Table 1.
Estimates of the effect of turnout on excess hospitalizations
| Coefficient (standard error) [p-value] | ||
|---|---|---|
| βt, effect of turnout for | γt, effect of turnout for | |
| departments with low | departments with high | |
| COVID-19 circulation by March 15 | COVID-19 circulation by March 15 | |
| March 25 | − 4.579 | 7.950 |
| (2.287) | (5.343) | |
| [0.054] | [0.147] | |
| March 26 | − 2.646 | 11.344 |
| (4.085) | (5.725) | |
| [0.522] | [0.056] | |
| March 27 | − 7.620 | 23.902 |
| (6.129) | (5.929) | |
| [0.223] | [0.000] | |
| March 28 | − 7.464 | 39.455 |
| (6.146) | (7.934) | |
| [0.233] | [0.000] | |
| March 29 | − 9.277 | 48.023 |
| (5.746) | (12.222) | |
| [0.116] | [0.000] | |
| March 30 | − 3.345 | 53.616 |
| (5.775) | (16.120) | |
| [0.566] | [0.002] | |
| March 31 | − 5.869 | 59.461 |
| (7.276) | (19.971) | |
| [0.426] | [0.006] | |
| April 1 | − 8.491 | 66.684 |
| (8.653) | (21.873) | |
| [0.334] | [0.005] | |
| April 2 | − 11.164 | 70.724 |
| (9.767) | (23.325) | |
| [0.261] | [0.005] | |
Estimates of βt and γt from Eq. 2 from March 25 onwards. See Section 3 for more details. See Fig. 2 for a graphical representation. P-values of two-sided tests in brackets. Standard errors clustered at the day and department levels between parentheses. The sample is made of 3,003 observations (91 departments × 33 days). Départements with high (low) COVID-19 circulation by March 15 are departments in the top two thirds (bottom third) of the distribution of cumulated hospitalizations for COVID-19 suspicion by March 26