Table 2.
Regressions results for the effect of one month localized lockdown on total VAT collection, estimated with two-way fixed effects at the municipality level, January-May 2020
| Dependent variable: VAT growth | Baseline | Excluding rural units | Excluding Greater Santiago | Conurbations | Conurbations excluding Greater Santiago | Conurbations and standalone municipalities | As in (1) controlling for deaths | As in (1) controlling for cases |
|---|---|---|---|---|---|---|---|---|
|
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
(7) |
(8) |
|
| Lockdown |
-0.125§ |
-0.132§ |
-0.162† |
-0.161‡ |
-0.153 |
-0.230§ |
-0.125† |
-0.135§ |
| (0.048) |
(0.049) |
(0.096) |
(0.064) |
(0.130) |
(0.059) |
(0.071) |
(0.052) |
|
| Standalone × lockdown |
-0.059 |
|||||||
| (0.104) |
||||||||
| Standalone |
-0.005 |
|||||||
| (0.038) |
||||||||
| New deaths per 100 000 |
0.00004 |
|||||||
| (0.002) |
||||||||
| New log cases per 100 000 |
0.002 |
|||||||
| (0.005) |
||||||||
| Observations |
850 |
785 |
570 |
360 |
195 |
455 |
850 |
850 |
| Adjusted R2 |
0.352 |
0.369 |
0.356 |
0.352 |
0.278 |
0.171 |
0.351 |
0.351 |
| Time effect |
YES |
YES |
YES |
YES |
YES |
YES |
YES |
YES |
| Municipalities | 170 | 157 | 114 | 72 | 39 | 91 | 170 | 170 |
VAT – value-added tax
*All specifications have both geography and time fixed effects. Robust standard errors in parenthesis. Column (1) shows regression results for the baseline sample (ie, municipalities over the 50th percentile of total 2018 VAT). Column (2) excludes units, with over fifty percent of the rural population. Column (3) is the same as (2) but excluding municipalities in Greater Santiago, the capital. Columns (4) includes only municipalities that are part of a large conurbation. Column (5) is the same as column (4), excluding municipalities in Greater Santiago. Column (6) includes municipalities that are part of large conurbations and standalone municipalities (Angol, Antofagasta, Arica, Aysén, Calama, Castro, Chañaral, Colina, Copiapó, Curicó, Osorno, Ovalle, Puerto Montt, Puerto Natales, Punta Arenas, Valdivia, and Vallenar). Columns (7) and (8) consider the baseline sample and controls for contemporaneous COVID-19 new deaths and lagged incidence per 100 000 population, respectively, at the municipality level.
†P < 0.10.
‡P < 0.05.
§P < 0.01.