Culture, Institutions, and the Gender Gap in Competitive Inclination: Evidence from the Communist Experiment in China


Cultural differences can impact the gender gap in competitiveness, as evidenced by three ethnic groups with differing gender role expectations in China.


While opportunities for women have broadened significantly in the past few decades, differences in “competitive inclination” – a willingness to enter competitive situations – may prevent women from taking full advantage of this potential gain.

Two leading explanations for gender differences in labor market outcomes are discrimination and gender differences in preferences, which includes reluctance by women to seek out competitive positions or promotions and contributes to on-going inequality and inefficiency.

Previous studies have shown that this gender gap in competition differs significantly across cultures, which highlights the role of socialization in structuring gender norms. It also suggests that changing cultural gender norms may be able to increase women’s competitive inclination, thereby reducing economic inequality and increasing the pool of qualified potential candidates for high-level positions. A natural experiment for changes in cultural gender norms were the policy changes that arose with the Communist reforms in China that began in the 1950s.

This study included three ethnic groups in China, the Han Chinese, Yi, and Mosuo. Traditional Han Chinese society and Yi society were both patrilineal, with little economic independence for women. However, the Han Chinese were subject to Communist reforms while the Yi were not. In particular, the reforms increasing the minimum marriage age and marriage autonomy, which have the potential to change gender role expectations, did not apply to the Yi.  While Han Chinese women began to close the gap in education and workforce participation, Yi women continue to hold subordinate positions and have low workforce participation relative to men. The Mosuo were also not subject to the reforms but are a matrilineal society with longstanding egalitarian traditions. By comparing these three groups within a single high school similar across all other major demographics, the author was able to isolate the effects of culture on the gender gap in competitive inclination.


Gender gaps in competitive inclination between the three ethnic groups, as measured by an economic experiment, can be explained by the differences in exposure to the egalitarian gender reforms from the Chinese Communist period.

  • There were no gender differences in performance on the math task or on probabilities of winning the tournament for any ethnic group.
  • Among the Han Chinese ethnic group (a Communist-reformed patrilineal society with strong egalitarian gender norms), there was no difference in competitive inclination between men and women (likelihood of entering the tournament after controlling for the probability of winning, overconfidence, and risk aversion).
  • Among the Yi ethnic group (a non-reformed patrilineal society with weak egalitarian gender norms), there was a 23.7 percentage point difference in competitive inclination between men and women (likelihood of entering the tournament after controlling for the probability of winning, overconfidence, and risk aversion).
    • A sub-analysis among the Yi of agricultural (i.e. not affected by reforms) and non-agricultural households, showed the gender difference is driven entirely by the non-reform households, as women are 40.5 percentage points less competitive than men.
  • Among the Mosuo ethnic group (a non-reformed matrilineal society with strong egalitarian gender norms), there was still a gender gap with a 29.7 percentage point difference in competitive inclination between men and women (likelihood of entering the tournament after controlling for the probability of winning, overconfidence, and risk aversion).
  • Across ethnic groups, Han Chinese and Mosuo women were similarly competitive, but there was a 20 percentage point difference in competition between Han Chinese and Yi women.
    • Notably, the lack of difference between the Han Chinese and Mosuo suggests the high position of Mosuo women in traditional culture has similar effects to the institutional reforms on the position of Han Chinese women.
    • The larger gender gap among the Yi was due to the lower competitive inclination of Yi women.
  • Across ethnic groups, Han Chinese and Yi men were similarly competitive, while Mosuo men had the highest level of competitiveness.
    • Notably, the lack of difference between the Han Chinese and Yi men suggests that even if institutional reforms increased female competitive inclination, they did not appear to influence male competitive inclination.
    • One potential explanation for the higher Mosuo male competitive inclination might be due to their polygamous practices, which might increase their return on competitive behavior (in terms of the ability to produce additional children without any additional financial responsibilities).

The lack of a gender gap between Han Chinese men and women suggests that the gender egalitarian reforms of the Communist period increased women’s competitive inclination relative to men. By contrast, Yi women have significantly lower levels of competitive inclination compared to both Yi men and Han Chinese women. This difference between two traditionally patrilineal societies indicates that cultural factors like gender norms can impact competitiveness. Despite the existing egalitarian values among the Mosuo, with the Han Chinese and Mosuo women similarly competitive, the Mosuo men were still more competitive. Thus, the gender gap among the Mosuo is likely due to the overall higher levels of competitive inclination among Mosuo men relative to all other men and women in the study.


This study sampled 272 high school students (96 Han Chinese, 96 Yi, 80 Mosuo) in years 11 and 12 from a single high school in Ninglang. Participants were of similar demographics in terms of age, number of siblings, and geographical area. They were also similar in socio-economic status and academic performance.  This minimized potential confounding and allowed for a “clean” comparison to determine the influence of culture of the three different ethnic groups on competitive inclination.

A total of 12 hour-long sessions with 20-24 subjects each were conducted, with each session limited to one ethnic group from the same year in school and with an even divide across gender. Participants were asked to complete a competitive math task in three rounds, with the order of first and second rounds randomized between sessions. In the first round, participants were asked to add as many sets of five two-digit numbers as possible in five minutes and were compensated based on the number of correct answers.  In the second round, participants entered a “compulsory tournament” of the same task in groups of four. Only the participant with the most correct answers was compensated, while the other three participants in the group received nothing. In the final round, participants were offered a “discretionary tournament” – they could either be compensated per correct answer, as in round one, or compete against the other members of their group for a potentially higher payout with greater risk, as in round two. This third round, in which participants could decide whether or not to participate in the riskier tournament option, served as the measure of competitive inclination.

The author compared gender gaps in competitive inclination across the three ethnic groups while controlling for each participant’s probability of winning the tournament, overconfidence, and level of risk aversion. Since the students were in the same school and grades are public knowledge, they might already have some awareness of their competitive advantage relative to the other members in their group. For the competitive rounds, in addition to the participant’s own scores, the models also took into account the scores of the three competitors as well.