Performance uncertainty and the level of female participation on tournaments

Project Information

GAPINCOMPETITION

Grant agreement ID: 101064641

DOI

10.3030/101064641

Project closed

EC signature date 23 May 2022

Start date 15 August 2022

End date 14 August 2024

Funded under

• Marie Skłodowska-Curie Actions (MSCA)

Total cost

No data

EU contribution

€ 171 521,28

Coordinated by

MIDDLE EAST TECHNICAL UNIVERSITY
Türkiye

Periodic Reporting for period 2 - GAPINCOMPETITION (Performance uncertainty and the level of female participation on tournaments)

Reporting period: 2023-08-15 to 2024-08-14

Summary of the context and overall objectives of the project

In business, government, academia, and the nonprofit sector, it is commonly observed that women enter competitive positions less often than men. There are two possible reasons for this behavior: 1) Women prefer competitive positions less than men. 2) Women are discriminated for competitive positions. Niederle and Vesterlund (2007) and follow up literature related to their work found that there is a gender gap in tournament entry and this gap is largely explained by traits like competitiveness and overconfidence but negligibly by risk aversion. Recent studies by Gillen et al. (2019) and Van Veldhuizen (2022) challenge this result by showing that gender gap in tournament entry can be explained by risk aversion difference fully.

To understand how gender gap differs in various environments, and whether information can mitigate this gender gap, the first part of the project focuses on symmetric environments between contestants when they know their own and others' skills fully but not know about the noise in the environment. In particular, by using a rank order tournament, when the performance is calculated as the sum of effort and noise, we analyze males' and females' tournament entry rates when the noise is drawn from a known or unknown distribution and when the noise is drawn once or twice (when it is drawn maximum of draws is taken into account for calculating the performance).

The second part of the project focuses on whether gender gap persists in competitive choices when there is an asymmetry between contestants. This is important since not all the competitions in real world are symmetric, we need to learn how different genders respond to asymmetry when they know the size of asymmetry. We create asymmetry in two different ways by using two components of a performance in a 2-person rank-order tournament: either one of the contestants has higher skill than other when both contestant have the same chance or when one of the contestants has higher chance than the other when both contestants have the same skill.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

In the first part of the project, by using a novel method, we control subjects' skills and analyze the role of noise on entry in a rank-order tournament. In our 2x2 experiment, we change whether the noise is drawn from a known or unknown distribution or whether the noise is drawn once or twice (maximum of them are taken into account when it is drawn twice). We expect no effect of the number of draws on tournament entry, decrease in entry when the noise is drawn from an unknown distribution, and there is no role of competitiveness trait in subjects' tournament entry choices as in Van Veldhuizen (2022). We have 4 main insights from our experiments: First, we find no effect of the number of draws. Second, when the noise distribution is unknown, females enter tournaments less often than when it is known. However, males' entry is not affected by whether the noise distribution is known or unknown. Furthermore, we find that although there is a gender gap in tournament entry when the noise distribution is unknown, there is not when it is known. Finally, unlike Van Veldhuizen (2022), we find that although both males and females choose a tournament more than the lottery with the same payoff value when the noise distribution is known, only males do so when the noise distribution is unknown.

In the second part of the project, we analyze how males' and females' tournament entry choices differ when they have asymmetric skills controlling for noise or asymmetric noise controlling for skill compared to baseline (symmetric contest, where both contestants have the same skill and noise). In this part of the project, the noise is drawn from a known distribution. As found in the first part of the project, here, we also find that when the skills of the contestants are the same and noise is drawn similarly for both contestants from a known distribution, there is no gender gap. Furthermore, when the noise is drawn from known distribution, gender gap disappears even under asymmetry: tournament entry choices of males and females are similar when they are advantageous or disadvantageous positions for chance or skill. Then we look at how asymmetry affects tournament entry choices compared to baseline. We find that although being advantaged does not lead higher entry in asymmetric skill and asymmetric chance treatments compared to baseline, being disadvantaged decreases entry in asymmetric skill but not in asymmetric chance treatments. This behavior stems from the beliefs. Although advantaged (disadvantaged) players in asymmetric skill treatment believe that their probability of winning is significantly higher (lower) compared to symmetric treatment, these do not hold in asymmetric chance treatment.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The project contributes to state of the art in a couple of ways: The first is methodological. We use a carefully calibrated real-effort task (a variation on the ball-catching task of Gachter et al., 2015) to control the ability component of participants’ final performance. To that end, we program the speed with which the (digital) balls drop on the screen and calibrate the time limit on the task so that, with considerable ease and minimal cognitive effort, every subject can catch exactly the maximum possible amount which differed in the second part of the project when we create asymmetry in skill. This calibration ensures that in the symmetric treatment (when all subjects have the same skill) the variation in final performance is driven (almost) entirely by a random noise component, allowing us to mitigate one of the primary impacts of overconfidence on the gender gap in tournament entry.

The first part of the project contributes to the literature analyzing the role of uncertainty or ambiguity on males' and females' tournament entry choices. As far as we know, only Balafaoutas and Sutter (2019) look at the role of uncertainty and ambiguity regarding the number of winners in a tournament on the entry decisions of males and females. They find that ambiguity increases the performance of men who choose to compete but has no such effect on women. Ambiguity also increases the entry rate of men, further contributing to a gender gap in the rate at which men or women win the tournament. In contrast with their approach, our study focuses on the effects of ambiguity and uncertainty with respect to the noise component of participants’ final performances in the tournament.

The second part of the project contributes the literature analyzing the role of asymmetry on males' and females' tournament entry choices by varying the asymmetry type in two components of performance: skill or chance. Although earlier literature shows discouragement effect (decreasing effort) in tournament due to possible asymmetry in contests, none of the studies used rank order tournament and analyzed the role two different asymmetries when subjects know about the extent of the asymmetry.

Random participants of our Behavioral and Experimental Lab at METU