Table A7.
Specification | (1a) | (1b) | (2a) | (2b) | (3a) | (3b) | (4a) |
---|---|---|---|---|---|---|---|
p-value | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 |
Specification | (4b) | (5a) | (5b) | (6a) | (6b) | (7a) | (7b) |
p-value | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 | 0.048 |
Notes: Since our data have (almost) no variation in the outcome variable in any region before the vaccination rollout, we use the vaccination share among 18–29 year olds as the main variable to match on. Specifications refer to: (1) all pre-treatment vaccination shares among 18–29 year olds, (2) the first three-fourths of the values, (3) the first half of the values, (4) odd pre-treatment values, (5) even pre-treatment values, (6) pre-treatment mean, and (7) three values. Specifications ending with b includes all additional eight covariates, while specifications ending with a, includes no additional covariates. The post-intervention effect size is largest in Uppsala in each specification, p-value = 1/21 = 0.048.