There’s evidence that the presence of unions and community support for President Donald Trump had a significantly bigger influence on school districts’ decisions about holding in-person classes than the local spread of the coronavirus, two researchers say.
The working paper from two political science professors released earlier this month looked at more than 10,000 school districts’ reopening plans for this school year, and studied the correlation between those decisions and indicators based on politics, public health, and market forces. However, the paper also embodies the challenge of measuring how political considerations have driven school reopening decisions during the pandemic amid a flurry of factors. And some would disagree that political and not practical considerations have been the overriding factor in many districts’ decisions.
As cases of the virus continue to spread, the role that schools do or do not play in spreading the virus, and the wisdom of keeping school doors shut to try to contain the pandemic, have become divisive subjects.
There’s early data suggesting that reopening in-person classes isn’t a major contributor to the spread of the coronavirus. Some critics of districts say Trump’s fervent push for schools to reopen their doors has driven a corresponding—but overly fearful and misguided—push to keep classes remote among district leaders and unions in heavily Democratic areas where virus rates are relatively low. Meanwhile, they stress, students in poverty are being harmed the most by the lack of in-person classes, and other services and social bonds public schools can provide.
But other say that in general schools need more support and resources to reopen safely, regardless of what Trump or other powerful actors think. There’s some concern that data about virus cases used in these debates is being self-reported by schools, which have experienced challenges with tracking cases, and therefore should be treated with caution. Older teachers and those with underlying health conditions have expressed concern about whether they can return to their classes safely. And as the virus has hit communities of color especially hard, large shares of Black and Latino parents have indicated they’re leery of sending their kids back to schools.
The working paper from Leslie Finger, of the University of North Texas, and Michael Hartney, of Boston College, won’t resolve the ongoing dispute. But it might help fuel debates about what factors schools have relied on and should be relying on the most when making big decisions that must balance the safety and well-being of their students, staff, and communities.
“Public education’s response to the COVID-19 crisis appears to have become another manifestation of technocratic decision-making being swallowed up by our polarized and nationalized political debates,” Finger and Hartney write in the working paper.
Perhaps the main message of the study, “Politics, Markets, and Pandemics: Public Education’s Response to COVID-19,” is best captured by the following findings:
- Finger and Hartney found that compared to districts in counties where 40 percent of voters voted for Trump in 2016, districts in counties where 60 percent backed Trump were 17 percentage points less likely to shut their doors and go with remote learning; the likelihood of that outcome declined from 27 percent to 10 percent.
- In roughly 1,500 districts in which the authors could specify whether districts deal with collective bargaining units, the likelihood that a district with a local union conducts only remote learning is 40 percent, while the likelihood of a district that doesn’t deal with a local union conducting only remote classes stands at 15 percent.
- Compared to districts with zero cases per 10,000 residents, districts with 20 cases per 10,000 residents were more likely to reopen remotely by just 1 percentage point. (The paper relied on the average daily case rate over a 14-day period prior to late August.)
The relative density of Catholic schools in the area, which the authors used to measure the power of market forces in education during the pandemic, was more closely correlated to districts’ reopening decisions, but less closely correlated to those decisions than Trump’s 2016 vote share and the presence of unions.
For the roughly 10,000 districts in the paper’s overall findings, Finger and Hartney used district size as a proxy for union strength when they looked for—and found—a notable correlation between union influence and district reopening decisions. They say this is a good stand-in, based on federal data.
But to test the correlation further, Finger and Hartney ran their analysis again for the districts where they could find information about whether districts worked with a collective bargaining unit. As we noted above, they found that a strong correlation persisted, although of course that sample size of 1,500 is less than 20 percent of the paper’s overall sample size.
In an interview, Hartney noted that the study isn’t “dynamic” and didn’t track changes over time, and that he and Finger didn’t dive into minutes from individual school board meetings to see how politics shaped decisions. Hartney also stressed that the study doesn’t show—and he is not trying to prove—that local officials are acting in bad faith.
“We’re not out there claiming school board members are hacks,” Hartney said, adding later that, “If I were a teacher and I paid my union dues, I would absolutely expect my union to elevate my safety above other issues.”
Nevertheless, he said, the study’s findings—which have yet to be subject to peer review—are coherent. Citing an old cliché that, “There’s no Democratic or Republican way of cleaning the streets,” Hartney said the same has not proven to be true for school reopenings.
“Everybody should be concerned that on average it doesn’t appear to be the public health indicators that are driving decisionmakers,” Hartney said. “It seems like tribalism is driving decisions.”
It’s also important to keep in mind that guidance for schools about reopening during the pandemic differs significantly between states. And some states haven’t set forth such guidelines at all.
Some might question whether the paper uses the correct measure of COVID-19’s impact on the community. Could Finger and Hartney have used the recent hospitalization or mortality rates linked to the virus, for example, instead of the average daily case rate?
Hartney said that he and Finger wanted to pick a statistical trend that local decisionmakers were likely to be aware of compared to other metrics. He also said he hasn’t seen compelling evidence that different public health measures are more strongly correlated with districts’ decisions than political considerations.
Trump, Polls, and Parents
What about the paper’s finding that links voter support for Trump and district decisions? It’s similar to what a Brookings Institution report found about schools in Trump country. That July study relied on a much smaller sample size—a database of district information compiled by Education Week well before the start of the academic year for many schools— than the work from Finger and Hartney.
Certain caveats cover both findings on this front; keep in mind, for example, that district borders and the county boundaries used to tally up 2016 voting behavior often do not align. But beyond the 2016 election, of course, Trump’s rhetoric about schools during the pandemic has become a recurring feature of debates about the wisdom of reopening classrooms.
“The president nationalized the issue. There’s no way to study a counter-factual here,” Hartney told me.
Many Democrats took a dislike to the idea of reopening school buildings in the wake of Trump’s rhetoric. But the gap that’s perhaps drawn the most attention is between white parents who support the resumption of in-person learning and more-cautious Black and Latino parents. That disparity has persisted when parents of different races are surveyed, as Education Week’s own polling has found.
To what extent could unions be, however deliberately, representing nonwhite parents’ concerns in addition to their own interests? Hartney responded that the disparity between white and nonwhite parents isn’t as big as some might make it out to be.
He pointed to a Pew Research Center poll released in August showing that four out of 10 Black respondents, for example, supported either a mix of online and in-person learning (32 percent of Black respondents backed this option) or in-person classes five days a week (8 percent of Black respondents wanted this). That aggregate level of support for school buildings to open their doors at least part of the time, he said, might be a minority, but the 43 percent of those respondents who wanted full-time online school isn’t overwhelming either.
“It’s not the case that large majorities of any subsets of Americans ... are vigorously seeking only remote learning,” Hartney said.
Public opinion has obviously played a role at different levels. But it might be a mistake to think that this summer’s flood of national polls about schools, and the related national political tumult, has had a dramatic impact on local school officials’ decision-making process during the pandemic. One trend we wrote about in August, however, is that large shares of those surveyed were wary of or opposed to schools resuming in-person classes in the fall (here’s an example of a July poll with such a finding).
One more technical caveat about the Finger-Hartney study: A few states and jurisdictions have ordered school buildings closed or put in conditions districts must meet before in-person classes can resume. Hartney said the study controls for these fixed regulations.