On Monday, Greg Abbott announced the beginning of Texas’s “phase two” easing of public safety measures. By the end of the week, restaurants can double capacity to 50 percent, while bars, bowling alleys, and bingo halls can reopen at 25 percent capacity. The order exempts El Paso County, where leaders have concerns about, among other things, hospital capacity, as well as the hard-hit counties in the Amarillo area, including Randall, Potter, Moore, and Deaf Smith.

While the numbers of confirmed COVID-19 cases, deaths, and hospitalizations around the state have trended upward in recent weeks, Texas hasn’t yet seen massive spikes in infections or had its hospitals overwhelmed. As part of a high-stakes gamble, state leaders are now betting that, even as they relax public safety measures, that will continue to be the case.

In April, Dr. Mark McClellan, one of Abbott’s medical advisers on reopening the economy, laid out benchmarks that states should reach before easing safety measures. These included a sustained, fourteen-day drop in cases; testing capacity that allows for everyone with symptoms to be tested; the implementation of robust contact tracing and health monitoring; and the hospital capacity to treat everyone with symptoms. Those are similar to the mid-April guidelines issued by the Trump administration about when it’s safe to relax safety measures.

Texas has performed well on only the last of the four benchmarks. Thus far, hospitals in the state haven’t been overwhelmed, and have managed to treat those who fall ill without running out of beds. That’s even true in less populous regions that have seen large outbreaks, such as Amarillo’s Potter County, where the number of cases exceeds that in Bexar County, which is more than fifteen times its size.

On the other three criteria, however, Texas’s performance has been mixed. First, the number of daily cases has trended up, not down, between the start of phase one, just over two weeks ago, and today’s announcement. On Friday, May 15, the state experienced its highest single-day spike in new cases, with more than 1,800 Texans diagnosed with COVID-19.

Second, since May 1, Texas has administered slightly more than 20,000 tests a day on average, a number that increased to more than 25,000 last week. While that marks a significant increase over the testing in April, those numbers are still relatively small, accounting for less than .001 percent of the state’s population. In tests per capita, the state ranks fortieth in the nation.

Third, the state lacks the capacity to interview everyone who tests positive, and to notify the people they’ve come in contact with so they can also get tested. There’s been movement toward implementing a more robust system for tracking the disease—the state currently has around two thousand contact tracing personnel, with hopes to double that number by the end of May—but public health experts estimate that Texas will need nearly eight thousand to successfully manage the disease.

There is one broadly hopeful piece of data, in Abbott’s favorite metric: the rate at which tests are coming back positive. After a brief spike early last week, the positivity rate has remained in the range of about 5 percent, meaning that 95 percent of COVID-19 tests administered on a given day are coming back negative. But basing decisions that have life-or-death implications on just one metric carries risk—especially given that the metric is flawed. As Texas A&M epidemiologist Rebecca Fischer said last week, “The numbers don’t include people with barriers to seeking care, or who are asymptomatic.” The uninsured, people who can’t take time off from work to see a doctor, and infected people who spread the disease but haven’t yet exhibited symptoms (and may never) aren’t getting tested in large numbers.

Further, according to a Department of State Health Services representative who spoke to the Texas Observer, Texas testing data includes results from antibody tests, which are generally less reliable than viral tests. If true, that would account for the large increase in tests, but it also means that the positive test rate could be less accurate. During his press conference on Monday, Abbott denied that the total number of tests provided by the state includes antibody tests, but state representative Chris Turner told the Dallas Morning News that when he followed up with the department after the press conference, it confirmed that the totals it provides include both viral and antibody tests. (Other, more reliable data points, such as hospitalizations, take days or weeks to show up, as it takes time for those with symptoms to feel the need to go to the hospital.)

It’s clear that Abbott is not using his own medical advisers’ benchmarks to calibrate the state’s reopening. Rather, the governor has taken a more shoot-from-the-hip approach to relaxing safety measures. On May 8, for example, he allowed hairdressers in the state to reopen earlier than phase two. There was no real data or science cited in the decision, but there was political pressure; Abbott focused on expressing solidarity with the Dallas hairdresser who, in opening her shop in defiance of his order, became a cause célèbre among opponents of restrictions on business openings, and among politicians seeking to appeal to them.

Texas hasn’t yet seen the sort of devastating outcomes that New York and Italy have, and the state’s leaders are gambling that, even as they relax public safety measures, this will continue to be the case. But given how thin the basis for these decisions is, that’s less of a science-guided approach to reopening than a roll of the dice.

It’s possible that this roll of the dice will work out, of course. Any conversation among a group of epidemiologists about our current attempts to control the disease is fairly glum these days, but most of them will acknowledge that there are a lot of things we still don’t know about the disease. Maybe for a reason that isn’t reflected in our current scientific understanding, the steps Abbott is taking won’t lead to our hospitals being overwhelmed. Perhaps nursing homes—where 43 percent of the fatal cases in Texas have occurred so far—will manage to keep the disease out. Maybe the state’s increased testing, while still limited, will succeed in giving us a picture of the situation inside high-risk facilities such as jails, prisons, and meatpacking plants, and suffice to keep outbreaks from spreading among those who live and work there.

It’s also possible that enough Texans will voluntarily reduce the social activities they participate in to keep the rate at which the disease spreads from person to person low. Maybe a sufficient majority of Texans will choose to wear masks if they go out, to keep themselves from potentially spreading the disease unwittingly.

But easing restrictions doesn’t make those outcomes more likely. If Texans are going to bars or spending more time in increasingly crowded restaurants, or going to bowling alleys (a place where almost everything you come into contact with has been touched by someone else!), there is no clear reason why the disease wouldn’t continue to spread. But the state is opening up regardless, and maybe the spread will be contained for reasons we don’t yet understand. As we watch the dice tumble, that’s really all we have to hope for.

Read More:
What Public Health Experts Are Saying About the Coronavirus in Texas Right Now

The Texanist: Will the Pandemic Let Texas Be Texas Once Again?

Matthew McConaughey’s Numerous Pandemic Public Service Announcements, Ranked