Fine particulate matter produced by human activity is responsible for between 85,000 and 200,000 deaths in the U.S. each year. “This is not a new problem,” said Christopher Tessum, a professor of civil and environmental engineering at University of Illinois Urbana-Champaign and the paper’s lead author.

When they set out, he and his fellow researchers had hoped to identify the specific types of pollution that most affected people of color, and thus be able to target those specific sources. Instead, they found that nearly all types affect people of color more, indicating that mitigation efforts may need to be undertaken at the neighborhood level instead. This may still mean focusing on one major emissions source, such as a nearby power plant, but will likely also need to incorporate newer policy strategies such as low-emissions zones and congestion pricing for vehicle traffic.

“In the past, air pollution policymakers have often worked by identifying a problem—such as acid rain—then identifying the main source of the problem—such as coal power plants, in the case of acid rain—and then controlling the emissions from that source,” Tessum said. “Our results show that in the case of environmental injustice, this type of approach may not work, because almost all of the source types are significant contributors to the problem.”

Researchers funneled pollution data from the U.S. Environmental Protection Agency’s 2014 National Emissions Inventory into a physics-based air quality model, which includes wind speed and direction to see where pollution ends up. They then overlaid this on a map showing demographic data by census tract. Theirs is the first study of its kind to track all 5,000 individual types of pollutants listed in the emissions inventory, although the researchers also grouped them into 14 broad types for ease of discussion.

Of the six emissions sectors responsible for the largest disparities compared to the national average, four were the same across Blacks, Hispanics, and Asians: pollution from industry, construction, light-duty gasoline vehicles, and heavy-duty diesel vehicles. Coal electric generation and agriculture were the only two sectors that affected Whites substantially more than average.

While the researchers stopped short of determining the reasons for these disparities, the paper cites the country’s history of discriminatory housing policies and practices as one possible cause. Tessum also hypothesized that gas combustion for home heating and commercial cooking—one of largest sources of relative disparities for all four non-White demographic groups—was a problem related to population density, and that communities of color tend to form disproportionately in urban areas. That said, “Even if you look only in urban areas, you see the same patterns,” he said.

Consistent with previous research, the paper finds that “racial disparities are not simply a proxy for economic-based disparities.” In other words, even though highly polluting facilities tend to be placed near poor areas, that doesn’t fully account for the disparate racial impacts. Even as incomes rise, families don’t necessarily move, leading to entrenched demographic health impacts and outcomes.

While the paper doesn’t prescribe specific policy solutions, Tessum said, “perhaps the most important thing is to listen to the needs and expertise of the members of the communities that are most affected.”