Exploring the relationship between COVID-19 and immediate 2-year college enrollment and persistence among Kalamazoo Promise scholars

INTRODUCTION

The onset of the COVID-19 pandemic swiftly altered U.S. postsecondary education. Statewide shelter-in-place orders in March 2020 forced educational institutions to immediately convert to virtual instruction (Collier et al., 2022). As the pandemic continued into the summer of 2020, economic constraints, health concerns, general uncertainty, and many other individual factors led to sharp declines in enrollment at public institutions. While enrollment at many public 4-year institutions has rebounded since the initial decline, enrollment at 2-year public institutions remains below pre-pandemic levels (National Student Clearinghouse, 2021). Nationally, enrollment at 2-year public institutions declined by 10.1% from Fall 2019 to Fall 2020 and fell again year-over-year in 2021 (3.4%), exacerbating pre-pandemic downward enrollment trends. These impacts were more severe than the enrollment changes in the public 4-year sector which saw a slight increase (.2%) from 2019 to 2020 and a decrease from 2020 to 2021 (3%; National Student Clearinghouse, 2021).

We have previously theorized that locales and states where tuition-free policies (a.k.a. Promises) exist may have eased or mitigated the effects of enrollment declines in the 2-year sector during the COVID-19 pandemic (Collier et al., 2021). However, early evidence suggests that not even the most generous Promise policies have insulated students and institutions from such enrollment downturns—including the Kalamazoo Promise (KPromise; Collier et al., 2021). KPromise is one of the most generous tuition-free programs—granting up to 100% tuition coverage to graduates of the Kalamazoo Public Schools (KPS) system and providing choices for students to engage in 2-year or 4-year institutions. Over the past 16 years of the scholarship, the local 2- year public institution Kalamazoo Valley Community College (KVCC) has served as a primary enrollment destination for KPromise students with about a third of KPromise students attending KVCC in any given pre-pandemic term.

Research on programs like KPromise can shed light on the potential of tuition-free programs to mitigate shocks to postsecondary enrollment like that caused by the pandemic. The local context of the Kalamazoo education and policy environment serves as an important case study, particularly because of the unique generosity and flexibility of KPromise. Data collection and findings on KPromise can serve as a benchmark for other tuition-free programs, 2-year institutions, and communities grappling with educating their youth. While every community and education system will undoubtedly experience different effects and outcomes, case studies can provide benchmarks for other local, state, and national programs. Thus, this chapter is guided by two overarching questions. First, how did immediate college enrollment of KPromise students at KVCC change from pre-COVID-19 (2019 cohort) to during COVID-19 (2020/2021 cohorts)? Secondly, how did KPromise student first-year stop out from KVCC change from the 2019 cohort to the 2020/2021 cohorts?

COVID-19 AND COMMUNITY COLLEGES

The pandemic will likely be framed as a disrupting event that will shape all ongoing research and will be the genesis of many completely new research questions. One concern related to the COVID-19 pandemic has been the steep decline in undergraduate enrollment. From Fall 2019 to Fall 2021, higher education experienced an overall 8% drop in undergraduate enrollment. Two-year public institutions experienced the steepest decline at 15% (National Student Clearinghouse, 2021). Before the pandemic, about 53% of students nationwide who began college at a 2-year institution returned their second fall to the same institution. For the class of 2019, the first-year retention rate fell to 51.6%, an indication that 2-year institutions across the board struggled to retain students at the same rates as before (National Student Clearinghouse, 2019). The financial pressures associated with lower enrollment and weakened retention have forced some 2-year college systems and institutions to reduce staff (Gramza, 2022; Weissman, 2022). Additionally, attracting talent may be increasingly difficult as more faculty and staff choose non-higher education pathways (Flaherty, 2022; Morales, 2022). For the foreseeable future, community colleges will be running lean and will be expected to do more with less money and fewer employees.

Enrollment declines may be due to various new realities. Wages have increased from the start of the pandemic in March 2020 when the average hourly wage for all private employees was $29.79 to $31.95 in the first quarter of 2022 (U.S. Bureau of Labor Statistics, 2022a). For individuals with a high school degree, the 2020 first-quarter median weekly earnings were $768, and by the first quarter of 2022 were $827 (U.S. Bureau of Labor Statistics, 2022b). An increase of $256 per month is likely enough to encourage some would-be students to choose engagement in the workforce over enrollment in community college.

Additionally, 4-year universities engaged in behaviors that are likely siphoning enrollment from community colleges. For example, many universities converted to test-optional admissions (Jaschik, 2022). At private institutions, test-optional policies were previously related to increased admissions for Pell-eligible students, non-White students, and women (Bennett, 2022). Among public institutions, test-optional policies have been linked with an increased number of applications, which in turn diversifies the student body through demographic changes (Schultz & Backstrom, 2021). We found suggestive evidence for just this trend among KPromise students as well who increased their applications to 4-year institutions (Collier et al., 2021).

Furthermore, both students and administrators expressed that instructional change, financial constraints, and mental health challenges led to higher barriers to college student success during the COVID-era (Hart et al., 2021; Prokes & Housel, 2021). Interviews with students confirm that the pandemic created strained relationships between students and their institutions, primarily driven by university decisions about COVID regulations (Vaterlaus et al., 2021). Within the context of tuition-free college, adults interviewed as part of the Tennessee Reconnect program expressed similar sentiments of disconnection from colleges (Collom & Cooper, 2022)—suggesting that tuition-free policies alone are unable to help community colleges stem enrollment declines or contribute to an immediate bounce back.

THE KALAMAZOO CONTEXT

The Kalamazoo Public Schools (KPS) district is a mid-sized, majority-poverty, and majority-minority school district in Michigan. The average total enrollment in the district is around 12,500 students, about 67% are eligible for free-or-reduced-price lunch, and the largest racial subgroups that comprise the district are Black students (38%), white students (35%) and Hispanic students (14%, MI School Data, 2022). The district is home to KPromise, a private, anonymously funded grant that awards up to 100% tuition coverage at any public Michigan postsecondary institution and 15 private institutions. KPromise is a first-dollar award, meaning it is applied to a student's tuition bill before other funding like Pell Grants. Eligibility for the scholarship is based solely on enrollment and residency in the Kalamazoo Public School (KPS) district. Students must enroll for a minimum of 4-years of high school (and graduate) to receive 65% of the total cost of tuition and mandatory fees—and students who begin enrollment in kindergarten receive 100%. During college, scholars must maintain a 2.0 GPA, which is the standard threshold for academic progress at most institutions (The Kalamazoo Promise, 2022).

At some point, more than half of all KPromise students will enroll at KVCC, and during any given pre-pandemic term about 27% of scholars were enrolled at KVCC (W.E. Upjohn Institute, 2022). Additionally, two employees at KVCC are funded by the program to advise and support KPromise scholars. Because no other institution receives more KPromise students, understanding the unique ways in which the pandemic impacted enrollment and persistence is important for understanding how to best assist recipients of tuition-free awards to start and return to college.

Early evidence of pandemic effects on the KPromise population and the local community college showed declines in the rate of new high school graduates enrolling for the first time at KVCC, from 35% of graduates in 2019 to 17% of graduates in 2020. These declines can be explained both by students with fewer socioeconomic advantages who likely would have attended KVCC opting not to enroll in college, and newly test-optional universities in Michigan skimming academically stronger, mostly White students who otherwise would have enrolled in KVCC (Collier et al., 2021). This early work began to illuminate the effects of COVID on the KPromise student population, and now that some additional time has passed, the study of what the longer-term effects of the pandemic might be for this population is possible.

DATA AND SAMPLE

This study used a dataset comprised of student-level administrative data from the Kalamazoo Promise. These data were acquired through a data sharing agreement, and include demographic details, KPromise eligibility and payment information, enrollment details, and grade records. About 90% of each graduating cohort from the public schools is eligible for Promise funding, but we include all graduates in our sample and generally describe all KPS grads as KPromise students, consistent with prior research (Collier & McMullen, 2020). In the case of ineligible students that enroll at KVCC, these students comprise a minuscule portion of the sample (0.6%), and we can reliably observe the college enrollment patterns of other ineligible students by using the StudentTracker report from the National Student Clearinghouse. Our sample includes KPS graduates from three recent classes, 2019, 2020, and 2021. We restricted our sample to just those graduates that earned a traditional high school diploma from one of two traditional high schools (N = 1770) or one alternative school in the district (N = 218) for a total of N = 1988. In our sample, 51% of students are Female, 39% are white, 37% are Black, and 56% are eligible for Free or Reduced-price Lunch.

METHOD

We examined two binary outcomes of immediate college enrollment at KVCC and stopping out within the first year of college. Consistent with Collier and McMullen (2020, 2021), immediate college enrollment was defined as starting college by the first fall after spring high school graduation. For regressions on first-year stop-out, we restricted our sample to individuals who enrolled in KVCC immediately after high school graduation (N = 377), and we defined first-year stop-out as not enrolling at KVCC for the second fall after high school graduation, conditional on starting college the first fall after high school. These decisions and definitions are consistent with prior work on KPromise student stop-out (see: Collier & McMullen, 2020, 2021). For stop out, we compared the 2019 and 2020 cohorts as the class of 2021 had not yet reached their second fall of enrollment at the time of writing.

Both outcomes were binary, therefore we conducted logistic regression analysis and report the marginal effects in terms of percentage points. Marginal effects indicate the positive or negative direction of the relationship between the independent and outcome variables (see: Williams, 2012). For example, we found that students from the class of 2021 were 10 percentage points (M.E. = –0.10) less likely to immediately enroll at KVCC compared to the class of 2019.

Our models included students’ race, sex, and free-or-reduced price lunch status, as well as the graduating cohort, high school, and final cumulative high school GPA. For our regressions on first-year stop out, we also included the first-year GPA of students at KVCC, and missing flags for both GPA measures as some students were missing GPAs in our data, an approach previously used for analyses on the KPromise population (see: Collier et al., 2021). We converted both GPA measures from continuous to categorical with five categories. Our reference groups for the logistic regressions were the class of 2019 (non-COVID class), one of the two traditional high schools, men, White students, non-FRL eligible students, and those in the middle category of each GPA variable (2.5–2.9).

RESULTS

The results of the study are reported for enrollment and first-year retention. As noted above, overall enrollments in community colleges across the nation declined 15% from Fall 2019 to Fall 2021 (National Student Clearinghouse, 2021).

DISCUSSION

The severe enrollment declines revealed at KVCC for students using the KPromise scholarship are troubling. Outcomes suggest that KVCC enrollment was primarily harmed by students choosing not to enroll in KVCC. We found year-over-year declines in KVCC enrollment—echoing wider trends for 2-year public institutions (National Student Clearinghouse, 2021). Although these changes vary by size across different demographic groups, all groups experienced declines in both years. Because KVCC is the geographically closest open-access institution where KPromise-eligible students may use their award, these findings have troubling implications for the academic progress of KPromise students in the years to come and illustrate an important limitation of tuition-free college.

One factor that may be contributing to enrollment drops is the flexibility of KPromise. With 10 years of eligibility, students facing continued uncertainty about the future direction of COVID and economic policies may simply choose to put off education until they are clearer on their path forward. Because KPromise does not penalize this choice, the policy's flexibility may be a contributing factor in current enrollment declines. Research demonstrates that if a student does not start college within a year of high school graduation, they are much less likely to ever begin, or to earn any kind of credential (Burke, 2020). Even though KPromise students have different options than the average graduating high schooler, it remains likely that many of these KVCC non-starters will never start college.

KVCC enrollments may be secondarily affected by 4-year institutions that engaged in test-optional admissions, including the two local KPromise-eligible institutions (Western Michigan University and Kalamazoo College). Likely, some students who would normally enroll at KVCC chose a 4-year institution instead. The percentage of KPromise students who started college that enrolled at a 4-year institution increased from 55% in 2019 to 61% in 2020 and 69% in 2021. This “trade-up” issue could be a net positive for KPromise and individual students, if they can earn a degree. However, from KVCC's perspective, this enrollment pattern is indeed problematic. Given that 4-year institutions have widely responded to the pandemic using these behaviors (see Nickel, 2021), it is likely that other 2-year institutions are experiencing the same squeeze-out issues in enrollment. Although this outcome is probably suboptimal for 2-year institutions, as these students were likely amongst their strongest, the trend could be seen as a net-positive outcome if these students can attain 4-year degrees or credentials.

While we focused on KVCC, it is important to consider in this context whether 4-year institutions are prepared to address the needs of students who would normally choose to enroll at a 2-year institution. An increase in stop out among these students at the 4-year campuses could result in higher loan debt loads and limited opportunities to repay, a frankly unacceptable policy outcome. This outcome possibility is a reality even for KPromise students, who have lower debt loads than pre-Promise cohorts but still have some debt (Bolter & McMullen, 2022). It is of great concern what happens to students who were able to trade up, but are unable to persist. In this vein, we call for more research trying to understand if this “trade-up” exists more widely, for whom, the retention rates of these students, and how institutions have improved or increased support for these students.

Our suggestive findings on first-year stop out begin to outline the severe effects that the pandemic had on students enrolled at KVCC. At over 63%, the first-year stop out of the 2019 cohort was astonishingly high at nearly 20 points higher than established historical trends. This outcome illustrates the extreme, immediate impact that the pandemic had on students’ persistence (Collier & McMullen, 2020). While remaining higher than established trends, fortunately, the stop-out rate has declined for the 2020 cohort to 57%. Although the regression illustrates that the 2020 cohort was less likely to stop out, the outcome is non-significant—which is likely due to a limited sample size.

Despite lacking a significant outcome, the subsiding of a first-year stop-out is encouraging and may illustrate that both students and KVCC have begun to adapt to the “new normal” of COVID-19. This stabilization is especially critical because the demographic makeup of KPromise students at KVCC has shifted—particularly, the proportion of students entering with lower high school GPAs below 2.5 has increased from about 30% of entering students to 59% when combining the COVID cohorts. The other student demographics for KVCC enrollment during COVID have not dramatically shifted on the level as those with below-average high school GPA—for example, the following student demographics of pre-COVID to 2020 enrollment patterns are as follows: Black (39% vs. 40%), Hispanic (20% vs. 18%), White, (32% vs. 40%), Female (45% vs. 51%), and FRL-eligible (66% vs. 65%). Although we promote caution in the interpretation of these data, this outcome is yet another signal that students with higher GPAs were pulled to other institutions. Both KVCC and KPromise must acknowledge that the population at 2-year institutions has changed and may require even more academic support than previous cohorts. This is an important pattern that should be explored further as more cohorts graduate high school.

While KVCC currently has some KPromise-supported infrastructure in place to support KPromise scholars, evidence indicates that the most marginalized student groups are expected to continue to struggle at KVCC (see—Collier & McMullen, 2020, 2021). As such, the current levels of support are unlikely sufficient to return trends to pre-pandemic projections. Therefore, increased socio-emotional, academic, and perhaps even financial supports are likely necessary just to simply return to “normal.”

Finally, we found no associated effect on first-year stop-out for students who were FRL eligible. Aligning with our prior work on more recent cohorts (Collier & McMullen, 2021), while many financial barriers remain for students from the Kalamazoo district, those barriers themselves appear to be exerting less influence on actual academic outcomes. To be clear, we are not saying that tuition-free policy eliminates the socioeconomic effects on academic performance, that would be false especially because pre-college performance is tied with socioeconomic advantage and that remains an important factor in performance and persistence pre-pandemic (Collier & McMullen, 2021) and as this study shows, in the pandemic. However, for college performance and persistence, the effects are lessened which means that KPromise or KVCC could focus more resources and attention on academic and racially-based supports. Again, due to the limited sample size of this model, we urge caution here—even if the outcome is aligned with our stronger research. Additionally, this finding may be unique to KPromise due to the generous terms of the scholarship and we do not expect to see this finding from less generous tuition-free policies (last-dollar programs). However, we cannot be sure of this outcome until more researchers engage in similar analyses.

LIMITATIONS AND FUTURE RESEARCH

These analyses are a mixture of descriptive and inferential and the outcomes should be considered within these contexts. Specifically, we urge much caution with the first-year stop-out outcomes given that we could only use one pandemic cohort—and that additional data on other cohorts may yield different outcomes. Additionally, because KPromise is among the most generous and flexible tuition-free policies, these findings do not speak to what analyses on other programs might find. However, we believe the evidence presented here can be a road map for similar research and comparisons.

Although it is tempting to make recommendations for other programs, more data and evidence on other programs and other contexts are needed to construct a clear picture of tuition-free program patterns during COVID-19. It is important that other programs take close account of their students that are not enrolling, and those that are perhaps choosing to enroll at a 2- versus 4-year institution. Our evidence is an important first step to understanding tuition-free student enrollment during the pandemic, but more evidence is needed to draw concrete policy recommendations for other programs.

Finally, our research remains an early effort at understanding how trends have changed and we need qualitative studies to examine why more students remain disengaged from higher education. While it is fair to assume that increased wages may be one mechanism that keeps more students out of college, we have little empirical knowledge suggesting this is the case for the average student who has chosen not to engage—there could be other issues not explicitly considered in this manuscript that qualitative inquiry may highlight. Finally, this early research on the COVID-era has illustrated that prior trends are no longer the “norm” and while it may be valuable to understand what tuition-free policies could accomplish under better conditions and aspire to return to those outcomes—we must better understand how the COVID-era has changed what we “know.”

CONCLUSION

Our findings suggest that even one of the most generous tuition-free programs could not mitigate the effects of COVID-19 on enrollment and persistence at a public 2-year institution. Enrollment of KPromise students at KVCC declined precipitously, and first-year stop-out increased substantially. While economic conditions may be encouraging some students to remain in the workforce, it is still troubling that so many individuals are not pursuing postsecondary credentials. Programs like KPromise and institutions like KVCC need to do more to address student needs and concerns to help them enroll and persist. While financial support may be a factor, the evidence from KPromise shows that other student concerns must be addressed including increased academic support, likely support for physical and mental health, and more.

The ongoing effects of the COVID-19 pandemic on 2-year institutions will have many implications moving forward. One pivotal concern for KVCC as a result of these declining enrollments is the decrease in funding from the KPromise program. The financial losses for KVCC from KPromise students alone are likely in the hundreds of thousands of dollars, and this will have implications for the remaining KPromise scholars enrolled there, as well as other students. In addition, our evidence suggests the population currently enrolled at KVCC likely has increased academic and socio-emotional needs than prior cohorts. As this dynamic plays out across the county, institutions will grapple with fewer resources and a student body with greater needs and must find creative solutions to continue to support the students they do have. It is also imperative that both KPromise and KVCC increase their outreach efforts to reach those students who are not enrolled and are not working in high-paying jobs. While KPromise students have lengthy eligibility, the sooner that these students can be enrolled and reach a credential, the better for their long-term economic stability, as well as that of KVCC.