In a week in which the majority of U.S. states began to ease social distancing restrictions, the White House’s coronavirus task force faced an uncertain future and the president toured a Honeywell factory without a mask while “Live and Let Die” played in the background, the number of confirmed cases of COVID-19 continued to rise — even above what experts had expected — as more and more tests have been conducted. The total count of positive cases now stands at more than 1.2 million, according to numbers from The COVID Tracking Project.
In our effort to make sense of the range of possible outcomes for the outbreak, we’ve continued to follow the results of a weekly survey of infectious-disease researchers from institutions around the United States. In this week’s survey, conducted May 4 and 5, experts predict that the total number of positive cases in the U.S. will be more than 1.3 million and less than 1.45 million on May 10. The experts also forecast a total of 255,500 deaths attributed to the coronavirus by the end of 2020.
The survey is organized by Thomas McAndrew and Nicholas Reich, both biostatisticians at the University of Massachusetts Amherst. Respondents are asked to provide their best estimate, as well as a best-case and worst-case possibility. Structuring the survey in this way lets the organizers generate probabilistic consensus forecasts, a tool that can answer questions about how likely various scenarios are.
Here’s what the experts had to say this week.
How many confirmed cases will be reported by May 10?
Last week, the consensus from the experts was that there would be between 1 million and 1.15 million confirmed U.S. cases by May 3. The reported number of 1,152,006 fell just outside that range, according to The COVID Tracking Project. This week, the researchers were asked to predict the number of confirmed cases by May 10. The expert consensus put the most probable range between 1.3 million and 1.45 million cases. (Like last week, the researchers were given five distinct ranges of total confirmed cases and asked to assign a probability to each bin.)
How many U.S. jurisdictions will report more new cases in September than in June?
In a new question, the experts were asked to forecast out to September and predict how many states and territories (plus the District of Columbia) would report more new COVID-19 cases in that month than they did in June. Questions like this require two separate predictions, with one largely conditional on the other. This increases the difficulty and uncertainty of the forecasts.
Accordingly, the distribution of answers given by the experts is more uniform across the range of possible outcomes. But in the mass of uncertainty, there is perhaps some good news: As summer stretches into fall, experts assigned the highest probability to fewer than four states reporting more new coronavirus cases than they had in June.
What will happen in Texas?
On May 1, Texas Gov. Greg Abbott allowed his stay-at-home order to expire. In the most recent seven-day period for which the state is reporting data — April 30 through May 6 — the average daily new case count is 1,053. The panel of experts were asked to forecast the average number of new daily cases in Texas during the week ending on June 13.
The experts predict that the new seven-day daily average in Texas is most likely to rise to between 1,350 and 1,800, affirming the widely held belief that social distancing measures have been effective at containing the spread of the disease. Experts placed a slightly smaller probability on the chances of Texas seeing between 900 and 1,350 new daily cases, on average.
Meanwhile, most experts believe that the average new daily case count would have fallen under the stricter stay-at-home guidelines that Texas previously had in place. The experts forecast a 31 percent probability that the average number of new cases would have been between 675 and 900 had the stay-at-home order not been allowed to expire.
How many deaths will there be by year’s end?
Finally, the experts were asked to forecast the total number of deaths attributable to COVID-19 by the end of 2020. The consensus estimate is that 255,500 die, with a 90 percent chance of the death total falling between 118,750 and 1,211,500 — a large range.
The low end of the expert’s range may be optimistic, however. The FiveThirtyEight model roundup shows that most models project 100,000 or more deaths by the end of May. Modeling the virus continues to be incredibly challenging, so it’s difficult to know which set of expectations to weight more heavily. But with uncertainty still high and spring in the air, perhaps some optimism in this moment is warranted — or at least forgivable.