Skip to main content
. 2023 Jul 19;9(29):eadg9434. doi: 10.1126/sciadv.adg9434

Table 2. Measuring vaccines and ads in rates.

This table displays estimates from modifications of Eqs. 1 and 2 where the dependent variable is the total percent of the county population vaccinated at a given point in time and the treatment intensity is measured as the number of ads a county receives per 100 residents. The number of observations is slightly higher here than in our main analysis (163,856 county-date observations rather than 151,945) because, for some observations, the vaccination count is missing on certain dates in the CDC data, although the vaccination rate is recorded. “***,” “**,” and “*” indicate significance (from a one-tailed test) at the 0.01, 0.05, and 0.10 levels. Standard errors, reported in parentheses below each estimate, are clustered at the county level. Randomization inference P values are from a one-tailed test based on 1000 permutations using the treatment effect as the randomization test statistic. Panel (A) reports results from an unweighted regression, and panel (B) shows results where observations are weighted by county population. Table S5 contains estimates of other coefficients from these regressions.

ITT effect ACR of 1000 ads
(1) (2) (3) (4)
A. Unweighted regression
Effect 0.570*
(0.437)
0.563*
(0.437)
0.0296*
(0.0227)
0.0291*
(0.0226)
Randomization inference P value 0.089 0.093
B. Weighted regression
Effect 0.448
(0.458)
0.411
(0.454)
0.0191
(0.0196)
0.0173
(0.0192)
Randomization inference P value 0.159 0.184
County fixed effects Yes Yes Yes Yes
Date fixed effects Yes Yes Yes Yes