Skip to content Skip to main navigation Report an accessibility issue

Studying the Impact of Social Distancing on COVID-19 Morbidity

Nicholas Nagle, associate professor of geography, and Liem Tran, professor of geography, are part of the UT Coronavirus-19 Outbreak Response Experts (CORE-19) studying the impact of social distancing on Covid-19 morbidity in the early months (up to May 31) of the pandemic.

Using county-level COVID-19 data and social distancing metrics from tracked mobile devices, they investigated how social distancing influences the total number of confirmed COVID-19 cases in a county between when the state and the county first enacted social distancing measures through May 31, 2020.

They created a mixed-effects negative binomial model to assess the association between social distancing and the change in the total number of confirmed COVID-19 cases/100,000 people while controlling for covariates that might introduce bias to this relationship. Marginal effects at the means were also generated to further isolate the individual influence of social distancing on COVID-19 from other factors.

A 1% decrease in percentage of mobile devices leaving home between March-May 2020 corresponded with 5.8 fewer total confirmed COVID-19 cases/100,000 people in a county (95% confidence interval [CI]: 0.46-72.6) from the mixed-effects negative binomial while marginal effects revealed that a 1% decrease in percentage of mobile devices leaving home between March-May 2020 led to 14.5 fewer total confirmed COVID-19 cases/100,000 people in a county (95% CI: 7.82, 21.16).

Social distancing plays a key role in keeping the number of COVID-19 cases low in counties across the US and should be encouraged until a COVID-19 vaccine becomes readily available.

The Howard H. Baker Jr. Center for Public Policy published a policy brief, Understanding COVID-19 Models, in partnership with the Coronavirus-19 Outbreak Response Experts (CORE-19), July 1, 2020.

Daily outputs are used in the Tennessee State Data Center’s Covid-19 dashboard.