When the money is on the line in Final Jeopardy, women bet no more cautiously than men. That is the conclusion of a study of thousands of wagers on the American quiz show, spanning episodes almost 40 years apart.
The paper, by economists Michael A. Leeds of Temple University and Eva Marikova Leeds of Moravian University, was published online in the Eastern Economic Journal on April 7, 2024 (read the paper). Its title is a nod to the show itself: “I’ll Take Gender Differences for $400.”
Why a quiz show is a useful test
Many laboratory studies find that women are more risk averse than men and more likely to avoid winner-take-all contests. If that holds in the real world, it could hold women back in careers where promotions work like tournaments. But a growing body of work suggests the gap shrinks or disappears when women operate in a field where they feel confident.
Jeopardy! offers a natural test of that idea. Contestants are anything but random: according to the paper, roughly 70,000 people take the qualifying test each year and about 400 make it onto the show. They have chosen to compete, and they know they are good at it. Final Jeopardy then forces each player to write down a wager, based on the category and the scores, before seeing the clue.
The authors built on an earlier model of optimal betting and focused on two situations. In a “runaway”, the leader has more than twice the score of the second-place player and only needs to bet small enough to stay safe. In closer games, the second-place player must decide how much to risk, and in some cases faces a clear choice between “going low” with a small bet or “going high” with a large one.
Nearly 2,200 episodes, two eras
The data came from the J! Archive fan site. The main sample covers 1,669 regular episodes broadcast between September 16, 2013, and November 23, 2021. A second sample covers 492 episodes from the first five seasons of the prime-time show, aired between September 13, 1984, and June 30, 1989. Gender was taken from contestant photos for recent seasons and from first names for the early ones, with ambiguous names dropped.
The researchers looked at bets by runaway leaders, bets by second-place players in close games, and whether second-place players chose to go high or low. They controlled for the scores of all three players, whether the contestant was a returning champion, whether they answered the final clue correctly, the share of women on the panel and the host’s gender.
No gender gap in risk taking
None of the three tests showed women betting more cautiously. The only direct gender effect ran the other way at first: women leading in runaways bet more than men, though the gap narrowed as scores rose, and women only began to bet less above $27,750, a score higher than the average entering Final Jeopardy. Neither specification of the go-high-or-go-low analysis found women choosing the safe option more often than men.
The result was the same in the 1980s data. In the authors’ words, the finding that women do not show greater risk aversion “continues to hold” across the generation gap.
Confidence mattered for everyone. Answering the final clue correctly, the study’s main proxy for confidence, was linked to bets roughly $1,200 to $1,400 higher among runaway leaders. A few gender differences turned up in the interactions, such as women leading runaways betting more when more women were on the panel, but these were weak and showed no consistent pattern.
The authors note that the data lack contestants’ age, education and, for the early seasons, race. Their reading is that women who select into a contest they feel equipped for behave like the men beside them, and on Jeopardy! that has been true since the Reagan years.
Study Details:
- Title: I’ll Take Gender Differences for $400: Using Jeopardy! to Analyze Attitudes Toward Risk
- Authors: Michael A. Leeds, Eva Marikova Leeds
- Journal: Eastern Economic Journal
- Publication Date: April 7, 2024
- DOI: 10.1057/s41302-024-00270-9