Does opening K-12 schools and colleges increase the transmission of SARS-CoV-2? Do mitigation strategies such as mask-wearing requirements reduce the spread of COVID-19 at school? These are hotly debated policy-relevant questions.
The existing empirical studies provide mixed evidence on the effect of school openings on COVID-19 diffusion. Using a natural experiment in Sicily, Italy, Amodio et al. (2021) find school openings tend to increase the number of COVID-19 cases. Bignami et al. (2021) provide evidence that school openings are associated with increased cases in Montreal neighbourhoods. Auger et al. (2020) use US state-level data to argue that school closures at the start of the pandemic substantially reduced cases, while in Chernozhukov et al. (2021a) we point out the difficulty of identifying the effect of school closures in the US given the lack of cross-sectional variations in the timing of school closures across US states. Goldhaber et al. (2021) find that in-person schooling is only associated with increased cases in areas with high pre-existing COVID-19 cases in counties in Washington and Michigan. Similarly, Harris et al. (2021) find that in-person schooling is not associated with increased hospitalisations in counties with low pre-existing COVID-19 hospitalisation rates using the US county data.
To quantitatively examine the role of school openings and mitigation strategies at school in determining the community transmission of SARS-CoV-2, in Chernozhukov et al. (2021b) we use the county-level panel data in the US on foot traffic and K-12 school opening plans to analyse how an increase in visits to schools and opening schools with different teaching modes (in-person, hybrid, and remote) and staff mask-wearing requirement is related to the 2-weeks forward growth rate of confirmed COVID-19 cases.
As shown in Figure 1(a), in the US county-level panel data from August to December of 2020, the average number of weekly cases starts increasing after 2 weeks of opening K-12 schools in-person or hybrid, especially for counties with no mask mandates for staff. Similarly, in Figure 1(b), the number of deaths starts rising after 3 to 5 weeks of opening K-12 schools for counties with in-person/hybrid teaching methods but requiring no mask-wearing among staff. These figures suggest mask mandates at school may be important for reducing the transmissions of SARS-CoV-2. In Figure 1(c), opening K-12 schools in-person or hybrid increases the number of per-device visits to K-12 schools more than opening remotely, especially when no mask mandates are in place. Figure 1(d) suggests that the opening of schools allows parents to return to work.
Figure 1 The evolution of cases, deaths, and visits to K-12 schools and restaurants before and after the opening of K-12 schools
Figure 2 presents the evolution of the number of cases by age groups, the number of visits to colleges and universities, and the number of visits to bars and restaurants in Dane county, WI, and Lane County, OR, where the University of Wisconsin-Madison and the University of Oregon are located. The number of cases for age groups of 10-29 sharply increased when these two universities start their fall semester with in-person classes. This sharp increase in cases among the 10-29 age cohort is associated with an increase in visits to colleges/universities, bars, and restaurants with a lag of two weeks. When the fall semester began in these two universities, many students started living together in residential halls and likely visited bars and restaurants, where properly wearing masks and practicing social distancing are difficult.
Figure 2 The number of cases by age groups and the number of visits to colleges/universities and bars in Dane county, WI, and Lane county, OR
The observations in Figures 1 and 2 may be driven by a variety of confounders. Therefore, based on the framework developed in Chernozhukov et al. (2021a), we analyse the association of opening K-12 schools and colleges with the spread of COVID-19 by debiased panel data regression with a set of county dummies and interactions of state and week dummies, to control for unobserved time-invariant county-level factors as well as unobserved time-varying state-level factors. We also include county-level NPIs (mask mandates, banning gathering of more than 50 persons, stay-at-home orders) lagged by two weeks, as well as the logarithm of past weekly cases with 2-, 3-, and 4-week lag lengths to capture people’s voluntary behavioural responses to new information of transmission risks. Although we use county-level panel data, we compute clustered standard errors at the state level to provide valid inference under possible dependency not only over time but also across counties within each state.
The main finding from our panel regression analysis is that the growth rates of confirmed cases increase when visits to colleges and K-12 schools increase. Specifically, the estimates indicate that fully opening K-12 schools with in-person learning is associated with a 4.7% (SE = 1.7) increase in the growth rate of cases. We also find that the positive association of K-12 school visits or in-person school openings with case growth is stronger for counties that do not require staff to wear masks at schools than for those without mask-wearing requirement by an additional 4.1% (SE = 1.9). These results have a causal interpretation in a structural model with unobserved county and time confounders.
Sensitivity analysis shows that our baseline results are robust to timing assumptions and alternative specifications. We also provide sensitivity analysis for changes to our regression specification and assumption about delays between infection and reporting. Our findings are robust to alternative time lags of 10 or 18 days, to a different way of computing case growth rates, and to adding a variety of additional controls, including past visits to restaurants, bars, recreational places and churches, visits to full-time and part-time workplaces, and a proportion of devices staying at home.
We also examine how an increase in visits to K-12 schools is related to the proportion of devices at full-time workplaces and that of staying-home devices by panel regression. The results show that opening K-12 school allows parents to return to work and spend more time outside. This suggests a possibility that an increase in the mobility of parents driven by school openings may be associated with an increase in case growth. On the other hand, analysing the relationship between college visits and bar visits, we find that bar visits are positively associated with college visits. This finding is consistent with a hypothesis that the transmission of SARS-CoV-2 due to college openings may be partly driven by an increase in visits to bars by students (Chang et al. 2021).
Our study is observational and therefore should be interpreted with great caution. With this caveat in mind, our analysis suggests that opening K-12 schools and colleges may increase the spread of COVID-19, especially when strict mitigation strategies are not enforced at schools. Because of accumulating evidence for the association of school openings with the transmission of SARS-CoV-2, the precautionary principle calls for local governments to promote enforcing mitigation measures at schools (universal and proper masking, social distancing, and handwashing) to lower the risk of COVID-19 spread. Furthermore, the government can prioritise vaccines for education workers in case of in-person school openings.
Amodio, E, M Battisti, A Kourtellos, G Maggio, and C M Maida (2020), “Schools opening and COVID-19 diffusion: evidence from geolocalized microdata”, COVID Economics 65.
Bignami, S, Y Boujija, J Sandberg, and O Drouin (2021), “Enfants, ecoles et COVID-19 : le cas montrealais”, working paper.
Chang, S, E Pierson, P W Koh, J Gerardin, B Redbird, D Grusky, and J Leskovec (2021), “Mobility network models of COVID-19 explain inequities and inform reopening”, Nature 589(7840): 82–87.
Chernozhukov, V, H Kasahara, and P Schrimpf (2021a), “Causal impact of masks, policies, behavior on early COVID-19 pandemic in the US”, Journal of Econometrics 220(1): 23–62.
Chernozhukov, V, H Kasahara, and P Schrimpf (2021b), “The Association of Opening K-12 Schools and Colleges with the Spread of COVID-19 in the United States: County-Level Panel Data Analysis”, COVID Economics 70.
Goldhaber, D, S A Imberman, K O Strunk, B Hopkins, N Brown, E Harbatkin, and T Kilbride (2021), “To What Extent Does In-Person Schooling Contribute to the Spread of COVID- 19? Evidence from Michigan and Washington”, NBER Working Paper 28455.
Harris, D N, E Ziedan, and S Hassig (2021), “The Effects of School Reopenings on COVID-19 Hospitalizations”, Technical report.
Hsiang, S, D Allen, S Annan-Phan, K Bell et al. (2020), “The effect of large-scale anti-contagion policies on the COVID-19 pandemic”, Nature 584: 262–285.