Absence of conflict of interest.
The study's objective was to examine the impact of Affirmative Action on employment outcomes for women and minorities.
The author used a nonexperimental design to study employment ratios before and after winning a federal contract. Using statistical models and firm-level longitudinal data from the U.S. Equal Employment Opportunity Commission, the authors estimated the impact of gaining a federal contract.
The study found that affirmative action regulations were significantly related to increases in employment shares for Black and Native American men and women, but with decreases for White women and Hispanic men at federal contracting firms compared to non-contractors.
This study receives a low evidence rating. This means we are not confident that the estimated effects are attributable to Affirmative Action legislation for federal contractors; other factors are likely to have contributed.
Features of the Intervention
Affirmative action became part of the labor market when it was signed into federal law in 1961 by President John F. Kennedy via executive order. Affirmative action legislation regarding employment mandates that all federal contractors actively seek to treat employees equally regardless of race, gender, religion, or nationality. Firms are required to comply with the Office of Federal Contract Compliance Programs (OFCCP) and the U.S. Equal Employment Opportunity Commission (EEO), which collect data, review required yearly affirmative action plans, enforce compliance, and deal sanctions. After 1996, several states began banning affirmative action in public employment. As of the publication of this study, these states included California, Washington, Michigan, Nebraska, Arizona, New Hampshire, and Oklahoma.
Features of the Study
The author used a nonexperimental design to study the relationship between federal contract status and the share of women and minorities in a given company. The main data source was the U.S. Equal Employment Opportunity Commission which provided firm-level longitudinal data on firms both with and without federal contracts. The analytic sample includes a total of 123,511 firms from 1973 to 2003 consisting of 53,110 federal contractors with over 50 employees and 70,401 firms without federal contracts with over 100 employees. Firms were observed for an average of 8.2 years, and 43 percent were federal contractors. The firms included in the sample are spread across the US geographically and come from many industries, principally manufacturing, services, finance, retail trade, wholesale trade, transportation, and construction.
The study divided treatment and comparison groups by federal contractors, beholden to affirmative action regulations, and firms without federal contracts. The author used a statistical model to compare the employment outcomes of the firms in the treatment and comparison groups. The author included fixed effects for year, region, and firm, and compared outcomes between 2 years before and 4 years after the treatment firm began a federal contract.
The study found that becoming a federal contractor was significantly related with a reduction in shares of White female employees (-0.12 percentage points) and Hispanic male employees (-0.058 percentage points).
The study found that becoming a federal contractor was significantly related with an increase in shares of both male and female Black employees (0.04 percentage points) and Native American male and female employees (0.014 and 0.08 percentage points).
The study found that becoming a federal contractor was also significantly related with an increase in the share of White male employees (0.09 percentage points).
The study did not find a significant relationship between becoming a federal contractor and shares of Hispanic female employees, Asian male employees, or Asian female employees.
Considerations for Interpreting the Findings
The author did not account for preexisting differences like age of employees between the treatment and comparison groups when establishing baseline equivalency or include the necessary controls. These preexisting differences between the groups—and not Affirmative Action—could explain the observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this is low. This means we are not confident that the estimated effects are attributable to affirmative action legislation for federal contractors; other factors are likely to have contributed.