DURING MOST of the coronavirus period, President Donald Trump has acted as if only part of America has been affected: the Democratic part. Early on, he called the virus the Democrats’ “new hoax”. He claimed that only Democratic states were asking for bail-outs from the costs of it and has sided with the many Republicans demanding the prompt reopening of locked-down states.
And for most of the course of the disease, Democrats have indeed been worse affected. It is a rule of American politics that Democratic control increases with population density. So do coronavirus infections. New York, New Orleans, Chicago and Detroit all had early, large outbreaks; all are Democratic strongholds.
Reuters, a news agency, calculated that counties which voted for Hillary Clinton in the presidential election in 2016 have had about 39 deaths per 100,000 people, three times the number in Trump-voting ones. According to the New York Times, at the end of May counties that Mr Trump won in 2016 had reported only 27% of cases and 21% of deaths, but contained 45% of the population. The Pew Research Centre calculates that, of the 10% of congressional districts worst affected by the virus, Democrats hold 41 and Republicans three. Crowded apartment blocks; large numbers of vulnerable ethnic minorities; globally connected cities—all these explain why the disease took off in areas under Democratic control.
But the pattern of infection is changing, and so is the politics of the disease. In early June, case numbers began to rise again after more than a month of slow decline, prompting fears of a new surge of infections in states that had opened up too early. Twenty-two states reported increased caseloads during the two weeks to June 11th. Twenty of them were in the South or West and 15 had either Republican governors or Republican House majorities or both. Texas, Florida and California, the three most populous states, all reported their highest-ever daily totals of new infections in June. In contrast, at the end of March, 90% of all cases had been either in the north-east or the Midwest.
William Frey, a demographer at the Brookings Institution, a think-tank, has delved deeper into the numbers by looking at counties where the disease appears in significant numbers (meaning, those that report 100 or more cases per 100,000 people) for the first time. Between April 20th and June 7th, he says, more than half of newly infected counties supported Mr Trump in 2016. At the end of March, only a third of such counties had been Trump backers. Of the 1,292 counties that joined the category in the seven weeks to June 7th, Mr Trump won 87%. Red America, in short, is increasingly infected.
As the disease spreads, it touches different sorts of people in different sorts of places. In late March, according to Mr Frey, more than four-fifths of the population of newly-infected counties lived in inner cities, especially in the north-east. By the end of May, more than half the residents of newly infected counties lived in outer suburbs, smaller towns or rural areas. Almost half of those counties were in the South.
The ethnic profile of victims is changing, too. At the end of March, minorities were disproportionately affected and non-Hispanic whites accounted for less than half the population of new covid-19 counties, well below their share of the national population. By the end of May, whites accounted for almost two-thirds. Suburban whites are older than average and may be more vulnerable for that reason: 37% of whites living outside big cities are over 55, making them, on average, four years older than blacks and Hispanics in those areas.
It is true that some of these patterns may exaggerate the underlying trends. Though newly infected counties may be white as a whole, minorities within them may still bear a disproportionate share of the disease burden, as they seem to, for example, in midwestern states with meat-packing plants staffed by immigrants. In Iowa, says the New York Times, Hispanics account for about a third of covid-19 cases even though they are only 6% of the state’s population. So although Republican areas are reporting more cases, Democratic voters may account for a large share of the victims. And despite the recent shifts, Democratic areas are still worse affected overall than Republican ones. Using Mr Frey’s classification, 50% of the population of the high-covid-19 counties by June 7th had voted for Mrs Clinton, compared with only 44% for Mr Trump.
Still, the shift towards a whiter, more suburban and more Republican patient population could have profound implications, especially if caseloads continue to rise. For the past few weeks, the diverging experience of the disease seems to have entrenched partisan differences of opinion and behaviour, with Republicans more impatient with the continuation of public-health measures. That difference could become less sharp as the disease bites deeper in Red states. As more Trump voters fall victim, it will be harder to argue that covid-19 only affects part of America.
Editor’s note: Some of our covid-19 coverage is free for readers of The Economist Today, our daily newsletter. For more stories and our pandemic tracker, see our coronavirus hub