Several studies - mostly conducted outside of academia - using field experiments investigate how different application or compensation settings increase the share of women applying.
Flory et al. (2015) show that women shy away from applications if the compensation package becomes more heavily reliant on individual relative performance. However, the authors also find several factors that mitigate the effect of competition-based pay on gender composition of applicants. For example, reducing the percentage of compensation determined by relative performance, hiring into jobs where the work is completed in teams, and hiring into versions of the job that have removed masculine connotations each have an important role in substantially attenuating the observed gender differences.
Flory et al. (2019) show that signaling explicit interest in employee diversity increased the interest in openings among racial minority candidates, the likelihood that they apply and are selected significantly, however, the effect on gender diversity was not significant.
Abraham & Stein (2019) show that words do matter for applications to job postings. It is tested whether women might hold themselves to higher standards than their male counterparts when deciding whether or not to apply for a job. Evidence suggests that choices about applying are correlated with skills for women but not men.
Coffman et al. (2019) find that very few talented, qualified women apply for an advanced job in a baseline condition. Once the study provides more detailed information on expectations of what is needed to receive the advanced job, the gender gap is reduced substantially, creating a more gender diverse, and more talented pool of applicants.
Samek (2019) shows that competitive compensation schemes that depend on relative performance disproportionately deter women from applying. However, the study does not find the same deterrence effect under a compensation scheme in which students face uncertain payoffs independent of performance. The charity setting increases the willingness to apply but does not affect the gender gap.
Gee (2018) shows that showing job seekers the number of applicants for a job posting on a large job-posting website increases the likelihood that a person will finish an application by 3.5%. Furthermore, women have a larger increase in their likelihood of finishing an application than men.
US natural experiment (with n almost 9,000) randomizing job seekers into different compensation regimes. By varying the role that individual competition plays in setting the wage and the gender composition, the authors examine whether a competitive compensation regime, by itself, can cause differential job entry.
To test whether men and women respond differently to employment contracts characterized by competition and uncertainty, the authors conduct a natural field experiment on job-entry decisions in 16 major U.S. cities. They posted employment advertisements to an internet job-board in cities with different market wages and randomized interested job-seekers into different compensation regimes for the same job. The set of possible compensation regimes was identical in all cities. Job-seekers were randomly assigned into treatments offering fixed-wage compensation, compensation depending mildly on individual relative performance, compensation depending heavily on individual relative performance, team relative performance, or on elements of uncertainty. Thereafter, each job seeker decided whether to apply formally for the position. In addition, two different versions of the job were advertised. One ad presented a version of the position with male connotations, while the other removed these connotations. Comparing the application patterns for these two versions of the position renders it possible to clarify the relevance of task-dependence and gender–job associations. First, the study finds as the compensation package becomes more heavily reliant on individual relative performance, the applicant pool becomes significantly more male dominated. In the limit, the gender gap in applications more than doubles when a large fraction of the wage (50%) depends on relative performance. This result is not driven by any attraction to competitive work environments among men. Women, and to some extent even men, are repelled by competitive work environments, but overall women have a stronger aversion to them. The authors also find several factors that mitigate the effect of competition-based pay on gender composition of applicants. For example, reducing the percentage of compensation determined by relative performance, hiring into jobs where the work is completed in teams, and hiring into versions of the job that have removed masculine connotations each have an important role in substantially attenuating the observed gender differences. Likewise, there a link between gender-based differences in competitiveness and the age of job-seekers: older age cohorts show much less of a gender gap than their younger counterparts. Moreover, the broader economic environment seems to play a role for entry into competitive workplaces. In particular, the gender gap in applications for competitive workplaces is correlated with prevailing market wages: the gender gap is most prominent in areas with higher local wages. Simply put, women are more likely to walk away from competitive workplaces if there are good outside options—i.e. positions with comparatively high fixed-wages—but not if these outside options are lacking.
In a second natural field experiment the authors shed further light on the underlying mechanisms at work. In this second set of experimental treatments with 2189 different job-seekers, the study finds that expectations about gender composition of the work environment, a potential driver of the findings in the first experiment, have little effect. That is, neither the gender of co-workers nor the gender of the supervisor affects the gender gap in applications for competitive workplaces. Furthermore, the results suggest that preferences over uncertainty can be just as important as preferences over competition per se in driving job-entry choices.
US natural experiment in of a major financial service corporation trying to attract minority candidates in the recruitment process (n=1,212). The corporation tested different types of signals regarding the extent and manner in which the employer values diversity among its workers. By randomly varying the content in recruiting materials (either by emphasizing how diversity leads to firm productivity, or to firm culture, or by attempts to encourage interest from a broader variety in major field of study than is typically associated with finance careers, like ethnic studies, nursing, and psychology) the authors find that signaling explicit interest in employee diversity more than doubles the interest in openings among racial minority candidates, the likelihood that they apply and are selected. Impacts on gender diversity are less sharp, and not significant.
US experiment with (n=616 ) edited job postings at Uber with about 8 percent of viewers who saw a job posting applying for the jobs (4,928 unique applicants out of 59,140 unique viewers). The paper evaluates one potential mechanism responsible for female under-representation in STEM: women might hold themselves to higher standards than their male counterparts when deciding whether or not to apply for a job. Thus, job postings that ask for “exceptional” expertise and a slew of bonus qualifications may disproportionately discourage women from applying if women believe they must meet all the listed qualifications. Men, the hypothesis goes, are less affected by the intensity of the language, and so are more willing to apply. If this is the case, then women may be differently screening themselves out of jobs for which they might be qualified solely due to differences in how each gender perceives language about the expected qualifications.
Therefore, in this study the treatment removed optional qualifications and softened language about the intensity of the qualifications. It is shown that job seekers meaningfully respond to language! The treatment significantly increased the total number of applications by 7 percent. Altering the language did not change the fraction of women who applied, but did close the gender skill gap. While female applicants in the control group are 6 percentage points significantly more likely to have graduate degrees than men applying for the same job, women and men in the treatment group are equally likely to have graduate degrees. The results confirm the importance of language in the self-screening process: words matter in different ways for women and men of different educational backgrounds, and materially affect job seekers' economic outcomes. Summing it up this paper provides further evidence that choices about applying are correlated with skill for women but not men.
Coffman, K., Collis, M., & Kulkarni, L. (2019). When to Apply? Discussion Paper
US laboratory and field experiments (with n varying between 500 and 1,500 ) testing the willingness to apply for higher return, more challenging work with a focus on gender differences:
The authors conducted a series of three experiments to address these questions, including a field experiment with real job applicants. In the first laboratory study, they provide evidence on how female and male students perceive their and their peers’ qualifications for real job openings. In the second, they create a controlled laboratory setting online that simulates a labor market, where they exogenously vary whether and how precisely the qualifications for a promotion opportunity are presented to participants. A similar exercise is then repeated as a field experiment on the online labor market platform Upwork with real job applicants.
Across our three studies, the authors draw three important conclusions. First, on average, women believe themselves to be less qualified for a given job opening than men. However, this gender gap is reduced when the qualifications for the job openings are more objective and clear. Second, men and women also differ in their beliefs about where “the bar” is for a given opening. Even holding fixed how they view themselves, women seem to believe it is more difficult for anyone to be qualified for a given opening as compared to men. Third, clearly stated qualifications help to reduce gender gaps in application rates.
In the field experiment, they find that very few talented, qualified women apply for an advanced job in a baseline condition. Once the study provides more detailed information on expectations of what is needed to receive the advanced job, the gender gap is reduced substantially, creating a more gender diverse, and more talented pool of applicants.
U.S. field experiment with more than 35,000 graduate and undergraduate university students exploring the impact of compensation scheme on willingness to apply for a job. In the experiment, students received one of several different e-mails advertising an opening for a short-term job. To identify the roles of competitiveness and risk preferences on the decision to apply, the author varied whether the compensation scheme for the project was (1) flat rate, (2) competitive, performance-based bonus or (3) bonus not based on performance. To learn if social preferences narrow the competitiveness gap in the labor market, the author also varied whether or not the job was described as helping a charity. The charity manipulation was conducted for both the flat and performance-based compensation schemes. If women are more likely than men to respond positively to the charity frame by applying for the job, this social preference manipulation could reduce the gender gap in worker entry in competitive environments. Moreover, if pro-sociality is negatively correlated with competitiveness, then the charity framing should have the largest effect on those who are the least competitive, exactly what may be needed to reduce the gender gap in earnings.
The study finds that competitive compensation schemes that depend on relative performance disproportionately deter women from applying. However, the study does not find the same deterrence effect under a compensation scheme in which students face uncertain payoffs independent of performance. The charity setting increases the willingness to apply but does not affect the gender gap.
Field experiment (n= 2.3-million-person ) that varies whether or not a job seeker sees the number of applicants for a job posting on a large job-posting website, LinkedIn. This intervention increases the likelihood that a person will finish an application by 3.5%. Women have a larger increase in their likelihood of finishing an application than men. Overall, adding this information to a job posting may offer a light-touch way to both increase application rates and alter the diversity of the applicant pool. In particular, women (especially those who apply for male-dominated jobs) and inexperienced job seekers are more affected by the treatment. This implies that these groups do not shy away when the information is presented because they fear competition, instead knowing the number of other applicants reduces information uncertainty and makes job applicants more comfortable with the idea of applying.