Gender Diversity, Decision Rules, and the Risk Appetite of Teams

Project Start:03/2017
Researchers:Renée Adams, Andreas Grunewald, Ferdinand von Siemens
Category: Law and Finance, Experiment Center
Funded by:LOEWE

Former EU Commissioner for Competition, Kroes, famously said “[…] if Lehman Brothers had been ‘Lehman Sisters,’ would the crisis have happened like it did? No.” The rationale was that women are more risk averse and give different perspectives than those of their male colleagues. For similar reasons, commission Barnier proposed mandatory gender quotas on bank boards in 2011 (Treanor 2011), and the Commission added targets for the underrepresented gender in the management body to its 2014 Capital Requirement Directives. The same argument is often brought up when promoting gender diversity in organizational groups outside the board room. Experimental evidence indeed suggests that women on average are more risk averse than men (Croson and Gneezy 2009; Bertrand 2010). However, few papers investigate how individual risk preferences of group members affect the risk preferences of their group (Schupp and Williams 2008, Ambrus, Greiner, Pathak 2015, and Masclet, Colombier, Denant-Boemnt, and Lohéac 2009). Moreover, the rules used by groups to come to their group decisions could influence the aggregation of individual preferences, although Goeree and Yariv (2011) indicate that how strongly rules affect outcomes is unclear if decisions follow deliberation. Rules can of course be set or at least influenced by organizations. It is thus important to know which rules allow gender diversity to beneficially impact group decisions, and whether decision rules matter at all. With our experimental study we would like to investigate how the gender and the individual risk preferences of group members influence risky group investment decisions. We also want to study which role different group decision rules play in the aggregation of preferences. We hope to contribute to the existing literature in two ways. First, we systematically vary the gender and risk preference composition plus the decision rules in groups. Second, we are planning to collect sufficient amounts of data to detect the interaction of gender and decision rules.