The prevalence of remote learning in schools educating our most vulnerable students was no accident.
The prevalence of remote learning in schools educating our most vulnerable students was no accident.
I f you were a school superintendent considering whether to keep your district open in-person or move to online, how would you decide? Most people would suggest you look at COVID-19 case numbers in your community. Perhaps you would consider the vaccination rate, and if you had students with auto-immune disorders or other risk factors, maybe you would consider that. Most Americans would find these sorts of considerations reasonable.
As it turned out, this was far from what happened in American schools last year. An analysis of school-closing data on the nation’s 150 largest school districts reveals something entirely different. Rather than the progress of the disease in a local community, the most important predictor of remote schooling was a school district’s historical propensity to prioritize the interests of its teachers over the competing interests of its students.
We looked at specific ways districts favor teachers over students, such as prioritizing teacher seniority over new teachers and teacher performance, granting teachers more days off, and limiting the number of hours students spend in school each day. Districts that had historically scored high on these metrics were significantly more likely to opt for the remote-learning format last year. In aggregate, these measures of district-level teacher favoritism do far more to explain remote vs. in-person school decisions than every other variable we tested, including the COVID-19 infection rates in the community. When investigating the demographic features of school districts, we found that student-favoring districts were significantly different from teacher-favoring districts. Student-favoring districts were wealthier, whiter, and less urban and had a higher percentage of families who spoke English at home. However, even controlling for these demographic variables, teacher-favoring districts were far more likely to opt for remote learning.
Although there are legitimate reasons to worry about the health risks of in-person school for unvaccinated children, a mounting body of literature has demonstrated that remote instruction is detrimental to students’ learning. While the full effects of pandemic-driven remote school on America’s schoolchildren will not be known for years, it is already clear that remote school has hurt the average student and that the damage has fallen disproportionately on low-income students, urban students, and students of color. Students in these groups are more likely to be in a remote school and are less able to learn in a remote classroom due to resource disparities at both the school and the household level.
It is sad but perhaps unsurprising that more than one million students did not enroll in their local school during the first year of the pandemic, according to the New York Times. More disappointing is how disproportionately these students were from low-income neighborhoods. Unfortunately, the more vulnerable students in these teacher-favoring districts were already starting well behind their peers in student-favoring districts. Even before the COVID pandemic, students in these teacher-favoring districts had measurably worse educational outcomes: Students in teacher-favoring districts tested at a level one full year behind students in student-favoring districts and half a year behind when controlling for demographics and geography.
Let’s break down the numbers. According to our analysis, students in remote districts were significantly different from in-person-district students, across several important categories. They were more racially and socioeconomically diverse (66 percent vs. 48 percent were minorities and 67 percent vs. 51 percent received free lunches). Their elementary-school test scores in math and reading were almost a full year behind, and they were far more likely to drop out of school (20 percent chance vs. 14 percent). Even more significant were the differences in how these districts treated their teachers. Teachers in remote districts were more likely to be offered tenure (95 percent vs. 76 percent), more likely to have their pay determined primarily based upon seniority (maximum salaries were 195 percent of starting salaries vs. 172 percent for in-person districts), and less likely to have pay incentives based upon performance (39 percent vs. 56 percent), and in-district transfer priority was much more likely to be determined based upon seniority alone (62 percent vs. 36 percent). Even the school days at remote schools look different; remote schools had a shorter day length (7.3 vs. 7.6 hours), with teachers spending less time preparing lessons (29.7 vs. 41.5 additional minutes), and fewer additional days training and preparing (8.5 vs. 10). Overall, these numbers paint a damning picture and an outrageous one, especially if you are a parent in a teacher-favoring remote district.
How does this happen? One potential explanation is that wealthier, whiter, and more suburban/rural districts tend to have more politically involved parents, and so in those districts, there was a counterbalance to teachers’ understandable health concerns. In the lower-income, more racially and linguistically diverse urban districts, teachers’ desire for remote learning was significantly more likely to have won out over the interests of their students in having in-person school. Another factor is state politics and the power of special-interest groups.
Rather than using the euphemisms “teacher-favoring districts” and “student-favoring districts,” let us be more direct. If you want to know why your children are in Zoom school, look to your local teachers’ union. The more power it enjoys, the more likely it is that your kids will be in Zoom school, regardless of vaccination rates, infection rates, or emergency-room capacity. Media coverage to the contrary, this should not be surprising. After all, teachers’ unions are supposed to protect the interests of teachers, not students. Most of the time, those interests are somewhat aligned, but when — as with the COVID-19 pandemic — teachers’ interests come into conflict with the needs of students, teachers’ unions become a serious obstacle. As a society, we need to understand that and factor it into the political debate. Teachers’-union officials should not be interviewed in the press as if they were unbiased experts. Instead, the media should treat them skeptically, like other special-interest groups, such as the National Rifle Association.
The pandemic revealed large inequities throughout our society. This process has caused many to rethink their views of fairness and to seek to understand the structural forces behind these inequities. Teachers’ unions should be near the top of any list of structural features in society that need to be reevaluated. Without a wholesale rethinking of how to better represent the interests of our most vulnerable students, we are laying the foundation for another generation of inequity.
Noah Benjamin-Pollak is a research associate at Harvard Business School. Joshua Coval is the Jay O. Light Professor of Business Administration at Harvard Business School.