The Washington PostDemocracy Dies in Darkness

How the pandemic spread and contracted on its way to killing 100,000 Americans

Analysis by
National columnist
May 28, 2020 at 10:13 a.m. EDT
The Post examines what President Trump predicted compared with what is now known about the death toll. (Video: The Washington Post, Photo: Jabin Botsford/The Washington Post)

The first coronavirus death in the United States wasn’t the first coronavirus death in the United States.

We learned on Feb. 29 that someone had died of covid-19, the disease caused by the virus, hours after President Trump boasted at a political rally of his administration’s success in containing the virus. That death, in Washington state, was actually preceded by several other deaths, uncovered only after the fact. For all of the attention paid to the early cases of the virus in the United States and to the first few victims of its effects, the actual first death came in California on Feb. 6.

At least, the first death that we know of.

Since then, more than 100,000 Americans have died of the virus, a figure that almost certainly represents an undercount of the actual toll. Over the past four months, bodies piled up in New York City — literally — as people succumbed in rural Nebraska or Iowa. The disease didn’t spare any state, but it hit some harder than most.

The United States has been particularly hard hit, relative to the rest of the world. Around two out of every seven deaths from the virus globally has occurred in the United States — at least, if you take the death tolls in countries like China and Iran at face value, which we recommend not doing.

We can tell one part of the story of the pandemic by breaking down those 100,000 deaths into 5,000-death increments, showing how the death toll accelerated, where it spread and where it lingered.

It took more than a month for the first 5,000 deaths to occur — longer if you use that Feb. 6 date as the starting point. From then on, each additional 5,000 deaths took less than a week to accrue. At several points, the country added more than 10,000 deaths a week, particularly during the period in early April when New York was reaching its peak.

(Because tallies of deaths are calculated daily, days on which a milestone was passed were included in the figures for both it and the preceding milestone.)

How did those deaths spread? We plotted each 5,000-death increment by county, using the New York Times’s public data set of coronavirus cases. The dots below are roughly scaled to size, with small dots representing up to 50 deaths in the identified period and the largest dots representing up to 400.

You can see how the earliest chunk of cases is weighted to Washington state, where a cluster of deaths in a nursing home marked the first significant isolated toll in the United States. The center of the outbreak quickly moved to the New York City area and then to the broader Boston-to-D.C. corridor. Cook County, Ill., also stands out — where Chicago is — as a consistent center of numerous deaths. This is to some extent misleading: New York City has five counties, each displayed separately below.

All those little dots serve to show how the death toll from the virus was geographically distributed, again, unevenly. It does obscure, though, how much worse the virus was in certain parts of the country. Displaying the same data as a heat map, with darker red regions indicating more clustered deaths, shows how the mortality hot spots spread and shifted as the number of cases climbed.

New York quickly gets dark red. It spreads up to Boston and down to the District. But compare the first chart with the last (which includes data through May 26): The red is a bit more saturated east of the Mississippi now than it was for the first 5,000 deaths.

That comports with the broad decline in the number of cases in the Northeast as the number of deaths increased, with the number of deaths in other parts of the country making up the difference. (For the chart below, we used Census Bureau designations for regions.)

Remember that the time frame for the milestones reflected below varies. We’re adding 5,000 deaths more slowly than we had been, which is obviously good news. That’s in part because of the fade in new deaths in the Northeast, where deaths are coming more slowly than they had. That means other regions can add more new deaths over the lengthier time period.

At 5,000 deaths, the pandemic was a novelty. At 25,000 deaths, it was a narrowly contained wildfire. Now it’s a slow burn, claiming victims by hundreds every few hours across the country.

One death is a tragedy and 100,000 a statistic, to modify an old saying. But statistics still have use in delineating and describing ongoing tragedies.