Table 4.
Mechanism (1) Victim-criminal match, high infection risk, and alcohol sales ban.
| All crime | Domestic violence | Lapsed alimony | Sexual crimes | Femicide | |
|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | |
| Panel A: 1(Large City within the Municipality) | |||||
| 1(Post-COVID-19) x 1(Large City in Municipality) | −2.468*** | −1.516*** | −0.380*** | −0.574*** | 0.001 |
| (0.229) | (0.195) | (0.057) | (0.073) | (0.013) | |
| 1(Post-COVID-19) | −1.350*** | −0.771*** | −0.339*** | −0.224*** | −0.015 |
| (0.202) | (0.172) | (0.050) | (0.064) | (0.011) | |
| Panel B: 1(Above Median Per Capita Deaths, March-October) | |||||
| 1(Post-COVID-19) x 1(High P.C. Deaths) | −2.937*** | −1.777*** | −0.666*** | −0.506*** | 0.011 |
| (0.302) | (0.257) | (0.075) | (0.096) | (0.017) | |
| 1(Post-COVID-19) | −0.469 | −0.253 | −0.023 | −0.169* | −0.024 |
| (0.290) | (0.247) | (0.072) | (0.092) | (0.016) | |
| Panel C: 1(Above Median Per Capita Cases, March-October) | |||||
| 1(Post-COVID-19) x 1(High P.C. Cases) | −3.095*** | −1.785*** | −0.782*** | −0.536*** | 0.008 |
| (0.306) | (0.260) | (0.076) | (0.097) | (0.017) | |
| 1(Post-COVID-19) | −0.322 | −0.238 | 0.079 | −0.141 | −0.022 |
| (0.294) | (0.250) | (0.073) | (0.093) | (0.016) | |
| Panel D: By Municipality-Alcohol Sales Ban (Post-Implementation) | |||||
| 1(Post-COVID-19) x 1(Post-Alcohol Sales Ban) | −2.738*** | −2.078*** | −0.143 | −0.511*** | −0.007 |
| (0.689) | (0.580) | (0.119) | (0.143) | (0.012) | |
| 1(Post-COVID-19) | −2.245*** | −1.219*** | −0.549*** | −0.465*** | −0.013 |
| (0.388) | (0.312) | (0.079) | (0.090) | (0.009) | |
| Observations | 49,140 | 49,140 | 49,140 | 49,140 | 49,140 |
| Pre-Lockdown Mean Dependent | 17.42 | 13.22 | 1.45 | 2.68 | 0.07 |
| Baseline FE | X | X | X | X | X |
| Time Trends | X | X | X | X | X |
Source: Mexico's National Public Security System (Secretariado Ejecutivo del Sistema Nacional de Seguridad Pública). See Appendix B for data sources used for the mechanisms. Notes: Difference-in-differences-differences estimates from Equation (3). Estimates show the grouped post-periods (month three 2020 onward) relative to the pre-pandemic (month three and before). The interacted estimate shows the post-period interacted with the mechanism of interest. Monthly linear time trends are included. Baseline fixed effects are included at the municipality, month, and year. The baseline specification is weighted by the muncipality-level population. Estimation includes January through October over 2019–2020. Robust standard errors are clustered at the municipal level. Significance levels: * , ** , *** Crimes are measured per 100,000 persons. Alternative specification with continuous measures shown in Table B.3.