When Ron DeSantis ran for governor of Florida in 2018, he made a politically savvy decision: He would be the most pro-President Trump candidate possible. He popped up on Fox News Channel repeatedly, understanding that it would be an effective way to get Trump’s attention. After he earned Trump’s endorsement in the Republican primary — and then won his party’s nomination — his campaign ran an ad touting how loyal he was to the president’s vision. At one point in the spot, he helped his young daughter build a wall with her blocks.

After narrowly winning the general election, he remained loyal to Trump. After the coronavirus emerged in the United States, he echoed Trump’s insistences that economic activity should not be constrained to slow the virus’s spread.

At one point in late May, DeSantis stood in the driveway of the White House after meeting with Trump, attacking members of the media for criticizing his decision not to limit business activity.

“You’ve got a lot of people in your profession who waxed poetically for weeks and weeks about how Florida was going to be just like New York,” DeSantis said. “‘Wait two weeks, Florida is going to be next. Just like Italy, wait two weeks.’ … We’re eight weeks away from that and it hasn’t happened.”

“We’ve succeeded and I think that people just don’t want to recognize it because it challenges their narrative, it challenges their assumption,” he later added, “so they’ve got to try and find a boogeyman.”

At the National Review, Rich Lowry echoed DeSantis’s rhetoric, asking where the Florida governor should go to get his apology.

Lowry should have waited a few weeks. On the day DeSantis spoke at the White House, Florida was seeing 724 new coronavirus cases a day on average. Three weeks later, the average was more than 1,200. Three weeks after that, more than 7,100, 10 times the figure as when DeSantis was taking his victory lap. At the worst point over the summer, Florida had nearly 12,000 new infections a day and 185 new covid-19 deaths.

Trump, frustrated that North Carolina’s governor wouldn’t allow him to hold a jam-packed convention of Republican devotees, announced that he would move key events to Jacksonville, instead, where there were no restrictions. Until cases in Florida spiked and the state was finally forced to clamp down. Trump’s convention became mostly virtual.

This intermingling of DeSantis, Trump, politics and the pandemic played out over the course of the year. DeSantis was a fervent supporter of the president’s reelection bid, including attending a Trump campaign rally in October where he high-fived supporters while not wearing a mask.

That overlapped with questions about how the state was tallying its coronavirus data. Shortly before DeSantis’s White House appearance, Florida fired its top data scientist, whose creation of a data-heavy dashboard providing information about the virus had been hailed by the White House. Part of DeSantis’s ire at the media stemmed from questions about the reliability of the state’s data, given that the fired official, Rebekah Jones, alleged that her termination stemmed in part from her refusal to doctor the numbers to bolster DeSantis’s calls to resume normal economic activity.

Despite that concern, there wasn’t any indication that the numbers were being fudged. Until this week: The Sun-Sentinel reported Tuesday that a review of data released by the state appeared to show that for about a month the state had paused its reporting of backlogged coronavirus-related deaths.

There’s not a uniform way in which states catalogue and report coronavirus data. (That’s one reason that Jones’s work on a dashboard was praised: It set an early standard for clarity.) Accurate data is hard to come by in part because recording deaths from covid-19, the disease caused by the virus, can happen only belatedly, after the death has occurred. Since the pandemic began, states have reported covid-19 deaths as soon as they could, although that has at times meant reporting as new deaths that occurred weeks prior.

What the Sun-Sentinel’s Cindy Krischer Goodman and David Fleshler found was that, for several weeks in late October and running into November, deaths that had occurred previously weren’t included in Florida’s daily death totals. The effect was to show fewer deaths each day than had been shown previously.

We can see that effect in data collected by the Centers for Disease Control and Prevention. The CDC catalogues covid-19 deaths once death certificates are finalized, but also shares data on reported deaths from states. In other words, we can see both the number of actual deaths each day in Florida and the number of deaths reported each day, which might include deaths from days or weeks prior.

The actual deaths are reported weekly. In the week ending Oct. 17, the state reported more than twice the number of deaths as were actually recorded, suggesting the inclusion of older fatalities. In the week ending Oct. 24, 1.7 times as many deaths were reported. By Oct. 31, the ratio was 1.1 to 1, suggesting that the number of deaths reported matched the number of deaths that occurred. In the next two weeks, that figure remained relatively steady.

During the week ending Nov. 7, of course, the presidential election occurred. The Sun-Sentinel pressed officials for an explanation; none chose to speak on the record.

“Whatever the intent, the change led to more favorable death trends as the election approached,” Krischer Goodman and Fleshler reported.

One of the metrics we’ve used to track deaths also shows a shift in Florida around the election. Comparing the number of deaths each day to the number of new cases shows that the ratio of deaths to cases fell much faster in Florida in the weeks before the election than it did nationally — the number of new deaths was lower relative to new cases than it had been. When we compare the number of new deaths to the number of new cases three weeks prior, which has been a better predictor of how many deaths we can expect, the plunge is even more significant.

The Sun-Sentinel reports that the death data are reverting to including backdated fatalities, which appears to be indicated in the graph showing data from the CDC. Confirmed fatality CDC data, though, tend to underreport more recent deaths, because they depend on death certificates, which take time to produce and transmit.

But DeSantis’s battle with coronavirus data continues.

A White House report recommending that Florida curtail the availability of indoor dining and other activity to slow the spread of the virus was reportedly muffled by DeSantis’s administration. On Tuesday, he appeared at an event to encourage restaurants to remain open, claiming — falsely — that restaurants aren’t a significant driver of new cases. (He also declined to refer to President-elect Joe Biden as president-elect.)

Last week, Jones’s home was raided by state police who seized computers she was using to create a standalone data tracker. She is accused of illegally accessing state computer and messaging systems; she claims the state sought to silence her.

In the end, Trump would almost certainly have won Florida no matter what the coronavirus death data showed. His victory was powered heavily by a shift among Hispanic voters in Miami-Dade County, where a 290,000-vote loss in 2016 became an 85,000-vote loss in 2020, as his statewide win margin grew to 260,000 votes. But with Trump leaving office and DeSantis sticking around for at least another two years, the question remains: Did his administration intentionally misrepresent coronavirus data for political purposes?

In other words: Did DeSantis’s loyalty to Trump and favored position as a keep-the-economy-open poster child manifest in less-dire death totals?