Table 3.
The relationship between long- and short-term exposure to pollution and the probability on dying from COVID-19 in Mexico City.
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| PM2.5 (μg/m3) | 0.0566⁎⁎ | 0.0484⁎⁎ | |
| [0.0227] | [0.0234] | ||
| PM2.5–2019 (μg/m3) | 0.0260⁎ | 0.0170 | |
| [0.0149] | [0.0132] | ||
| PM2.5–14 (μg/m3) | 0.0038⁎ | 0.0035⁎ | 0.0037⁎ |
| [0.0021] | [0.0020] | [0.0021] | |
| Municipal-level covariates included | Yes | Yes | Yes |
| Individual-level covariates included | Yes | Yes | Yes |
| N | 71,620 | 71,620 | 71,620 |
| Municipalities | 14 | 14 | 14 |
| Localities | 380 | 380 | 380 |
| Pseudo R2 | 0.260 | 0.260 | 0.260 |
Notes: The dependent variable is a dummy equal to one if an individual diagnosed with COVID-19 dies and zero otherwise. Estimations are done using a probit model. Cluster robust standard errors at the municipal-level shown in brackets. Municipal-level covariates are: Population density, population, density of hospital beds, percentage of population without access to health care, percentage of population with moderate or severe food insecurity, and percentage of labor force with jobs that can be done from home. Individual-level covariates are: gender, age and age squared, obesity, diabetes, hypertension, smoking status, and day in which symptoms started. The Mexico City sample only considers the 14 municipalities for which, in addition to long-term exposure to PM2.5, we know PM2.5 for 2019 and 2020.
p < 0.1.
p < 0.05.