Both came out today via the National Bureau of Economic Research, and they’re available here and here. The upshot is that lockdowns reduce the spread of COVID-19, though the studies have their limits.

The first paper looks at two different data sets, one of countries and one of U.S. states. In the country-level analysis, the authors are able to separate out various components of lockdowns to estimate which policies had which effects. (As I’ve said before, I’m skeptical of this kind of thing since the policies rolled out so quickly everywhere.)

Here are the main results; in reading the following, note that the numbers can be roughly interpreted as percentages (e.g., 0.05 would correspond to about 5 percent):

You Might Like

The policies most strongly and statistically significantly associated with slowing the [daily] growth rate of . . . confirmed cases in order of magnitude of impact were public transport closures (-0.09 . . .), enforced workplace closures (-0.0784 . . .), limited domestic travel ([-0.065] . . .), and restrictions on international travel (-0.0639 . . .). School closures . . . and limits on public events . . . are negatively related to growth rates of confirmed cases but were not found to be statistically significant.

Note, however, that this measures confirmed cases, which depend heavily on testing. The results for deaths are generally not statistically significant, perhaps because of “our short sample and long lags between implementation . . . and effects on death rate.”

For U.S. states, the paper focuses on stay-at-home orders, so it doesn’t distinguish among the various specific restrictions they entail. Once again, the results are strong for case growth — “a state’s own policy was associated with a reduction of the growth rate of 16.9 log points [i.e., in the ballpark of 17 percent]” — but not deaths. The paper further finds “spillover” effects among states, where a state’s own trends depend, in large part, on what other states are doing.

This paper also looks at effects on the economy. Interestingly, it finds that the economic damage has been mostly national in scope, not limited to states that enacted stay-at-home orders — though it is predictably concentrated among industries that were shut down or whose workers can’t telecommute.

On to the next paper, whose main analysis is a similar dive into case and death growth in U.S. states — but which adds that some states benefit more than others from locking down. Here’s part of the abstract, which sums things up nicely:

This study explores the impact of SIPOs [shelter-in-place orders] on health, with particular attention to heterogeneity in their impacts. First, using daily state-level social distancing data from SafeGraph . . . we document that adoption of a SIPO was associated with a 5 to 10 percent increase in the rate at which state residents remained in their homes full-time. Then, using daily state-level coronavirus case data collected by the Centers for Disease Control and Prevention, we find that approximately three weeks following the adoption of a SIPO, cumulative COVID-19 cases fell by 44 percent. Event-study analyses confirm common COVID-19 case trends in the week prior to SIPO adoption and show that SIPO-induced case reductions grew larger over time. However, this average effect masks important heterogeneity across states — early adopters and high population density states appear to reap larger benefits from their SIPOs. Finally, we find that statewide SIPOs were associated with a reduction in coronavirus-related deaths, but estimated mortality effects were imprecisely estimated.

The findings here are broadly similar; the results for case growth are strong, and while these authors do find a result for deaths, they concede it’s “imprecisely estimated.” And the addition that places with high population density benefit more from shutdowns is important.

As I wrote the other day, the key question now isn’t so much whether lockdowns work but what we do next, because this can’t go on much longer whether it’s working or not. But for future reference, we need to figure out how effective these policies are, and it’s nice to see some early work on that question.