Since the end of May, more than 11,000 additional Americans have died of covid-19, the disease caused by the novel coronavirus. There was a burst of attention paid to the death toll when it passed 100,000 at the end of last month, but since, thanks in part to other acute crises, awareness faded somewhat. That shift was also certainly a function of states beginning or continuing to scale back measures meant to contain the virus, shifts that signaled that the crisis was fading even if that wasn’t necessarily the case.

In recent days, there have been reports of new surges in coronavirus cases in a number of states. This can be misleading, since, as President Trump pointedly noted on Monday morning, more testing can uncover more cases even if rates are declining. But that’s not always the case.

In six states — Kentucky, Missouri, Oklahoma, South Carolina, South Dakota and Wyoming — the seven-day average of new cases has increased since May 31 while the average number of daily tests being conducted has declined, according to data collected by the COVID Tracking Project and the New York Times. In 14 other states, the rate of new cases is increasing faster than the increase in the average number of tests.

There are more concrete warning signs in some places. In 10 states, the seven-day average of the rate at which tests are coming back positive has increased more than 2 percentage points since the end of May. In 11 other states, the seven-day average for the number of new deaths is up at least 5 percent since the end of last month.

There’s a lot of variability here, particularly when comparing two individual days. The tool below allows you to select your own time range for the sake of comparison: maybe you’re more interested in the shifts since May 1, when states broadly began reopening, or since Memorial Day, on May 25.

Show changes in seven-day average in
over the past 14 days.

Your browser cannot display this graph.

(The graphs above show the range of change for each metric over the past 72 days.)

By default, you see Alabama. There, the seven-day average number of new cases per day has increased faster than the number of tests being conducted. The percent of tests coming back positive has jumped significantly — a metric that’s independent of how many tests are being conducted but that may be subject to variation depending on who’s being tested.

Compare that with New York over the same period, and you get a much different picture. In New York, testing has continued to increase and new cases have continued to decline.

Again, these are snapshots of averages, an assessment that is twice removed from what’s happening in the moment. The virus can take up to two weeks to manifest, meaning that New York’s reopening could already be leading to new infections that haven’t yet been detected. By contrast, what’s happening in, say, Missouri might already have begun to improve.

What’s presented, though, is a picture of a variety of different scenarios. New York, which had more than half of the new cases nationally in late March, had fewer than 4 percent of new cases nationally on June 14. On March 23, when 55 percent of new cases were in New York, the states of Kentucky, Missouri, Oklahoma, South Carolina, South Dakota and Wyoming — which have about 2 million more people combined than New York state — had only 2 percent of new cases nationally. On June 14, they accounted for more than 6 percent of the national total.

Again, those aren’t states where the increase is necessarily a function of more tests. What the United States is approaching is a scenario in which states normalize at an apparently politically acceptable rate of new cases. If the existing mortality pattern holds, about 5 percent of those cases will lead to a patient’s death.