Tiger Women: An All-Pay Auction Experiment on the Gender Heuristic of the Desire to Win
Peking University - HSBC Business School
Zhuoqiong (Charlie) Chen
London School of Economics & Political Science
May 5, 2013
Many real effort experiments have shown that males are more competitive and less risk averse than females. We therefore hypothesized that maleness would be a heuristic for “desire to win” (DTW) and lower risk aversion in “common value” all-pay auctions, where ability and confidence in ability cannot be confounds. We informed paired subjects of the opponent’s gender and developed theory to separate valuation from risk attitude in willingness to pay in FF, FM, MF, FF treatments, where F=female and M=male. However, in five separate experiments with nearly 600 subjects at two different tiers of universities, we never found that maleness signaled higher DTW and lower risk aversion. In the 400 subjects, where we controlled for self-selection into the experiment, we consistently found that femaleness signaled higher DTW. Among elite graduate students, femaleness even signaled lower risk aversion. Though like prior contests/auction experiments, women bid higher than men (on average), they also made more money. Our design can distinguish overbidding from strategic high bidding, as well as mixed strategy Nash equilibrium (MNE) play from random bidding, demand effects, and omitted preferences. Unlike prior work that identified MNE play using relative frequencies of mixing, we used the comparative statics induced by our gender pairings. We provide initial evidence that women may be more competitive than men (outside of verbal tasks), as measured by DTW, if ability, confidence in ability, risk attitude, self-selection, and demand effects are fully controlled for.
The appendices for this paper are available at the following URL: http://ssrn.com/abstract=2141181.
Number of Pages in PDF File: 41
Keywords: experiments, all-pay auction, overbidding, gender differences, heuristic, stereotype
JEL Classification: C91, D44, J16, I20working papers series
Date posted: December 30, 2011 ; Last revised: December 17, 2013
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo5 in 0.329 seconds