President Trump’s approach to scaling back social distancing measures aimed at containing the novel coronavirus is, to put it mildly, inconsistent.

Earlier this week, he claimed to have complete authority to rescind the measures, which he didn’t. His administration, in its recommendations that it released Thursday for considering whether to resume normal economic activity, instead recognizes that governors had that power. And his insistence that businesses reopen is offset by his formal embrace of those standards, which would keep them closed.

Trump is also aware that some of his supporters are vocally opposed to the measures that exist. So on Friday, he embraced a weird sort of middle ground, supporting opponents of distancing measures indirectly by calling on states to be “liberated” from the restrictions — less than 24 hours after offering guidelines to governors about how to do so safely.

He began with a rousing battle cry aimed at Minnesota. Given his enthusiasm about comparing the pandemic to a war, one might be forgiven for assuming he was referring to a liberation from the purported enemy: the coronavirus. But, no, he was offering his support to protesters in the state who object to social-distancing mandates.

He soon followed that up with similar exhortations for Michigan, where there were protests earlier this week, and Virginia. That each has a Democrat as governor is probably not a coincidence.

Again, it’s not as though these state are simply imposing the restrictions without any rationale. It was Thursday afternoon when Trump was on the phone with governors, offering the federal government’s insights into the factors that might prompt a state to consider allowing some level of increased social interaction, including business activity.

At a news conference, White House coronavirus task force member Deborah Birx explained the initial “gating criteria” states should use for those decisions. They are broken into three categories, looking at symptoms, confirmed cases and hospitalization rates.

Trump’s battle cries seem even more bizarre when you consider that looking only at the first two categories Birx outlined, none of the three states Trump wants to “liberate” actually meet the criteria.

Let’s consider just the first two categories, symptoms and cases.

To resume economic activity the White House — Trump’s White House — has a two-part benchmark measuring symptoms of covid-19, the disease caused by the virus.

  1. The Centers for Disease Control and Prevention’s measurement of influenza-like illness in the state should have trended downward for two weeks.
  2. Its data on emergency room visits for covid-19-like illness should similarly trend downward over the same period.

The category for cases is an either-or. Either:

  1. The number of new cases should trend downward over a two-week period, or
  2. The number of positive tests conducted should trend downward during that period.

So let’s consider Minnesota. How does it fare?

Well, the state just saw an uptick in its ILI activity in the April 5 to 11 period, data for which were released on Friday. The CDC is likely to have more current, daily data, but based on this public data, that alone is enough to indicate that the state shouldn’t begin “liberating.”

The only data available for measuring covid-19-like illness — “syndromic illnesses,” in Birx’s articulation — is offered regionally. In Minnesota’s region, the number of covid-19-like illnesses trended down from late March to early April (the most recent data available). The CDC has daily data on syndromic illnesses in smaller geographic regions which we don’t, so take that metric with a grain of salt.

(Vertical lines on these graphs mark two weeks before the most recent data. The most recent data for ILI is from April 11. For syndromic illnesses, it’s April 4. For cases and tests, it’s April 16.)

The daily case totals in a state can vary widely and depend on the number of tests completed. That said, the picture in Minnesota, using data from Johns Hopkins University, doesn’t show a consistent downward trend relative to two weeks ago. Data from the COVID Tracking Project indicates that the percent of tests coming back positive has been trending up. Without hitting one of the two case-total benchmarks, Minnesota fails the test.

While these recommendations have the feel of being predicated on hard data, there are clearly some gray areas, even beyond our access to immediate information. Consider Michigan.

There, influenza-like illness activity has been flat for two weeks, not down. Is that acceptable if the number is low? If it is, then Michigan would appear to be in decent shape.

Confirmed case totals are down in Michigan, meeting that requirement. Syndromic activity is also down — though this is again regional data. Michigan and Minnesota are in the same region, but it’s fair to say that the spread of the virus in the states is not equivalent. (Minnesota has the lowest per capita case total while Michigan has one of the highest.)

On paper, Virginia also fails the test. ILI activity is down, as is regional syndromic activity. The number of daily cases is up, though, as is the rate of positive tests in the state. Virginia doesn’t hit the benchmarks.

Curious how your state fares? We took state-level data from the above sources and created simple graphs for each state. Again: the CDC data for ILI and syndromic cases is narrower and newer than what’s publicly available. So don’t use these charts to reopen your state, governor.

Show benchmarks for .

Your browser cannot display this graph.

This is a complicated issue, one that requires nuance and thoughtful consideration about how to balance the public health of a state with its economic health. Every governor — every American — wants a return to normal. The White House task force offered its thoughts on how best to get there.

And then Trump started tweeting.