In Volume 11 of The Journal of Benefit-Cost Analysis, Thunström, Newbold, Finnoff, Ashworth, and Shogren presented their findings on the national benefits and costs of physical (or social) distancing measures. Their benefit-cost analysis shows a net benefit of $5.6 Trillion to the US economy over 30 years. However, work on the economics of education and family by Hanushek, Boyd, and others suggests that the long-term impacts on present and future productivity of one aspect of physical distancing policies, virtual learning and school shutdowns, may be more severe than this initial model supposes.
There are two forces that, ceteris paribus, would seem to be eroding productivity, and thus diminishing future economic growth: the quality of K-12 education and maternal workforce participation. For both mothers interrupting their careers to meet the educational needs of their children during the pandemic and those children, these forgone earnings represent a loss of future productivity to the economy at large.
A comprehensive assessment of the costs—and benefits—of a virtual K-12 education rather than a face-to-face education in 2020-2021 will take years to fully study and appreciate, but we know that teacher quality, even for just one year, can have profound impact on students’ future earnings. Despite teachers’ unquestioned hard work, it seems reasonable to suppose that a rapid implementation of a completely novel medium of instruction will reduce teacher quality during the transition.
For a back-of-the-envelope calculation of the educational costs associated with virtual learning, let us suppose that the effect on students’ future earnings, is analogous to the difference between having a teacher who is 1.0 standard deviation above average compared to an average teacher. If that supposition is correct, that will mean that students in online classes this year will have $12,263 lower lifetime earnings in net present value (2020 dollars; Hanushek estimated $10,600 per student in 2011 dollars). Multiply this impact across 54 million K-12 students, and we can expect a decline in earnings of over $662 billion, net present value, over the course of the next 30 to 40-odd years.
Second, early evidence indicates that women’s careers are being disproportionately affected by the pandemic generally, but school closures in particular. To the extent that this scenario is analogous to women interrupting careers for motherhood in general, it is reasonable to expect that working mothers’ earnings will also be diminished if and when they return to the workforce after schools reopen. Based on a model for maternal time off for child rearing from the Center for American Progress, an estimate of the effects on lifetime earnings is likely between $100,000 and $150,000 for most women (e.g. a 30-year-old woman working since age 22 and earning $44,000 who leaves the workforce for one year would lose $130,564 in total lifetime income; the model isn’t transparent, but I’m assuming it’s adjusted to net present value).
Estimates of how many women have or will temporarily leave the workforce due to school closures are conflated with the high numbers of layoffs generally. However, in 2018, the Census Bureau reported that there were 23.5 million mothers of children under 18 working full-time; if we assume 1.0% of those mothers left the workforce voluntarily due to school closures, that would conservatively be a decrease of $2.35 billion in lifetime earnings for this group.
Combined, these two effects could decrease the net benefit of lockdowns by approximately $0.665 trillion—roughly 12% of the Thunström estimate—if lockdowns include closing K-12 education. These back-of-the-envelope calculations are intentionally conservative, so it is not surprising that others have found potentially larger costs to school closures in the U.S. and elsewhere. Even more concerning, the long-term rate of GDP growth of 1.75% used by Thunström, et al. is likely overly optimistic for scenarios in which schools are primarily remote or virtual for one or more years because it does not include probable decreases in future productivity and growth, further attenuating the net benefits of lockdowns. This analysis does not address the costs of daycare closures or virtual college learning. Nor does it address the racial and wealth disparities associated with all of these costs, which are undoubtedly substantial.
I am not advocating that we abandon all physical distancing policies. A 12% decline in the net present value of these wages losses is quite impactful, but it does not change the results of the analysis: physical distancing policies have a net positive economic benefit; especially in the medium term. Rather, I am arguing that the policy mix within the umbrella of physical distancing matters. The closure of restaurants and bars (for example) has a very well-appreciated immediate impact on the economy, but the closure of schools, on the other hand, will have much larger long-term effects than is often appreciated. Even a small change in future productivity can and will have very long-lasting costs, indeed.