Knowing Your Opponent's Gender Changes Your Behavior in Contests?

We experimentally examine gender differences in contests. We used a Tullock contest because it described the practical competitive environments where both luck and efforts determined the winner of a contest. Each subject was randomly paired with another subject in each round and was explicitly informed about the gender of his/her opponent. Over-expenditures relative to the Nash equilibrium were observed in all treatments which were consistent with the literature. Males invested more than females in the single gender contest, but females invested more than males in the mixed-gender contest, which differed from the literature. We demonstrate that either risk attitude, or non-monetary value of the winning (willingness to win) did not play a major role, even though almost half subjects showed their values of willingness to win strictly positive. More than that, the difference in the self-worth of themselves as university students between males and females subjects played a significant role in the mixed-gender contest. Females invested 34% more when their opponents were males than when they were matched with other females. The result is applicable to the educational institutions and the labor market.