It’s not new that the White House would shift the goal posts in its assessments of the coronavirus pandemic. Twice this week, senior officials have waved away the recent uptick in new cases by insisting that the rate of growth is still lower than when the virus first emerged. (National Economic Council Director Larry Kudlow did so Thursday; Vice President Pence did so at a briefing Friday.)

President Trump, of course, continues to insist the increase is just a function of more testing.

“We have more cases because we do the greatest testing,” Trump told Fox News’s Sean Hannity on Thursday. “If we didn’t do testing, we’d have no cases.”

Well, if we did no testing, we’d have no confirmed cases. If we did no testing, though, we would see an exponential growth in actual infections, as we did at the outset of the pandemic, when testing was scattershot. Back when Trump was insisting that the virus would soon go away.

It would be hard to match that early level of growth simply because it’s easier to see a 5,000 percent increase in 100 cases (an additional 5,000 cases) than in 100,000 cases (an additional 5 million). So it’s useful, then, to step back and evaluate where we are and how the pandemic is changing in absolute terms.

First, we are seeing a renewed surge in new cases. The past two days of data (for Wednesday and Thursday) have seen the largest increases in new daily cases on record. The seven-day average of new cases (the most commonly used metric, since it smooths out unevenness in reporting data) hit a new high Thursday.

At the same time, the seven-day average of daily deaths is at its lowest point since the last day of March.

This is unquestionably good news, so it was naturally a focal point of Pence’s during Friday’s coronavirus task force briefing. He noted that on two days this week, the number of deaths was under 300, something he and the task force attributed to expanded access to medicine, better understanding of treatment and a higher density of new cases being detected among younger people, where the virus has proved to be less dangerous.

That there are fewer deaths as the number of cases climbs is somewhat perplexing. Over time, the number of deaths in the United States has been about 5 percent of the total number of cases. But that’s skewed by a relatively high rate of deaths early in the pandemic. Since early May, the daily ratio of deaths to cases has consistently fallen, dropping below 2 percent in recent days.

That’s a bit misleading, since deaths necessarily trail infections. If we compare the daily average number of deaths to the number of new cases one week prior, we see an even steadier decline in the number of cases leading to death.

That the ratio has dropped is itself attributable to some extent to the increase in testing. Daily data compiled by the COVID Tracking Project show how, after a lull in mid-April, the number of tests being conducted each day has steadily risen. That’s why more young people are being detected; we now have the capacity to conduct tests even on asymptomatic people.

That we detected so few cases in April (relatively speaking) means that a lot of cases went undetected — and that the people most likely to be tested were the sickest. That skewed the ratio of cases to deaths, since healthier people who had contracted the virus weren’t included in the confirmed case totals.

It also meant that the rate of positive tests was higher, since sicker people were more likely to get tests. As we’ve deployed more testing, the rate of positive tests has dropped.

Or, had dropped, until earlier this month, when it started climbing again.

That’s why Trump’s “it’s because of more testing” line is obviously misleading. More tests are coming back positive even as testing expands, meaning that more people are being infected.

If we look at the change in testing and new cases by state since June 1, we see how those things correlate. States with more testing and more cases relative to the beginning of the month (at upper right) generally have also seen increases in the rate of positive tests being returned.

In the case of Florida and Arizona, they are seeing a lot more new cases each day than they did June 1, while the rate at which testing has expanded has grown more slowly. (That dashed diagonal line shows a rate of growth in testing and in new cases that’s equivalent. Circles over that line are seeing cases grow faster than testing.) States where new cases have fallen are also seeing dropping positivity rates. It is a good indicator of how things are evolving.

Florida and Arizona stand out in another way: They are two of the four states powering the new surge in cases. The others are California and Texas. Those states together are adding an average of more than 16,000 new cases a day — far more than New York added each day on average at its peak.

When you adjust for population, the increase becomes more stark. Arizona, for example, has seen a huge surge in per capita daily cases, nearly a seven-day average of 40 per 100,000 residents. At its peak, New York was averaging about 50 new cases per 100,000 people each day. By this metric, too, California fades a bit: It’s adding a lot of cases, but it’s a big state.

As the administration points out, the growth of new cases in these states isn’t even. Los Angeles, for example, is seeing more growth than other parts of California. But that argument collapses a bit when you consider that the same is true of the United States more broadly: It’s these four states, not every state, which are particularly problematic. There are hot spots, as there have been from the beginning. The issue is whether we treat these like an ongoing problem with a unified solution or as one-off crises to be dealt with as they emerge.

The administration prefers the latter approach. The data above show one result of doing so.