Table 2 Results of the sensitivity analyses for the estimation of the total effect for both responses r and growth.new.cases. For the FE, DFE and double machine learning approach, the \(95\%\)-confidence intervals (\(CI_{95\%}\)) are constructed via two-way clustering on the canton and week (Canton-Week). For the RE approach, we compute the standard errors as described in Sect. 4.2. For the sake of comparison, we add the main results of the main text at the top of the table
| Â | r | growth.new.cases | ||
|---|---|---|---|---|
Point estimate | Confidence interval | Point estimate | Confidence interval | |
Main results | ||||
 FE | \(-0.16\) | \([-0.28, -0.05]\) | \(-0.17\) | \([-0.35, 0.00]\) |
 DFE | \(-0.17\) | \([-0.28, -0.05]\) | \(-0.19\) | \([-0.36, -0.01]\) |
 RE | \(-0.22\) | \([-0.32,-0.12]\) | \(-0.29\) | \([-0.48, -0.10]\) |
E.1 Alterations to confidence interval construction | ||||
 Month FE | \(-0.16\) | \([-0.23,-0.10]\) | \(-0.17\) | \([-0.31, -0.04]\) |
 Canton-Month FE | \(-0.16\) | \([-0.27,-0.06]\) | \(-0.17\) | \([-0.25,-0.09]\) |
 Month DFE | \(-0.17\) | \([-0.23,-0.10]\) | \(-0.19\) | \([-0.33,-0.06]\) |
 Canton-Month DFE | \(-0.17\) | \([-0.24,-0.09]\) | \(-0.19\) | \([-0.36,-0.02]\) |
E.2 Alterations to the data and point estimation | ||||
Additional information variables | ||||
 RE |     - |          - | \(-0.22\) | \([-0.40,-0.05]\) |
Half-Cantons | ||||
  FE | \(-0.17\) | \([-0.27,-0.07]\) | \(-0.12\) | \([-0.33,0.10]\) |
 DFE | \(-0.26\) | \([-0.36,-0.15]\) | \(-0.22\) | \([-0.43,-0.01]\) |
 RE | \(-0.21\) | \([-0.34,-0.09]\) | \(-0.25\) | \([-0.46,-0.03]\) |
Timing of information variables | ||||
 FE | \(-0.19\) | \([-0.28,-0.11]\) | \(-0.10\) | \([-0.29,0.10]\) |
 DFE | \(-0.27\) | \([-0.36,-0.19]\) | \(-0.15\) | \([-0.35,0.04]\) |
 RE | \(-0.28\) | \([-0.45,-0.10]\) | \(-0.16\) | \([-0.28,-0.03]\) |
Outliers | ||||
 FE | \(-0.08\) | \([-0.15,-0.01]\) | \(-0.21\) | \([-0.40,-0.03]\) |
Very short sample period | ||||
 FE | \(-0.24\) | \([-0.52,0.04]\) | \(-0.20\) | \([-0.44,0.05]\) |
Short sample period | ||||
 FE | \(-0.16\) | \([-0.27,-0.05]\) | \(-0.16\) | \([-0.43,0.12]\) |
 DFE | \(-0.11\) | \([-0.21,0.00]\) | \(-0.08\) | \([-0.22,0.07]\) |
 RE | \(-0.12\) | \([-0.20,-0.03]\) | \(-0.04\) | \([-0.18,0.11]\) |
Double machine learning | ||||
 DML | \(-0.70\) | \([-1.36,-0.04]\) | \(-1.00\) | \([-2.10,0.11]\) |
Lag-1 response variable as covariate | ||||
 FE | \(-0.12\) | \([-0.21,-0.03]\) | \(-0.26\) | \([-0.47,-0.04]\) |
 DFE | \(-0.12\) | \([-0.21,-0.03]\) | \(-0.27\) | \([-0.48,-0.05]\) |
 RE | \(-0.22\) | \([-0.32,-0.12]\) | \(-0.44\) | \([-0.65,-0.22]\) |
 DML | \(-0.71\) | \([-1.38,-0.04]\) | \(-1.08\) | \([-2.25,0.08]\) |