The More You Know: Information Effects in Job Application Rates by Gender in a Large Field Experiment

Authors:

Showing the number of current applicants for a job posting increases the likelihood that job seekers—especially women—will apply.

FindingsMethodology

Women are underrepresented in many high-status industries including tech and finance. These gender occupation gaps contribute substantially to the gender wage gap. In recent years, employers have begun to recognize the importance and performance benefits of a more gender-balanced workforce.

As companies aim to increase their workforce diversity, they may be further encouraged by a greater proportion of women in the candidate pool. However, the factors that affect candidates’ interest in jobs are not well understood, with previous research covering limited professions and using fictitious candidate profiles. Additional information on job postings can affect women’s and men’s desire to apply to certain positions. For example, if you see the current number of applicants for a job opening, this information could be interpreted positively or negatively; either as a signal of a job’s desirability or legitimacy, a signal of competition, or simply more information reducing ambiguity around the job’s position. How such information is interpreted may vary by gender, given that women are often more averse than men to risk and competition. 

This study tested a “light-touch” intervention among a large volume of real job seekers on LinkedIn. LinkedIn, a business networking website, has over 350 million users globally and hosts job postings with heavy representation from high-tech and finance industries. The experiment randomly assigned a quarter of job-seeking users to see the current number of applicants for each job opening, while the remaining users saw no applicant numbers on otherwise identical postings. Over 16 days, the study tracked over 2.3 million job seekers as they viewed and applied to over 100,000 job postings worldwide.

Findings

Displaying the current number of applicants to a job posting significantly increased the likelihood that job seekers—particularly women—would apply.

  • When job seekers saw applicant numbers for the first job posting viewed, they were 3.6% more likely to start a job application on another website, 2.0% more likely to start an application on the LinkedIn website, and 3.5% more likely to finish an application on the LinkedIn website, compared to job seekers who saw no applicant numbers.
    • These small effects translate to an increase of roughly 1,500 applications started and 250 applications finished per day on LinkedIn.
  • Women who saw applicant numbers were 0.212 percentage points more likely to finish a job application on the LinkedIn website compared to women who saw no applicant numbers, while there was no significant change for men.
  • For masculine-dominated jobs—those for which ≥ 80% of job seekers in the control condition were men—showing applicant numbers increased the number of women starting an application (by 11.7 percentage points for an external application, 8.8 percentage points for an application on LinkedIn) or finishing an application on LinkedIn (by 4.7 percentage points).
  • For both women and men, showing applicant numbers increased the number of new applicants (job seekers who might otherwise not have started any applications on LinkedIn). Those who saw applicant numbers started an average of 0.548 applications versus those who did not who started 0.539 applications, a statistically significant difference, which goes away when looking only among at those with at least one application.  
  • Showing applicant numbers produced stronger effects for women, inexperienced job seekers, and those viewing jobs at lesser-known companies, suggesting that job seekers were encouraged to apply simply by having more information and less ambiguity.

These findings demonstrate that providing information on applicant numbers is a light-touch, low-cost intervention that can increase the gender diversity of job applicants. A more diverse applicant pool creates opportunities to reduce gender disparities in industries such as technology and finance, and potentially reduce the wage gap.

Methodology

The study took place on LinkedIn over 16 days in March 2012. Within the website’s regular business practice, one-quarter of job-seeking users were randomized to the treatment group and saw the number of current applicants on each job posting. The remaining three-quarters in the control group saw no applicant numbers. Job postings were otherwise identical, and 95% were viewed by job seekers in both the treatment and the control groups. Job applications were either internal, using the LinkedIn platform, or external, referring the job seeker to another website. The study tracked whether job seekers started an application after viewing a job posting, and whether they finished an internal application. The main analysis was restricted to the first job posting a person views and not all postings viewed because in subsequent viewings people could have compared information on the number of applicants across postings. Over 16 days, the study collected data from 2,326,207 job seekers (estimated 36.5% women) from 235 countries and areas, viewing over 100,000 job postings from about 23,000 companies.


MLA: Gee, Laura K. “The More You Know: Information Effects in Job Application Rates by Gender in a Large Field Experiment.” Management Science, revised and resubmitted.
APA: Gee, L.K. (revised and resubmitted). The more you know: Information effects in job application rates by gender in a large field experiment. Management Science.
Chicago: Gee, Laura K. “The More You Know: Information Effects in Job Application Rates by Gender in a Large Field Experiment.” Management Science (revised and resubmitted).