With the onset of the global pandemic caused by COVID-19, nearly all K-12 students in the United States have experienced an unprecedented interruption to their formal schooling. Students and parents with lack of access to technological devices, high-speed internet, and information to navigate online learning are among the most likely to face growing inequalities. This includes many low-income and rural families. As Susan Dynarski described in her recent New York Times article, “[COVID-19] has exposed and intensified enormous gaps in schools’ and families’ capacity to support children’s learning.”
We agree that the circumstances are dire. The current situation is without precedent, so researchers and parents alike are scrambling to find out the effect of near-universal online instruction. To this end, one case worth considering is a relatively new learning environment with a set of schools that uniquely deliver instruction in an online mode: virtual charter schools. While some insights can be gained, we recommend caution when comparing the two cases.
Virtual charter schools deliver instruction in online or blended (online with some face-to-face instruction) formats and operate in 21 states. Like other charter schools, virtual charters are publicly funded, do not have admissions criteria, and have their charter (or contract) authorized by a state-approved entity. Families may choose to enroll their children in a virtual charter school for myriad reasons. At first glance, virtual charter schools offer students and their families the ability to tailor learning experiences to specialized needs. However, if such advantages are available to students in virtual charters, there is no evidence that the benefits transfer into gains on student test scores.
We recently published a peer-reviewed study in Educational Researcher in which we examine the effects of attending a virtual charter school in Indiana on student outcomes in grades 3-8. In the study, we analyze longitudinal student records provided by the Indiana Department of Education from 2011-2017. As of the 2016-17 school year, Indiana had 10,984 K-12 students enrolled in four virtual charter schools. Because the state maintains administrative records in concert with annual assessments (called the ISTEP), we were able to analyze the performance of students in virtual charter schools. We do this by identifying students who switch from traditional public schools to virtual charter schools. We match these students to their traditional public school classmates with similar characteristics (i.e., race/ethnicity, sex, poverty status, and achievement), and then compare their performance in virtual charter schools to their former peers.
We find the impact of attending a virtual charter on student achievement is uniformly and profoundly negative, equating to a third of a standard deviation in English/language arts (ELA) and a half of a standard deviation in math. This equates to a loss of roughly 11 percentile points in ELA and 16 percentile points in math for an average virtual charter student at baseline as compared to their public school peers (see Figure 1 above). There is no evidence that virtual charter students improve in subsequent years. We could not “explain away” these findings by looking at various teacher or classroom characteristics. We also use the same methodology to analyze the impact of attending brick-and-mortar charter schools. In contrast, we find that students who attended brick-and-mortar charters have achievement no different from their traditional public school peers (see Figure 2 below). Our confidence in these results is further buoyed by other studies of virtual charter schools in Ohio and nationwide having similar findings.
It is entirely possible that such dismal findings of virtual charter schools seemingly mirror the anticipated negative consequences of online instruction stemming from coronavirus-related school closures. However, we hesitate to connect the two because of many notable contrasts.
On one hand, some differences suggest that students in public schools transitioning to online learning resulting from COVID-19 may end up having worse outcomes than a typical virtual charter school student. Virtual charter school operators have an established infrastructure to deliver online learning and students self-select into these schools. These schools also engaged students in online learning environments and provided resources for them long before the pandemic. Meanwhile, most traditional public schools and districts (and even brick-and-mortar charter schools and private schools) had minimal online learning provisions and infrastructure for students and their families prior to COVID-19. In fact, many established virtual charter providers have advertised resources and consulted with public schools that are rapidly transitioning to online instruction.
On the other hand, some differences suggest that students in public schools might not experience as steep of a setback as peers who attended a virtual charter year-round. In our study, we find that virtual charter schools have an average of 100 students per class, which is substantially higher than their counterparts in traditional public schools (24) and brick-and-mortar charters (22). We have no evidence that student-to-teacher ratios in traditional public schools have increased as a result of the shift to online learning. Compounding this issue is the fact that all virtual charter schools in Indiana (and most virtual charters nationwide) are operated by for-profit educational management organizations. We do not know the degree to which these organizations are driven by profit maximization, which may have negative consequences on student learning, but we do know that virtual charter schools in Indiana are mired in financial scandal.
In addition, students who transitioned to online instruction in public schools in the past few months have spent most of the school year with their teachers. The relationships they have built in person could ease the transition to virtual learning. Despite the difficult circumstances of the current conditions, teachers in public schools across the country appear to be putting forth determined efforts to meet the needs of their students.
Considering these issues in conjunction with the research evidence, we believe that virtual charter schools are ill-equipped to take on a more prominent role in light of this global crisis, and recommend that both parents and school administrators be extremely wary of virtual charters’ attempts to expand during this crisis. Based on their dismal track record, policymakers should instead focus on greater oversight and accountability for these schools. Perhaps the worst policy response during the COVID-19 crisis is to promote these schools, though this is precisely what Education Secretary Betsy DeVos is currently pursuing.
At the same time, we simply cannot say whether the poor performance of virtual charter schools reflects a disadvantage inherent to online instruction, the unique disfunction of this particular sector of schools, or some combination of the two factors. As a result, we are hesitant to ascribe the magnitude of virtual charters’ deleterious impacts on student outcomes to the online learning environment resulting from COVID-19.
We view the dismal performance of virtual charter schools in the range of possible performance outcomes for students transitioning to virtual schooling, though we genuinely hope most students outperform this benchmark. Although we expect most students will be worse off because of the forced transition to online instruction, we suspect that whether or not the harm of the transition is comparable to students attending virtual charter schools will depend on how well individual schools managed the transition.
As most school systems appear to be making good-faith efforts to continue engaging students’ growth during closures, the virtual charter result is likely a worst-case outcome for most students. This is hardly cause for optimism, and our educational system is a long way from delivering effective online instruction to all students. We anticipate that disadvantaged students stand to lose the most ground. Therefore, a more appropriate policy response would be to focus on equity-based support systems and resources to aid students and their families in their recovery during and after this global pandemic.
 We attended to this concern in several ways in our study. One way is that we accounted for test score trends at baseline, so that selection into a charter school and subsequent achievement outcomes could not be ascribed to achievement that was already on a downward (or upward) trend. Another way is that we also accounted for student disciplinary incidents at baseline, as these may be a prime signal for a student to opt for a virtual charter school over a traditional schooling environment. We also subjected our results to many alternative model specifications and sensitivity analyses. (Back to top)