| |
Wisconsin April Primary Election Did Not Boost COVID-19 Cases
|
| |
| |
Mark Mascolini
An April 7 primary election in Wisconsin, in which 453,222 people voted in person at the polls, had no impact on the state's daily COVID-19 new-case rate, according to an analysis comparing new cases before and after the primary in Wisconsin with new cases across the United States [1]. But the researchers who conducted the study note that coronavirus worries boosted absentee voting in the Middle West state to 1,551,711, more than triple the in-person voter turnout. The report appears as a medRxiv preprint that has not had peer review.
Concern about the highly contagious COVID-19 virus, SARS-CoV-2, has already upset 2020 elections across the United States, including the presidential race. Vast public rallies, the meat and bones of US presidential campaigning, have disappeared. Most states delayed primary elections, in which political party members pick candidates for the November 2020 contest, while COVID-19 continued its breakneck spread. Despite protests from Democratic presidential candidate Bernie Sanders and many others-including shaken Wisconsin voters-the state went ahead with its in-person April 7 primary.
Researchers from Florida, Alabama, and Wisconsin analyzed COVID-19 rates before and after the primary election in Wisconsin, comparing them to rates across the United States. The investigators tracked daily new cases through online national and local sources for the United States, all of Wisconsin, and the three largest voter contingents in Wisconsin: Milwaukee county, Dane county, and Waukesha county. For each of those 5 blocs, they calculated daily new-case rates per 100,000 voting-age adults. The national US data excluded Wisconsin residents.
For each of these 5 jurisdictions, the investigators figured the new COVID case rate for the 10 days before the Wisconsin primary election and for April 12 through 21, that is, the 10 days following the median SARS-CoV-2 incubation period of 5 days after the April 7 vote.
The ratio of Wisconsin's new-case rate to the US new-case rate fell from the 10 days before April 7 voting to the 10-day period starting on April 12 (Table). The decline compared with the US rate held true for the entire state of Wisconsin and for the three largest counties, Milwaukee, Dane, and Waukesha:
The authors explain that these ratios indicate that the new-case rate in Wisconsin versus the rest of the United States decreased after the election compared with what one would expect if the relationship between Wisconsin and the rest of the United States continued at its preelection ratio.
The strongest evidence that the Wisconsin primary did not affect COVID-19 incidence would come from a trial randomizing people to vote in person or vote from home, then monitoring emergence of new cases in each group. But the researchers note that such a trial would be unethical. "The next best option," they suggest, would be a comparison of two voter groups matched for COVID-19 risk factors like age, gender, race, comorbidities, sick contacts, and so on. But that approach would be "arduous" and "impractical." Thus the authors propose that their practical approach gives a reasonable estimate of how in-person voting affected new COVID-19 rates in Wisconsin.
The researchers note that concern over COVID-19 spreading at the polls "caused considerable turmoil" in Wisconsin in the days before the election and led to an increase in absentee voting "that may have been a large factor in preventing an increase in COVID-19 activity." They add that voters at high risk for COVID-19 may have "self-selected themselves out of the live voting process," and protective measures at the polls may have damped down spread of SARS-CoV-2. One might also surmise that in-person voting in a state with more voters and a higher election-day COVID-19 prevalence than Wisconsin could lead to a postelection bump in new cases.
These results offer some reassurance that almost a half-million people can vote in person on a single day during the COVID-19 epidemic without inflating new-case rates. Wisconsin primary data also underscore the readiness of voters to cast their ballots from home before election day. Turnout will surely be much greater for the general election in November. For example, for the Clinton-Trump 2016 contest, about 644,000 people cast ballots in New York City's Manhattan borough alone [2]. In the 2016 election, Wisconsin counted 2,976,150 voters [3].
References
1. Berry AC, Mulekar MS, Berry BB. Wisconsin April 2020 election not associated with increase in COVID-19 infection rates. MedRxiv. https://doi.org/10.1101/2020.04.23.20074575 (The report appears as a medRxiv preprint that has not had peer review.)
2. New York Times. New York results. August 1, 2017. https://www.nytimes.com/elections/2016/results/new-york
3. New York Times. Wisconsin results. August 1, 2017. https://www.nytimes.com/elections/2016/results/wisconsin
|
|
| |
| |
|
|
|
