We’ve been hearing similar rhetoric from the White House for a while. The effort to contain the coronavirus, manifested most obviously in state-level stay-at-home orders, has had the effect of reducing the number of new confirmed cases and deaths from covid-19, the disease caused by the virus. If things are headed back down, it would seem fair to begin opening back up.
The problem with that, though, is that the back of the mountain doesn’t look the way the front did. We saw a steady, exponential rise in confirmed cases and deaths each day for several weeks. But particularly with daily case totals, the period after the peak nationally has looked more like a plateau than a downward slide.
Data from Johns Hopkins University and the COVID Tracking Project shows how the daily number of cases and deaths have evolved over time. On the graphs below, the actual daily figures are shown as columns, but since there’s day-to-day variability, the focus is on the three-day moving average for each category.
There’s an obvious peak in the daily death toll — but the number of deaths at this point is about what it was April 10, when that first, smaller peak occurred. The number of new cases is trending downward, but it, too, is about at the point we saw 20 days ago.
The number of tests being completed each day, a necessary component of the number of new confirmed cases, was steady for much of April before jumping to a new peak in the past week. In the past few days, that peak has slipped. (For tests, of course, we want to see an upward trend, unlike cases and deaths.)
Those national figures are driven by data from each state. In the states, the picture varies quite a bit. Some are clearly past peaks. Others are seeing similar plateaus. We created an interactive showing the three-day average in each state relative to the peak. Using the provided slider, you can quickly visualize the extent to which states have moved past the peaks they’d seen in cases or deaths — or whether they are at a peak in testing. (Click a state to see a larger version.)
Show data for .
Highlight states below 90% of peak.
(There are three ways to view the data: the 3-day average relative to the peak, the actual daily average for each metric scaled to the national total and the relative daily change in each metric, comparing the change in a day to the peak change for that state.)
The White House acknowledges the variability you see above. The plans offered by the White House coronavirus task force use localized measures of the spread of the virus to recommend whether states should resume normal economic activity. Even within states, as in New York, there may be different guidelines for different areas.
The challenge isn’t that the variability exists, it’s that the pandemic isn’t fading as smoothly as it emerged. While the worst effects of the outbreak may be behind us, that doesn’t offer as much assurance as one might expect. A wildfire that’s 10 percent contained poses a very different risk than one that’s 90 percent contained. What’s more, in many states we’ve also seen declines from peaks followed by new increases.
We don’t want to be gloomy. It’s important to realize, though, that the opposite of exponential increase isn’t necessarily exponential decrease. Sometimes instead of things getting better, you have to settle with things not getting worse.