In-person voting didn't lead to spike in COVID-19, but concerns remain

Three weeks after Wisconsin residents cast ballots, researchers see no spike in COVID-19 cases attributable to in-person voting, though they say the effect from the election may be hidden in the numbers and difficult, if not impossible, to ever detect.

Predicted by some experts and officials, a surge in cases from the chaotic day of voting may have been prevented by precautions embraced at the polls. Wisconsin residents also appeared to follow the stay-at-home orders carefully in the days after the election, data show.

For roughly a week after the April 7 election, residents in most counties in the state rarely traveled from their homes and didn't gather in large numbers, according to cellphone data analyzed by the University of Wisconsin-Madison and other universities across the country.

In other words, the vote occurred during what appears to be Wisconsin's most compliant period of social distancing.

More recent data show that compliance with the stay-home orders is waning, both in Wisconsin and nationwide.

Live Updates:The latest on coronavirus in Wisconsin

Daily Digest:What you need to know about coronavirus in Wisconsin

Share Your Story:We want to talk to doctors, nurses and others affected by coronavirus

Starting around April 15 in Wisconsin, cellphone data show residents traveling from home more often, even though the state remained under a restrictive stay-at-home order. The same thing is happening in other states. The cell data, which tracks signals not individuals, show movement and gatherings.

It's unclear to experts what is behind the new movement and gatherings. It could be warming weather, increased discussion of government relaxing rules or simply "quarantine fatigue."

"I am actually worried about this," said Oguzhan Alagoz, an expert in infectious disease modeling at the University of Wisconsin at Madison who is using the cellphone data in his research. “It's not just mobility, but the question is how do I behave during that mobility. It makes a big difference. I hope people are being careful as they are moving.”

The movement trends are not uniform, Alagoz said. For instance, residents of Rock County are moving more than most, while Dane County residents appear to be more compliant with orders to stay home.

State health officials also have been monitoring this movement and gathering data.

"It is one piece of the puzzle that gives us a warning of a consequence of what we call quarantine fatigue," Dr. Ryan Westergaard, the chief medical officer at the Department of Health Services, said in an interview Tuesday with the Milwaukee Journal Sentinel.

"The potential impact of physical distancing interventions can really wane if people become fatigued by the required behavior changes," he said.

'Like a wild goose chase'

Alagoz and other public health experts predicted Wisconsin would see a surge in COVID -19 cases after the state Supreme Court overturned Gov. Tony Evers' last-minute order to halt in-person voting.

While the election saw a big increase in absentee voting, an estimated 413,000 people voted in person, including 18,803 in Milwaukee. There were only five polling locations in the city instead of the usual 180 because of a lack of poll workers. Voters, many wearing masks, waited in line for hours trying to stay 6 feet apart, even as workers scrambled to clean.   

Given the incubation period of the coronavirus, epidemiologists said a surge in new cases would have appeared statewide last week. Except for a slight increase that could have been explained by other reasons, no surge appeared.

"I don’t think that the in-person election led to a major effect, to my surprise. I expected it," Alagoz said, adding they don't expect to see a spike in the future.

Mustafa Hussein, an epidemiologist with the University of Wisconsin-Milwaukee, agreed there did not appear to be evidence of a surge of cases in the raw numbers. But he noted researchers lack crucial data regarding COVID-19.

The illness is so new and moving so quickly there hasn't been time to study it in a statistically sound way, the experts said. To learn the effect of voting, researchers would want to interview a large group of people who did and did not vote in person, but that would be very difficult.

Even without a surge, the experts maintain having an in-person vote heightened the exposure and likely increased the spread.

"Bringing people together in such large numbers over many hours at a high-contact rate per hour, does that pose a high rate of exposure for people? Yes, absolutely," Hussein said. "Did that materialize into an uptick in cases and deaths? We are not picking up a signal we can attribute to that event, and it is difficult to do without more nuanced data and sophisticated analysis."

One theory is that without the in-person election, the numbers of cases in Wisconsin would have dropped even lower. 

"A lot of this is speculative and difficult to pinpoint and say, 'The election caused this,'” Hussein said. "All of these factors completely muddy the waters and make this very difficult. It feels like a wild goose chase to go after. It is a valid question, an important one, but very difficult to answer."

Officials at the Department of Health Services said they also did not see a spike.

"We didn’t observe any overall significant surge in infections associated with that," Westergaard said. "There is reason to think some people may have been exposed based on interviews done locally but not enough to create a trend or signal that we can detect on a statewide basis."

City of Milwaukee Health Director Jeanette Kowalik said seven people may have contracted the virus through voting. She later raised it to 40, but she said more research is needed, which should be done later this week.

The state said about two dozen people may have been infected on election day. 

Some have characterized these numbers as an "uptick," but the experts are cautious.

"With the data we have, we can't prove an association," Westergaard said. "It would be speculative to say that was definitely the cause without really investigating closely and being clear that somebody really had no other potential exposure to infected people. I don’t think we have the resources to really do that to know definitely."

Brown County drives increase

Wisconsin saw an increase in daily COVID-19 positive cases starting April 22 and continuing for five days. The increase came at the same time as a bounce in testing. The state loosened protocol for who can be tested.

For that five-day period, the state recorded more than 1,000 cases. Over that same period, there was the highest amount of testing in the month of April. Regardless of the amount of testing, the percentage of positive tests has been 9% to 11% since April 1, an analysis of state data by the Journal Sentinel shows.

While the raw number of cases went up, the percentage of people being hospitalized for COVID-19 has dropped. In Milwaukee County, the number of patients in hospitals with the virus reached a high of 265 on April 9. Monday it stood at 177.

The number of deaths had been trending down, but on Tuesday the state reported 18 deaths, the highest daily total since April 4.

Westergaard said the spike of new cases during the five-day period appears to be driven by an outbreak at meatpacking plants in Brown County. 

That county had been accounting for less than 10% of new cases statewide until April 13, when cases there began to climb. On April 24, Brown County accounted for half of all the state's new cases.

There may be other reasons for Brown County's numbers besides the meatpacking plant outbreaks, but the experts doubt it was the election.

For that to be the case, Brown County voters would have had to act differently than voters in Milwaukee and other counties where no bump appeared.

The chance for infection in a meatpacking plant could be much greater than in-person voting and other sources, the experts said.

"That number of people packing meat, shoulder to shoulder, that is a much more concentrated environment than even voters in a polling station or walking down a grocery aisle or walking past someone at the park," Hussein said. "It is pretty reasonable to say the rise in cases in Brown County is driven by what happened in those plants."

Erin Caughey of the Journal Sentinel staff contributed to this report.

Contact John Diedrich at (414) 224-2408 or jdiedrich@journalsentinel.com. Follow him on Twitter at @john_diedrich, Instagram at @john_diedrich, LinkedIn or Facebook.