The COVID-19 pandemic hit higher education on March 6, 2020 when the University of Washington became the first major U.S. university to cancel in-person classes and have students take courses and finals remotely. What followed was a tidal wave of shuttered campuses, canceled study abroad programs, and students and faculty alike scrambling to make sense of remote learning amid spiking infection rates across the country.
Surviving the end of the academic year, however, was only the beginning of a larger, looming higher education crisis. Over the summer, schools wrestled with the difficult question: what should be done about the fall? Davidson College’s College Crisis Initiative (C2i) collaboration with the Chronicle of Higher Education tracked colleges’ fall decisions, categorizing them as online, in-person, or some combination of the two. We use that data to examine trends in schools’ decision-making, shedding light on the contributing factors and eventual consequences of those decisions.
In our analysis, we find that while nearly every school in our sample offered in-person experiences in March, only about 30 percent were planning to maintain that in-person experience by September. Public schools and community colleges were particularly likely to go online. We examine some of the factors underlying those switches as well as the consequences. We find that differing financial pressures on public schools, private schools, and community colleges likely played a role in their decisions. We also find that those decisions appear to have had a significant impact on the enrollment decisions and educational outcomes of students attending each type of school, potentially changing the makeup of this generation’s college students. As schools began making decisions about the spring 2021 semester, it was especially important to reflect on what decisions were made in the fall, why they were made, and what effects these reopening plans could have on students.
Using the C2i panel data, we look at the reopening decisions of 2,958 colleges and universities over the course of the pandemic. The non-representative sample contains 1,588 public schools and 1,355 private schools. Four-year schools are overrepresented in the sample, with 2,007 four-year institutions and 936 two-year institutions. By far, the largest group in the sample is private, non-profit four-year schools, which represent 43 percent of the total sample.
Between March and September, roughly 1,800 schools changed their reopening plans at least once. The peak of switches came after mid-July, when many schools were scrapping their intended plans for the fall and making a new reopening decision.
Figure 1 looks at a subset of schools for which we have complete data on decision making. It shows that while only 24 schools in the sample offered exclusively online education when the pandemic began in March, 1,160 schools had decided to go exclusively online by July. By September, 50 schools that had intended to be in-person and 32 that intended to be hybrid as of July had also moved to being fully online. While most schools that made changes moved toward going online, 617 schools moved to hybrid models by July. That number decreased to 586 in September as more schools moved online. Finally, in September, 742 schools chose to remain entirely in-person. There was a significant amount of public concern about schools changing their plans at the last minute. However, Figure 1 shows that in fact most schools stuck with the decisions they had made earlier in the summer even with spikes in COVID-19 cases in some parts of the country.
In Figure 2, we break down the decisions made by each type of school as of September 10. We find that four-year private institutions had the highest share of schools choosing to remain in person in the fall, with only 381 schools, or 34 percent, going fully online. Community colleges had the highest share of online reopening decisions, with 543 schools, or 63 percent, choosing to entirely move away from in-person classes.
This disparity is likely in part due to a decade-long trend of moving toward online learning in higher education at two-year schools. Prior to the pandemic, community college students were more likely than their counterparts at four-year public or private institutions to be enrolled in at least one online course, with 14% of all community college students studying exclusively online (although not necessarily at an exclusively online institution) in 2018. However, the differential spike in online enrollment clearly extends beyond enrollment trends that predate the pandemic. We examine one reason for the trend in the next section.
Decisions on how to hold classes are complicated and involve many factors. However, we find that one important factor associated with those decisions is the financial makeup of different types of institutions. The three sectors mentioned above (four-year private, four-year public, and two-year) receive their funding in different ways, with additional differences between public and private two-year schools, as shown in Figure 3. While both private schools and public schools have faced pressure to keep their doors open, public schools seem to have had less financial incentive and fewer resources to do so.
Four-year colleges and universities, particularly private schools, tend to offer a residential component, while only 28% of two-year programs provided on-campus housing in 2015, though it is becoming more common. As a result, four-year schools are more vulnerable to a loss of revenue from canceling in-person experiences, such as residential housing, dining services, and parking. These services represent an average of $5,000 and $3,000 in revenue per student for private and public four-year institutions respectively, compared to less than $2,000 per student for both public and private two-year institutions, according to Figure 3.
When contrasting public and private institutions of higher education, other differences in funding arise. Private schools rely on tuition and fees, private gifts, and endowment funds for revenue, while public schools rely more heavily on state and local funding sources (Figure 3). Appropriations for public schools have been shrinking, with state funding being slashed by over $7 billion since 2008 leading to relatively constrained budgets for public schools. As a result, public schools had fewer resources than private schools to install technology like keycard systems, rapid testing, and other systems that made in-person learning possible during the pandemic.
Moreover, among all the different types of schools, endowment assets are most common in the private sector, and are highly concentrated. Not every school has utilized its endowment fund in the current crisis, as these funds are often thought of as intended for future investments. Nonetheless, endowments provide a cushion in funding that private two-year schools and public schools generally lack.
The Missing Cohort
This section highlights which kinds of students have forgone higher education in recent months or dropped out, likely in reaction to movements to online classes. Online learning can offer flexibility and accessibility, but it also can present challenges to learning outcomes. In particular, there are some groups of students for whom online education works poorly. Anticipating these challenges and more, 55 percent of students not returning for fall 2020 cited changes in class format.
Because of the nature of the populations that different schools serve, lower-income students are more likely to find their classes moved online. As shown above, two-year institutions were the most likely to move their classes from in-person to online. Roughly 15 percent of students at two-year schools come from families in the bottom income quintile, while the share is 9 percent of those at four-year public schools and 6 percent at four-year private schools. What this means is that low-income, underrepresented students are more likely to be struggling with the consequences of online schooling. And, if those students choose to drop out, that has dire consequences on lifetime earnings, given that those with bachelor’s degrees earn twice that of high school graduates.
Indeed, the discrepancies in experience have led to marked decreases in enrollment, especially for underprivileged students, for whom home environment or internet access may be more likely to present a challenge. New Census data suggests that students from households making less than $75,000 per year were almost twice as likely to forego college this year as those making more. The enrollment changes have also been marked along racial lines, with Black undergraduate students having been the most likely to see drops in enrollment for summer 2020. Community colleges, which typically serve students of color or of low-income backgrounds, accordingly held the largest share of enrollment drops for fall 2020, with first time student enrollment at public two-years plummeting by nearly 23 percent.
Compounding these problems, individuals who drop out are unlikely to re-enroll. From 2013-2018, within a group of 29 million observed students, 87 percent of college dropouts failed to re-enroll within the observation period.
Colleges have faced difficult trade-offs when deciding if, and how, to open their physical doors to students this year. While holding in-person classes can affect the health and safety of students, it is important to note the consequences of holding online classes for those who are typically underrepresented in institutions of higher education.
As the fallout of the pandemic continues to unfold, educational institutions must consider the broad impacts of online learning on retention, learning opportunities, and student achievement. They face difficult decisions for the upcoming semester and should choose policy solutions that protect and support the most academically vulnerable students.
Economic inequality persists in the United States, particularly along racial lines. A college degree is one significant pathway for lower-income young adults to improve lifetime earnings. All told, the greatest toll of the pandemic will be the disruption in upward mobility for the current cohort of low-income students.
The authors are grateful for valuable insight from Wendy Edelberg, Lauren Bauer, Jimmy O’Donnell, and Sarah Reber. Thank you to Chris Marsicano and Don Troop for their data collection, without which this analysis would not be possible. We would also like to thank Stephanie Lu and Moriah Macklin for incredibly valuable help with the research.