Recent social movements have generated a renewed emphasis on promoting diverse and inclusive workplaces. For example, institutional investors have increased their investments in firms that demonstrate strong commitments to diversity, and regulators also increasingly require firms to describe the extent to which their culture is diverse and inclusive. Yet, it remains unclear whether employees ultimately value diversity information and whether it factors into their job search. This issue is of particular importance given the scarcity of diversity information available to employees, with 17 percent of public firms disclosing either numerical metrics of gender or racial workforce diversity in their 2020 public disclosures filed with the Securities and Exchange Commission. Our objective is to understand how information about the diversity of a potential employer affects individuals’ job-seeking behavior, whether this varies based on worker demographics, and what factors determine whether firms voluntarily disclose such information publicly.
We conduct a field experiment with Zippia, a leading career advice agency in the U.S. that, among other things, collects and aggregates firm-level diversity information. In our field experiment, we present jobseekers with an email indicating jobs that might be of interest, and randomly vary whether the email includes information about prospective employers’ levels of workforce diversity. We then measure interest in jobs that include or do not include the corresponding diversity signals.
Our experiment was conducted over an 11-week period beginning in June 2021 and ending in August 2021. Participants are 267,494 unique jobseekers who signed up for and received nearly 5.4 million email notifications about job listings from Zippia during the 11 weeks. These job postings relate to 107,810 identifiable companies, of which 66,694 provide requisite information for Zippia to calculate a diversity score. We use a 1×3 between-participants design, where participants are randomly assigned to receive either “baseline” job listings (Baseline condition), job listings with diversity scores (Diversity condition), or job listings with salary scores (Salary condition). Diversity and Salary scores rank an employer’s workforce diversity and offered salaries against those of peers from similar locations. All other information is held constant, including median salaries in dollars. Our dependent variable is participants’ click-through behavior (i.e., whether the participant clicks on particular jobs to learn more about and apply for those positions). Participants in our sample ultimately click on job listings for 9,311 identifiable companies, with 8,568 companies having diversity scores. We find no significant differences across our three experimental conditions with respect to participant education, gender, ethnicity, or job level (e.g., “Junior”, “Senior”, “Executive”, etc.), indicating successful randomization.
Our headline result is that job listings clicked under the Diversity condition have higher diversity scores than the job listings clicked under the Baseline condition. That is, including information about diversity influences how participants direct their attention and clicking behavior when presented with a menu of job options. This result yields two key insights. First, jobseekers on average value a firm’s workforce diversity. Second, holding underlying workforce diversity constant, and in line with our focus on information content, jobseekers respond to additional diversity information. In other words, this information has value.
While our field experiment helps us use a natural setting to understand whether jobseekers find diversity information useful, it is not equipped for understanding why diversity matters to jobseekers. There are several possible mechanisms underlying our main result. First, workers may intrinsically value employment at a diverse firm (i.e., a “Preferences” channel). Second, jobseekers may choose jobs in worker-diverse firms to avoid discrimination and increase the chance of promotion and inclusion at the firm (i.e., a “Discrimination” channel). Third, diversity information might signal that the firm is of higher quality across other unobservable metrics (i.e., a “Signaling” channel). To provide a better understanding of why jobseekers may or may not value diversity information, we conducted a follow-up survey in December 2021 that yielded responses from many of our original participants. From the jobseekers who value diversity information, we find supporting evidence for all three channels. Furthermore, we find that, among the three channels, there is the most evidence for workers believing that diversity is an important social issue, that is, a Preferences channel.
Informed by the results in our field experiment, the final part of our study seeks to understand how firms publicly communicate their workforce diversity levels. We study the recent Human Capital Disclosure (HCD) mandate imposed on firms by the SEC in November 2020, around which diversity was a commonly discussed theme. The mandate entails substantial judgement and discretion because it requires firms to disclose only the features about their employees that they deem material. We explore whether firms consider the importance of diversity information for employees (highlighted by our field experiment) when making their HCD disclosure choices. Specifically, we extend our field experiment and show that firms’ own diversity performance (our experimental stimuli) and the relevance of diversity information to jobseekers (our experimental treatment effect) are both associated with a greater probability of firms disclosing precise, numeric diversity metrics in their 10-Ks. These findings suggest that firms’ diversity disclosure choices are partly explained by the usefulness of diversity information to jobseekers.
Our study breaks the link between a firms’ underlying diversity and the disclosure of that diversity level. By demonstrating the importance of workforce diversity information in the job search process, our study provides important insights for policy-makers considering whether to prescribe more disclosure about diversity and for firms that might use disclosure to attract and retain a diverse workforce. We also highlight how disclosure of a firm characteristic can elicit heterogenous stakeholder responses.
This post comes to us from professors Jung Ho Choi at Stanford University Graduate School of Business, Joseph Pacelli at Harvard Business School, Kristina M. Rennekamp at Cornell University’s SC Johnson Graduate School of Business, and Sorabh Tomar at Southern Methodist University. It is based on their recent article, “Do Jobseekers Value Diversity Information? Evidence from a Field Experiment,” available here.