Schools and their leaders had a decision to make last summer. Should schools open fully remote, fully in person, or somewhere in between? Different leaders made different decisions. The question is, why?
We wish that this question were merely academic, interesting from a 鈥渨hat happened last summer?鈥� perspective. But unfortunately, with the rise of the Delta variant and recent spikes in coronavirus cases, we鈥檒l be debating the same questions as the beginning of next school year draws near.
Together with our colleagues, we explored what motivated school reopening decisions in the 2020-21 school year, both and . The results turned out to be complicated. Ideally, health considerations would have driven school reopening decisions last summer. Specifically, in places where COVID-19 was widespread and growing, schools would go fully remote. In other places, they would open in person, at least until a threat emerged.
Early research, however, suggested that the strongest predictor was partisanship. Schools in Republican counties opened in person, and those in Democratic counties opened remotely or, sometimes, in a hybrid model. Similarly, many teachers鈥� unions, whose leaders are more politically aligned with Democrats, were vocal in their call for remote education. It is not surprising that politics played a role in the decision to reopen school buildings. Then-President Donald Trump and then-Secretary of Education Betsy DeVos strongly urged schools to open in person.
However, the reality behind school building openings is much more complicated than just politics. Cause and effect are always difficult to tease out. Correlation is not causation, and the problem with these correlational analyses is that race, ethnicity, income, occupation, and population density are all strong predictors of both political affiliation and COVID-19 health risk. And it is very difficult to disentangle them.
For instance, people of color have contracted, been hospitalized for, and died from COVID-19 at higher rates than white people. People of color are also much more likely to vote Democratic. Could it be that this population鈥檚 greater health risk is driving the apparent relationship between Democratic counties and a lower likelihood of reopening in-person schooling?
COVID-19 transmission is, additionally, more likely in large and densely populated areas where people live in closer proximity and with greater physical contact. Population density also tends to be higher in places where there are more people of color, more people working in personal-service industries, and with a greater share of Democrats. It is difficult to isolate one of these interconnected factors as the leading cause of school building closures.
Also, unions鈥� calls for remote instruction may not really have been rooted in politics. Conflating the two obscures teachers鈥� unions鈥� legitimate concern about the health risk to teachers, school staff, and students. This distinction reinforces the difficulty of isolating political and health factors when they鈥檙e intertwined.
One way of handling this jumble of potential causes is to predict school opening decisions based on all these (and other) factors, holding each one constant. In both studies鈥攐f reopening decisions in Michigan and nationally鈥攖hat鈥檚 what we did. Here is what we found:
- Politics and partisanship were indeed important. Consistent with other studies, ours found that a higher Democratic vote share was associated with more remote instruction. The results are similar when we focused on polling data.
- Districts with stronger teachers鈥� unions were more likely to open remotely. However, this relationship was not as strong as with Democratic vote share and less significant than what some anti-union critics have suggested.
- Demographics鈥攅specially race and poverty鈥攁lso strongly predicted fall 2020 in-person school reopenings. Specifically, school districts with more Black and Hispanic residents were more likely to have remote instruction. Other factors being equal, districts with more people living in poverty and in large, densely populated areas were also more likely to have remote instruction.
- Direct health measures also predicted school reopening decisions. Districts with higher rates of COVID-19 were less likely to open for in-person schooling. However, after controlling for the other measures, COVID-19 played a smaller role than politics and demographics.
We also examined factors like school funding and broadband access, which, when holding constant all the above factors, did not predict school reopenings.
What do all these data tell us about what drove school reopening decisions? There is no simple answer. Our best guess is that the direct and indirect measures of COVID-19 health risks, including demographics, all affected school reopenings, as did identity-oriented politics and polarization. Some groups were more likely to favor reopening schools in person because of their personal circumstances. Their political leaders seemed to mostly follow the leads of their constituents.
Studying the causes behind school reopening decisions raises a related question: Why aren鈥檛 more people getting vaccinated? Why are so many people risking their own lives鈥攁nd the lives of their parents, siblings, children, neighbors, church members, and others鈥攂y refusing the vaccine? After all, the vaccines are safe, effective, and widely available to anyone 12 and older. Here, too, politics plays some role, but it is likely not the only factor.
Whatever role politics played in prolonging the pandemic, it is almost certainly part of the solution. Our political leaders鈥攍ocal, state, and federal鈥攏ow need to step up. They can no longer simply follow the loudest crowd. They must lead with accurate information and focus attention on the damage being done by low vaccination rates and the resulting consequences for our communities, including schools. We need leadership and we need it fast.
