New study reveals limitations in quantitative research on structural racism and provides solutions

The study of structural racism—the laws, policies, practices and norms that create unequal access to resources and opportunities, particularly in health care—has matured into a robust field since its founding in the 1990s. Researchers have moved beyond commentaries and opinion pieces to data-driven research that measures its real-world impact.

Now, a comprehensive review of the literature on these empirical studies has uncovered two critical limitations in the ways most scholars use area-based data—and tools for addressing them.

The paper, by Kristi L. Allgood, PhD, a social epidemiologist with the Texas A&M University School of Public Health, and colleagues from Tufts University, the University of Michigan and the University of California-Irvine, was published in the American Journal of Epidemiology.

Using federal policy domains from the 1968 Kerner Report and its 50th anniversary update, the team identified relevant federal policies and existing indicators of structural racism in the literature.

“In short, we found that empirical studies of structural racism typically use a limited set of area-based measures, such as comparing the number of Black and white residents who are college graduates in a specific county,” Allgood said. “Our findings suggest that these may not always be the best choice, especially if researchers are considering the policy context that led to subsequent inequalities.”

Allgood said that researchers could improve area-based measures by expanding them in two ways: by explicitly communicating any links between these indicators and racist policies and by using a broader range of these indicators.

“There is often an unstated assumption that area-level racial and ethnic inequities are the result of racist policies, but showing a clear and specific link between the two removes any speculation about alternative causes, such as individual behaviors or preferences,” Allgood said. “This addition also makes the needed points of intervention easier to identify.”

Similarly, Allgood said researchers could identify the most relevant area-level indicators of structural racism by expanding the range of indicators considered.

“Most empirical research uses just a few, easily obtained area-based indicators such as the Decennial Census and the American Community Survey, but these might not capture the full extent of racial health inequities,” she said. “In addition to using other data sources, researchers could consider adding years of measurement or studying other racial and ethnic minorities.”

The team supplemented the literature review with practical tools for research. These include a comprehensive table of discriminatory federal policies, a list of common and novel indicators of structural racism across multiple domains, data sources, and an applied example showing how to connect policies and indicators ofstructural racism.

“Our list spans different domains that can be used together to better capture the multi-dimensional nature of structural racism,” Allgood said.

Allgood called the paper a “starting point” for examining the various domains of structural racism over time and the impact of co-occurring policies on racial health inequities.

“We hope our work will inform and support research that identifies ways to counteract and repair the historical discriminatory effect of older policies,” she said.