Table 1.
Descriptive Statistics (Feb 2020 – Jun 2021).
Fig 1.
Monthly intensity of stay-at-home orders by policy type.
Monthly number of days with recommended (black) and mandatory (dark grey) stay-at-home orders across southern U.S. states from February 2020 to June 2021.
Fig 2.
Monthly changes in residential and workplace mobility during the COVID-19 pandemic.
Monthly average percent changes in mobility from February 2020 to June 2021 across southern U.S. counties, relative to the pre-pandemic baseline (January 3–February 6, 2020). Residential mobility is shown as a dashed gray line; workplace mobility as a solid black line.
Fig 3.
County-level mobility changes relative to pre-pandemic baseline.
Panel A shows average workplace mobility changes by county, and Panel B presents residential mobility changes across 1,422 counties in the southern United States. Color gradients represent average percentage change from baseline (January 3–February 6, 2020), with darker blue indicating greater declines in workplace mobility and darker red indicating greater increases in residential mobility. Black-shaded counties indicate missing Google mobility data.
Table 2.
Baseline Estimation Results (OLS Regression).
Table 3.
Estimation results.
Fig 4.
Bayesian-predicted county-level mobility changes from the spatio-temporal model.
Panel A displays predicted workplace mobility, and Panel B shows predicted residential mobility derived from the Bayesian spatio-temporal model specified in Equation (4). Darker purple shades represent larger workplace mobility declines, and deeper red tones represent larger increases in residential mobility. The model incorporates CAR spatial structure, AR(1) temporal dependence, and a space–time interaction term, producing smoothed predictions that account for spatial and temporal dependence while accommodating missing data.