Regardless, Sen. John Cornyn (R-Tex.) did precisely that on Sunday afternoon, questioning why Dallas County (home of the city of Dallas) and Harris County (home of Houston) had similar death tolls but divergent case totals.
Fairly predictably, responses quickly centered on criticism of the question itself. The coronavirus pandemic has been neatly woven into America’s sharply partisan politics, particularly on social media — something Cornyn has played some small part in ensuring.
It is, nonetheless, an interesting question. Why has Dallas seen more deaths than Houston? The answer probably has to do with two factors: timing and location.
What the senator presented for his point of comparison was a snapshot of where the counties are at the moment. It’s useful to look instead at the trend — the number of cases and deaths in each place (and in the surrounding metropolitan regions) since the virus emerged.
On those graphs, you can see Cornyn’s point: The dashed Houston line representing the case total is much higher than that of Dallas — and has been since April — though the cumulative death tolls have been consistently similar. Put another way: 1.6 percent of Dallas County’s confirmed coronavirus cases have led to death. Only 1.1 percent of Houston’s have.
If we look at how the pandemic has evolved in each place, considering the seven-day averages of new cases and deaths in each place, we notice the same overall pattern. Houston and Dallas have both seen two surges. Houston saw one in early April and another, far larger one more recently, after containment restrictions were scaled back. Dallas saw one in mid-May and is seeing one now. As above, the daily number of deaths in the two places has been about equivalent.
The recency of the Houston surge is important. In the middle of June, Houston had seen 1.2 cases for every case in Dallas; now, it is seeing 1.5. Since June 15, Harris County has added more than 18,600 cases to Dallas’s 10,241.
Deaths generally lag behind case confirmations to some degree, often by as much as a month. Experts who spoke with The Washington Post indicated that the recent surge in new coronavirus cases across the South and West — including in Texas — would soon mean an uptick in the number of people dying of the virus. If and when that happens, we can expect the number of new deaths in Houston to outpace those in Dallas.
Put another way: Cornyn’s observation might be a temporary anomaly.
That said, the lagging death toll doesn’t entirely explain why Dallas’s death toll has tracked with Houston’s over the long term. Even on June 1, Dallas County was seeing a higher rate of death than Harris: 2.2 to 1.9 percent.
If we look at the mortality rate over time, we see that the rolling seven-day mortality rate in Dallas County has consistently been higher than in Houston, particularly in March and April.
It’s important to remember that, before May, the state was conducting fewer than 500 daily tests per million people, well below the national figure. In other words, there were a lot of cases before then that simply went unmeasured.
That has implications for the mortality rate and for Cornyn’s question. If the number of cases in Dallas were actually higher than recorded in March and April, that would explain the higher mortality rate. If there were 1,000 cases of the coronavirus but only 100 were confirmed through tests, 10 deaths from that population would be a 10 percent mortality rate according to the available data but, in reality, reflects only 1 percent of cases.
Put another way: If Dallas County was harder hit than Harris County in March and April, we might expect to see more deaths (a reflection of reality) than confirmed cases (a function of testing capacity).
There’s another factor, though. Many of the country’s coronavirus deaths have occurred in nursing homes and other long-term care facilities for elderly people. When the virus gets a foothold in a confined place where a number of high-risk individuals live, the results can be dramatic. Data from the state of Texas shows that while the number of cases confirmed in elder-care facilities in the Dallas and Houston regions have been similar, the virus has resulted in a larger death toll in the Dallas region.
These are large regions, encompassing more than the metropolitan areas, but the Dallas area has seen 72 more deaths in nursing homes and assisted-living facilities than the Houston area. Interestingly, the two regions had similar death tolls until mid-May, with 16.5 percent of cases in the Dallas region leading to death and 17 percent of cases in the Houston region leading to the same outcome. Since that point, 15.5 percent of confirmed cases in elder-care facilities in the Dallas region have resulted in death, compared with only 10 percent of those in the Houston area.
All of this said, the answer to Cornyn’s question is still encompassed in much of the same fog that surrounds data about the pandemic more broadly. It may be that Houston cases will soon grow more rapidly in parallel with the recent surge in cases. It may be that the Dallas region had more undetected cases earlier in the pandemic. It may be that Dallas has been less effective at tamping down the death toll in the most at-risk facilities.
Answering complicated questions like Cornyn’s demands not only a broad exploration of the available data but robust testing for the virus at the outset. It probably would have required, in other words, a different path from the one the federal government took months ago. Meaning Cornyn and his peers might have some questions of their own to answer.