۱- Introduction

۲- Background

۳- Jeopardy! background and data

۴- Empirical strategy

۵- Empirical findings

۶- Discussion

۷- Conclusion

References

Abstract

Analyzing 8,169 contestants over 32 years of the US game show Jeopardy!, we find that women compete more aggressively, become (marginally) more competitive, and take on more risk when paired against men. Specifically, a woman is more likely to win a Jeopardy! episode and, within the show, to respond to a clue. Once responding, she is also marginally more likely to respond correctly. Potentially most surprising, even the sizeable gender gap in risk-taking (analyzing Daily Double wagering decisions) disappears once a woman competes in an otherwise all-male field of competitors. Men, on the other hand, wager significantly less when paired against women only, but the gender of opponents does not affect their competitive performance otherwise. Our rich sample allows us to control for a host of potentially confounding factors and player-fixed effects, thereby eliminating potential biases from unobservable individual characteristics. Our findings are consistent with an explanation that emphasizes an adaptation to “social norms” applied to gender.

Introduction

It has almost become a stylized fact that, on average, women are more likely to avoid competition, under-perform in competitive environments, and exhibit higher risk aversion than men. Persistent social phenomena, such as the gender wage gap or the under-representation of women in highly competitive occupations and job positions, have been linked to such observations. One prominent hypothesis to explain this phenomenon relates to the idea that the gender of one’s opposition could influence competitive behavior (e.g., see Gneezy et al., 2003, Booth and Nolen, 2012b, Booth et al., 2014). More generally, people may behave differently when competing against adversaries from the opposite sex. If true, this would imply wide-ranging consequences in a number of settings. For instance, numerous work environments are characterized by persistent under-representation of one gender. Women are especially under-represented in jobs that are generally associated with highpressure environments and large stakes, such as financial management (the share of females at Wall Street remains at approximately 10 percent) or CEO positions in the US (2.5 percent).1 Other areas with low female employee shares include IT- and math-related occupations, where women usually occupy less than 20 percent of positions (e.g., see Department of Labor, 2016). On the other hand, men are traditionally under-represented in professions such as nursing, teaching, social work, and counseling. Other prominent settings of gender separation include sports and at times education (e.g., single- versus mixed-gender educational environments).2 Unfortunately, real-life settings do not offer many opportunities to test whether the gender of opponents can influence performance and risk-taking since people likely select into gender environments they prefer (consciously or subconsciously) and causality remains difficult to isolate. In the following pages, we study data from the US edition of the popular game show Jeopardy! (J! from hereon) to analyze whether the exogenous assignment of male or female competitors (i.e., contestants cannot choose to compete against female or male opponents) affects women’s and men’s performance and attitudes to risk.