Abstract Using the data on the boards of directors of public companies across the world, we show that female representation in corporate boards does not affect the value of the company. To identify the causal effect of interest we use a novel instrument that exploits natural discontinuities in the gender composition of the board. We further explore if this zero result is due to the lack of power or due to the effect heterogeneity. By splitting the sample into groups of companies that could be expected to have different (or even opposite) effects, we find no heterogeneous effects. This holds even in specifications with reasonably high first-stage F-statistics, where power is likely not an issue. We further show that percentage of women may still affect certain board characteristics (qualifications, independence, quality of governance, etc). However, as there is still no effect on performance, this may imply that the exact composition of the board does not matter in the end. Authors: Olga Kuzmina (New Economic School) and Valentina Melentyeva (University of Mannheim)