There is a stereotypical view of the places in America that COVID-19 has affected most: they are broadly urban, comprised predominantly of racial minorities, and strongly vote Democratic. This underlines the public’s perception of what kinds of populations reside in areas highly exposed to the coronavirus, as well as some of the recent political arguments over social distancing measures and the states easing their restrictions.
While that perception of high-prevalence areas was accurate during the earlier stages of the pandemic, COVID-19’s recent spread has changed the picture. During the first three weeks of April, new counties showing a high prevalence of COVID-19 cases are more suburban, whiter, and voted more strongly for Donald Trump than counties the virus hit first. These findings result from a new analysis of counties with high COVID-19 prevalence rates (more than 100 confirmed cases per 100,000 population) based on data available from The New York Times and the U.S. Census Bureau.
As of March 29, 59 counties (representing 7.8% of the U.S. population) registered high COVID-19 prevalence rates. Over the subsequent three weeks, through April 19, an additional, 142, 254, and 262 counties registered a high prevalence—increasing the shares of the U.S. population living in such counties by 11.8%, 11.2%, and 15.7%, respectively. Thus, as of April 19, 48% of U.S. residents lived in counties with a high prevalence of COVID-19 cases.
Source: William H. Frey analysis of New York Times data for confirmed COVID-19 cases and 2019 census population estimates
Note: High-prevalence counties showed rates of COVID-19 cases per 100,000 equal to or exceeding 100 for date shown
Map 1 displays the progression of these counties from March 29 through April 19. Initially, these counties were heavily concentrated in the Northeast region, especially around the greater New York metropolitan area, with some representation in other parts of the country, including around Seattle, New Orleans, and Detroit. During the week of March 30 to April 5, more high-prevalence counties sprung up, especially in the Midwest (Chicago, Indianapolis, and Milwaukee) as well as in the South (Georgia, Mississippi, and further spread in Louisiana).
The following week, April 6 to April 12, expansion of the virus in the South continued dramatically, especially around Miami, Atlanta, Baltimore, Washington, D.C., and Memphis and Nashville, Tenn., as well as many smaller metropolitan and nonmetropolitan counties. The disease also extended into Northeast and Midwest counties located in suburban areas and smaller metropolitan areas. COVID-19 also spread in the western region, especially in the greater Denver metro area.
The week of April 13 to April 19 introduced even larger numbers of counties with a high prevalence of COVID-19 into the South, across states that already witnessed an expansion as well as in Texas and North Carolina. And in California, the populous counties of San Francisco, Los Angeles, and Riverside reached the tipping point for high prevalence, along with increases in smaller counties elsewhere in the West.
Overall, there was a noticeable increase of counties with a high COVID-19 prevalence in the Midwest, South, and West over the three-week period between March 29 and April 19. These newly designated high-prevalence counties showed different attributes in terms of urban status, political orientation, and other demographic measures than earlier ones.
The shift was most marked with respect to urban residence status. On March 29, four-fifths of high-prevalence county populations resided in the urban cores of large metropolitan areas. This fell sharply in the subsequent weeks, as more shares of residents in new high-prevalence counties lived in suburbs and smaller areas. Among residents of counties which entered high-prevalence status between April 6 and April 12, for example, only 31% lived in urban core counties, while 45% lived in suburbs and 24% lived in small and nonmetropolitan areas.
Counties entering high-prevalence status during the April 13 to April 19 period showed a rise in shares of urban core residents, due largely to the inclusion of the three populous California counties. But overall, there has generally been an increased spread to areas outside of urban cores.
COVID-19’s spread has also impacted the overall politics of high-prevalence counties, as measured by 2016 presidential election voting patterns. As of March 29, high-prevalence areas leaned strongly Democratic—among 2016 voters living in those counties, Hillary Clinton held a 62-34 advantage over Donald Trump. But during each subsequent week since then, the Clinton advantage decreased—down to 52-42 for counties that reached high-prevalence status between April 13 to April 19.
A similar comparison of counties’ demographic attributes based on when they reached high COVID-19 prevalence can be seen in Table 2. The stereotype that high-prevalence counties are largely comprised of racial and ethnic minorities does in fact apply to the situation as of March 29, when whites comprised just 48% of those counties’ populations. Over the subsequent two weeks, however, whites came to comprise a 58% share of residents in high-prevalence counties. This fell to 52% for high-prevalence counties new to the April 13 to April 19 period, primarily due—again—to the addition of a few heavily minority California counties.
It is also noteworthy that the largest Black population share (20%) of high-prevalence county residents occurred for counties added between April 6 and April 12, as many of these counties were located in parts of the South with large Black populations. Similarly, the highest share of Latino or Hispanic residents (24.4%) lived in counties that registered high-prevalence status during the April 13 to April 19 period—reflecting, in large measure, counties in California and Texas with large Latino or Hispanic populations.
These counties also contributed to a rise in the foreign-born share (17.7%) of residents in high-prevalence counties, compared to residents in counties that registered a high prevalence during the previous two weeks (16% and 12.2% foreign-born shares, respectively). Nonetheless, the foreign-born share of residents in high-prevalence counties as of March 29 (25%) was far greater than the share in counties that reached that status later.
One further demographic dimension useful for comparison is the household income distributions of these groups. As shown in Table 2, the distribution for households in high-prevalence counties on March 29 shows a lower share of middle-income residents making between $20,000 and $99,000 per year (48%) than in counties that later reached high prevalence (ranging between 53% to 56%). Thus, as COVID-19 spreads, it appears to be occurring in areas with more middle-income populations.
Overall, this review shows that as the coronavirus pandemic continues its spread outward to new areas, those areas are less likely to fit demographic stereotypes that the public and some political activists attribute to COVID-19-affected places. It is important to continue to track these demographic trends as the pandemic evolves in order to gain a more accurate perspective of its impact on the nation.