Table 6.
Social and economic factors mediating the transmission of COVID-19
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Jan 19–Feb 1 | Feb 2–Feb 29 | |||
| IV Coeff. | IV Coeff. | |||
| Average # of new cases, previous 14 days | ||||
| Own city | − 0.251 | 0.672*** | ||
| (0.977) | (0.219) | |||
| × population density | 0.000164 | − 0.000202** | + 495 per km2 | |
| (0.000171) | (8.91e-05) | |||
| × per capita GDP | 0.150*** | − 66, 667 RMB | 0.0102 | |
| (0.0422) | (0.0196) | |||
| × # of doctors | − 0.108* | + 92, 593 | 0.0179 | |
| (0.0622) | (0.0236) | |||
| × temperature | 0.0849* | − 11.78∘C | − 0.00945 | |
| (0.0438) | (0.0126) | |||
| × wind speed | − 0.109 | 0.128 | ||
| (0.131) | (0.114) | |||
| × precipitation | 0.965* | − 1.04 mm | 0.433* | − 2.31 mm |
| (0.555) | (0.229) | |||
| × adverse weather | 0.0846 | − 0.614*** | + 163% | |
| (0.801) | (0.208) | |||
| Other cities | 0.0356 | − 0.00429 | ||
| wt. = inv. distance | (0.0375) | (0.00343) | ||
| Other cities | 0.00222 | 0.000192 | ||
| wt. = inv. density ratio | (0.00147) | (0.000891) | ||
| Other cities | 0.00232 | 0.00107 | ||
| wt. = inv. per capita GDP ratio | (0.00497) | (0.00165) | ||
| Wuhan | − 0.165 | − 0.00377 | ||
| wt. = inv. distance | (0.150) | (0.00981) | ||
| Wuhan | − 0.00336 | − 0.000849 | ||
| wt. = inv. density ratio | (0.00435) | (0.00111) | ||
| Wuhan | − 0.440 | − 0.0696 | ||
| wt. = inv. per capita GDP ratio | (0.318) | (0.0699) | ||
| Wuhan | 0.00729*** | 0.0125*** | ||
| wt. = population flow | (0.00202) | (0.00187) | ||
| Observations | 4032 | 8064 | ||
| Number of cities | 288 | 288 | ||
| Weather controls | Yes | Yes | ||
| City FE | Yes | Yes | ||
| Date FE | Yes | Yes | ||
The dependent variable is the number of daily new confirmed cases. The sample excludes cities in Hubei province. Columns (2) and (4) report the changes in the mediating variables that are needed to reduce the impact of new confirmed cases in the preceding 2 weeks by 1, using estimates with significance levels of at least 0.1 in columns (1) and (3), respectively. The endogenous variables include the average numbers of new cases in the own city and nearby cities in the preceding 14 days and their interactions with the mediating variables. Weekly averages of daily maximum temperature, precipitation, wind speed, the interaction between precipitation and wind speed, and the inverse log distance weighted sum of these variables in neighboring cities, during the preceding third and fourth weeks, are used as instrumental variables in the IV regressions. Additional instrumental variables are constructed by interacting them with the mediating variables. Weather controls include these variables in the preceding first and second weeks. Standard errors in parentheses are clustered by provinces
*** p < 0.01, ** p < 0.05, * p < 0.1