A new study that examined anonymous cell phone tracking data shows that shelter-in-place orders worked better in some regions of the United States than others. The findings, which appear in PLOS One, suggest that political partisanship and other factors played an important role.
The authors of the study were interested in learning more about why responses to public health interventions meant to slow the spread of COVID-19 varied from state to state.
“Responding to the COVID-19 pandemic was urgent. States took different paths to trying to deal with it. Some policies were evidence-based and others weren’t. If we could figure out which policies worked and in what settings, we could contribute an important piece of missing data at a critical time,” said study author Yevgeniy Feyman of the Boston University School of Public Health.
For their study, Feyman and his colleagues analyzed county-level changes in population mobility using data from Google Community Mobility Reports for February 15, 2020 through April 21, 2020.
The researchers found mobility decreased nationwide in the wake of shelter-in-place orders, but the effectiveness of orders varied widely. They were most effective in states with higher population densities, lower vote shares for Donald J. Trump in 2016, and higher per capita incomes.
“Shelter-in-place orders worked: they kept people at home. These policies were especially helpful in more urban areas, and when they came with other policies like school closures that make it easier for folks to comply with the orders,” Feyman told PsyPost. “Our study also underscores how politicized these laws have become. Areas with more Trump voters had smaller responses to these orders.”
Shelter-in-place orders also appeared to be more effective in states with lower Black share of population, but this “may be partly be a proxy for the concentration of essential workers,” the researchers said.
Feyman and his colleagues also observed a reduction in population mobility before shelter-in-place orders were even issued. The first case of COVID-19 in the United States was confirmed on January 21, 2020. But most states implemented shelter-in-place orders between March 19 and April 7, 2020.
“Another thing our study allowed us to do is distinguish between mobility reductions that were already in process (due to fear) vs. true effects of shelter-in-place orders. We see that household mobility fell by an average of 26% in the days before these orders became effective,” Feyman explained.
“The orders themselves were only responsible for another 12% reduction, and may have just accelerated this pre-existing trend. This is why claims that shelter-in-place orders ruin the economy are bogus. Fear ruins the economy. There was tremendous uncertainty in these early days, especially due to insufficient testing and contact tracing. If people don’t feel safe, they won’t fully resume normal activities even if they are legally allowed to do so.”
But it is unclear whether the shelter-in-place orders became less effective as the pandemic dragged on. “It’s possible that the effects of shelter-in-place orders are temporary and wear off over time,” Feyman said. “Additionally, our study doesn’t explain why states that enacted similar COVID-19 policies had very different outcomes; there’s more work to be done.”
The study, “Effectiveness of COVID-19 shelter-in-place orders varied by state“, was authored by Yevgeniy Feyman, Jacob Bor, Julia Raifman, and Kevin N. Griffith.