Reducing workplace discrimination: Data help identify what doesn’t work

Federal law prohibits discrimination against employees and job applicants because of their race, color, religion, sex (including gender identity, sexual orientation, and pregnancy), national origin, age, disability, or genetic information. The law also makes retaliation related to a discrimination complaint illegal. State and local governments also have and enforce laws prohibiting discrimination.

Training and grievance procedures are among the most common workplace practices to combat discrimination and harassment. Case law has established that these practices can protect firms against punitive damages in associated lawsuits. Organizations thus have legal incentives to provide training and use grievance procedures. Moreover, some states have enacted policies that require employers to provide anti-harassment training, further solidifying training as a cornerstone of antidiscrimination and anti-harassment strategy.

Community Development and Engagement at the Federal Reserve Bank of Minneapolis is focused on supporting low- and moderate-income workers and communities with better data and evidence. Recognizing that policymakers, employers, and workers alike benefit from knowing the effectiveness of different workplace practices in reducing discrimination, we examined whether added use of antidiscrimination training and grievance procedures results in good workplace-equity outcomes for employees. We find no clear evidence of effects.

No clear evidence that additional training reduces discrimination

The Notification and Federal Employee Antidiscrimination and Retaliation Act of 2002 (No FEAR Act) requires all federal employees to take periodic training on what constitutes illegal discrimination or retaliation. While this training isn’t the most exhaustive available on antidiscrimination and anti-harassment, it covers topics the Equal Employment Opportunity Commission recommends to include in antidiscrimination training and is similar to the training offered at many organizations. But is requiring employees to take additional instances of this training an effective way to reduce workplace discrimination?

To understand the effect of additional antidiscrimination training, we combined data on the timing and office locations of No FEAR Act training with data on complaints of discrimination and harassment raised by the employees of a large, quasi-public-sector workplace employing many low- and moderate-income workers and subject to the requirements of the No FEAR Act. (Learn more about this analysis.) Figure 1 shows that estimated effects are negative, ranging from around -5.0 percent to -3.5 percent. However, because the error bars around those estimates include zero, or no impact, we cannot say definitively that the training reduced either reports or incidents. In other words, while more training is associated with reduced reports and incidents of discrimination in our data, the relationships are not statistically significant.

We also looked at whether training for supervisors influenced the number of reports of subordinates and whether retaliation concerns increased in response to the training. We found that training has no consistent discernible effect on subordinate reports or retaliation concerns. Although our estimates from these analyses are somewhat imprecise, they align with research showing a limited impact of training on learning and attitudinal measures.¹

Examining the impact of increased grievance-procedure use

In addition to training, organizations frequently use grievance procedures, which provide workers with an official internal dispute-resolution process to help identify and prevent discriminatory behavior. Does encouraging more usage of these procedures lead to beneficial outcomes? Increasing the use of grievance procedures could, in theory, produce several outcomes, ranging from intensifying harassment and discrimination due to backlash from supervisors who have complaints brought against them, to improving equity outcomes by increasing supervisors’ accountability, to having no significant impact because of limited reporting in the first place.

In order to learn more about the efficacy of grievance-procedure use, we studied a change in grievance-procedure policy and subsequent patterns of hiring and worker exits. In particular, we examined a shift from a traditional procedure using the phone and mail to online reporting, which made it easier for a worker to report. In response to this policy change, we observe no change in women’s turnover, hiring, or promotion outcomes relative to men’s.

Building a future research agenda

Training and grievance procedures are two frequently used workplace practices encouraged by federal case law and certain state policies to combat workplace discrimination and harassment, yet our research suggests that more of those practices may not always improve discrimination outcomes. Unlike other research in this area that examines attitudinal changes, our approach looks at behavioral outcomes using a large sample and looking at the impacts of actual workplace training.²

More research is needed to understand which workplace practices and features are more effective in reducing discrimination and harassment. Existing research suggests that stress-reduction policies related to pay and working conditions could be one way for workplaces to reduce this type of misconduct.³

Different organizations may benefit from other approaches; the findings here simply reflect the experience at the workplace studied. Building upon existing data collection, aggregation, and analysis efforts will be important in growing our understanding of how to reduce workplace discrimination, and in turn improving conditions for low- and moderate-income workers. Testing novel solutions, which could include new types of training or alternative grievance-procedure models and more inventive practices, will help organizations expand their antidiscrimination toolkits and inform further refinement of federal and state policies.

Endnotes

1 This research includes a 2018 study of sexual harassment training (Roehling and Huang 2018), a literature review on prejudice-reduction interventions (Levy Paluck and Green 2009), and a quantitative and qualitative assessment of prejudice-reduction experiments (Levy Paluck et al. 2021).

2 Studies looking at behavioral outcomes often feature relatively small sample sizes and novel types of training that are not typically provided in the workplace (e.g., Chang et al. 2019; Devine et al. 2017; Dobbin and Kalev 2019; Sharma 2021).

3 See Narayan 2022 and Narayan 2023.