Descriptive norms and gender diversity: Reactance from men

When employers were informed that others had hired more employees of one gender, they did not exhibit norm-abiding behavior (on average). 

• When no information on previous employers’ choices was provided, neither male nor female employers showed significantly stereotypical hiring tendencies. 
• Male employers hired fewer female employees when informed that others had hired more women.
  • After learning that other employers favored women, male employers only hired 35.7% female majority groups, which is significantly below the equal split. In fact, male employers were 20% more likely to hire a female majority group when no norm information was provided than when the norm favored women.
  • When the norm information was framed to highlighted women’s gains, male employers were 33% less likely to choose female majority groups compared to the control, where no gender specific norm information was provided. An example of a female frame would be to tell employers that “62% of people chose more women than men,” rather than that “38% of people chose more men than women.”  
  • In contrast, male employers’ hiring decisions were not affected by norms that favored men.
• Learning that other employers had hired more women or men did not affect female employers’ hiring decisions.

Therefore, this study found that descriptive social norms did not lead to norm-abiding behavior as they do in other contexts. The results suggest that men, the traditionally high-status group, react to others’ behaviors threatening the representation and status of their gender group. As only men showed this behavior, and only when hiring norms favored women rather than men, generic concerns about equality as a motivator of behavior can be excluded. Additionally, men’s reactance was particularly pronounced when the norm information was framed to emphasize women’s gains, further suggesting their reactance was due to perceived intergroup threat by the high-status group.

The effect of descriptive gender norm information on hiring decisions that involve male and female ‘employees’ (i.e., laboratory subjects assigned to this role) was examined using a series of laboratory experiments. The experimental design consisted of two stages.  The first stage of the experiment gathered real data to establish gender norms in hiring by observing the hiring decisions of a set of ‘employers’ and the employees they selected for a stereotypically male math task and a stereotypically female verbal task.  In four sessions, 53 employer participants were asked to hire five employees out of 20 candidates (10 male and 10 female) for each task. 

The second stage of the experiment examined the impact of the gender norms established in the first stage on hiring decisions of another set of employers. In 25 sessions, 192 participants were presented with the exact same instructions and candidates as the participants in the first stage. The only difference was that the participants in the second stage were informed of the hiring norms established by employers in the first stage. There were four treatment conditions, including female favoring and male favoring norms that were paired with a female frame and male frame. In the respective two control conditions, no information was provided on the fraction of men and women hired during the first stage.