“It’s going to be very disturbing,” Fauci said of the future of the pandemic. “I will guarantee you that, because when you have an outbreak in one part of the country, even though in other parts of the country they’re doing well, they are vulnerable. … It puts the entire country at risk.”
“I would not be surprised,” he added, “if we go up to 100,000 a day if this does not turn around, and so I am very concerned.”
That grim prognosis seems to be at odds with what’s being presented by the White House. President Trump retweeted a tweet Tuesday suggesting that the rate of positive coronavirus tests was actually decreasing — a claim that’s not true. His administration has repeatedly pointed to the relatively small geographic representation of emerging coronavirus hotspots, identifying no more than 4 percent of U.S. counties as places experiencing new outbreaks.
On Monday, the Associated Press made an important point about that claim: the 4 percent of counties identified by the administration adds up to more than a fifth of the country’s population.
It’s not entirely clear what criteria the White House is using to identify those particular counties, but we can approximate the same discrepancy using data collected by The Post’s data team.
Let’s consider that same metric: change in the seven-day average of new cases in each county. We have data for most counties and can look at the positive or negative change in the average over a two-week period. So if, for example, the seven-day average of new cases in Washington County on June 1 was 10 and the average on June 15 was 20 cases, that’s an increase of 100 percent — a doubling of the total.
The graph below shows the percentage of counties and the percentage of the total population which passed four different benchmarks on each day since Feb. 10: any increase over the prior two weeks, a 25 percent increase relative to two weeks prior, a doubling in the average number of new cases and a tripling of that average.
A few things to notice.
The first is that more counties have seen an increase in new cases relative to two weeks previously than at any point before now. That doesn’t equate to more Americans being at risk, because that increased geographic diversity includes a lot of relatively low-population counties. That big spike in early April represents the New York City area, as you might expect, and it would (happily) be hard for nearly anywhere else in the country to match that number of new infections.
Notice, too, that the current number of counties that have seen new cases double in the past two weeks is also higher than at any other point. The vast majority of those counties are still seeing less than 50 new cases per day on average, which is why the percentage of the population represented by those counties is relatively low.
But it’s not that low. More than a third of the country’s population lives in a county where the average number of new cases has doubled over the past two weeks.
It’s important to reiterate that the pandemic is something of a epidemiological fractal pattern. Globally, there are some countries worse than others. In the U.S., some states are worse than others. In each state, some counties are worse than others. In those counties, some cities are worse than others, and so on. So just because a third of the country lives in a county in which the rate of new cases has doubled doesn’t mean that each of those people is at immediate risk of contracting the virus.
What the graph above does reflect, though, is the extent of the reemergence of the virus in the past few weeks. Fauci’s worst-case scenario may not come to pass, thanks, in part, to his warnings. But it’s clear why he’s worried.