DP16761 The Gender Gap at the Top: How Network Size and Composition Impact CEO Pay

Author(s):	Sharmin Sazedj, José Tavares
Publication Date:	November 2021
Keyword(s):	CEO compensation, Firm sorting, Gender Gap, networks
JEL(s):	G34, J24, J30, L14
Programme Areas:	Organizational Economics
Link to this Page:	cepr.org/active/publications/discussion_papers/dp.php?dpno=16761

This paper advances the literature on the gender pay gap amongst top managers, by explicitly assessing the relevance of professional networks. We use data on the universe of firms in Portugal, where female top managers earn 25% less than their male counterparts, conditional on age, education and firm tenure. We estimate that 20% of the above femalemale earnings difference is due to differences in networks across gender. Making use of Gelbach's decomposition, we find that the network effect can be ascribed to firm sorting, i.e. well-connected managers tend to be associated to higher paying firms. By focusing on episodes of transitions between firms, and relying on a propensity score matching procedure, we estimate that around 90% of the gender pay gap emerges during the hiring process, and is only slightly aggravated thereafter, due to biased career progression. Roughly one third of the gender gap can be attributed to firm sorting, two thirds of which to differences in networks. We then examine the gender composition of female and male CEOs' networks. While we find no evidence that females benefit differently from network size, we do find evidence that male connections are more valuable. If, however, we proxy for the inner circle of a manager, taking into account the proximity of connections, we conclude that same gender connections gain relevance. These results suggest that connections between females do play an important role in the existing corporate framework where males are over-represented. We conclude that policies furthering female representation in leadership positions can have positive spillover effect for other women.