New research by Colorado College faculty and students has identified and quantified three significant impacts of structural racism on the health of American communities during the pandemic.
Economics and Business Professor Daniel Johnson worked with Claire Altieri ’21 and Siena Blelloch ’22 last summer to collect nationwide data on Coronavirus infection and mortality rates across time by state. Then they used three levels of econometric analysis to discover statistically significant and important reasons that Black, Indigenous, and other people of color continue to face worse health outcomes than statistically similar white populations.
First, they found that some states elect not to collect or report health or mortality statistics by gender or race, a result that they found highly correlated with worse BIPOC outcomes. In other words, a Heckman correction to the data (invented by alumnus and Nobel laureate James Heckman ’80 for situations in which reporting is incomplete in a nonrandom manner) was warranted in the analysis.
Next, they confirmed that socioeconomic variables like average income, income inequality, rural/urban balance, unemployment rates, age distribution, and policy choices (like mask requirements and stay-at-home orders) all played quite predictable roles in the transmission of the virus and resulting fatalities last summer. What didn’t affect infections as expected were public demonstrations, whether for Black Lives Matter or other social concerns, which showed little effect on infection data. Interestingly, states voting more Republican showed lower mortality rates (once other factors were controlled), perhaps partially explaining partisan differences in policy urgency to contain the pandemic.
Finally, they explicitly modeled policy responses to the pandemic as a function of socioeconomic variables using an endogenous treatment approach. There they found that while COVID containment policies (mask requirements and stay-at home orders) clearly worked, they were less likely to be implemented in areas where more of the population identifies as Black, and where the population is more rural or votes more Republican. Even when introduced, those specific policies were far more likely to positively influence the health of white populations than the health of BIPOC populations in the same area.
As a result, the study concludes that during the summer of 2020 alone, differential outcomes by race caused over 873,000 infections (30% of the national total), a loss of at least $1.72 billion in GDP, and a staggering 39,303 BIPOC lives lost nationwide. Stated another way, if all Americans had faced the infection and mortality rates of Black Americans, the U.S. would have added another 1.16 million infections, a loss of at least $2.94 billion in GDP, and an additional 115,047 COVID deaths within only that three months.
The article has been professionally reviewed, accepted, and is forthcoming in the journal Applied Economics Letters later in 2021.