Social distancing causally impacts the spread of SARS-CoV-2: a U.S. nationwide event study

BMC Infectious Diseases

Table 3 Summary statistics for covariates

	Unmatched			Propensity score matched
	Non-protest	Protest	Difference	Non-protest	Protest	Difference
Males	0.501	0.495	0.006***	0.495	0.497	− 0.001
Asian	0.011	0.038	− 0.027***	0.024	0.026	− 0.002
Black	0.099	0.109	− 0.01	0.088	0.096	− 0.008
Hispanic	0.085	0.125	− 0.040***	0.111	0.113	− 0.002
White	0.850	0.819	0.031***	0.850	0.839	0.01
60-years+	0.261	0.233	0.029***	0.242	0.240	0.002
Diabetes prevalence	0.108	0.099	0.010***	0.099	0.101	− 0.002
Hypertension prevalence	0.401	0.377	0.024***	0.378	0.383	− 0.005**
Obesity prevalence	0.389	0.354	0.035***	0.364	0.366	− 0.003
Smoking prevalence	0.245	0.208	0.036***	0.217	0.220	− 0.003
ln(Population density)	3.784	5.378	− 1.593***	4.808	4.763	0.045
ln(Per Capita RGDP)	10.480	10.789	− 0.309***	10.659	10.683	− 0.023
Social distancing restrictions	4.891	5.050	− 0.159*	5.006	4.896	0.11
Social mobility	1.552	0.675	0.877***	0.979	1.024	− 0.045
Number of counties	2077	541		356	356

This table reports the mean value of the covariates described in “Covariates” section, for non-protest versus protest counties for all the counties included in our sample (unmatched), in the first panel, and for the propensity score matched counties, in the second panel. ***, **, and * denote the statistical significance of t-statistics from tests of the null hypothesis that the difference between the means for non-protest and protest counties is statistically indistinguishable from zero at the 1%, 5%, and 10% level, respectively

ISSN: 1471-2334