Because most of these local NPIs are adopted in February and our earlier results indicate that the transmission of COVID-19 declines during late January, we restrict the analysis sample to February 2–February 29. We also exclude cities in Hubei province, which modified the case definition related to clinically diagnosed cases on February 12 and changed the case definition related to reduced backlogs from increased capacity of molecular diagnostic tests on February 20. These modifications coincide with the adoption of local NPIs and can significantly affect the observed dynamics of confirmed cases. The adoption of closed management or stay at home is likely affected by the severity of the epidemic and correlated with the unobservables. Additional weather controls that have a good predictive power for these NPIs are selected as the instrumental variables based on the method of Belloni et al. (2016). Details are displayed in Appendix A. The estimation results of OLS and IV regressions are reported in Table 8.
Table 8 Effects of local non-pharmaceutical interventions
(1) (2) (3) (4) (5) (6)
OLS IV OLS IV OLS IV
Average # of new cases, 1-week lag
Own city 0.642*** 0.780*** 0.684*** 0.805*** 0.654*** 0.805***
(0.0644) (0.0432) (0.0496) (0.0324) (0.0566) (0.0439)
× closed management − 0.593*** − 0.244*** − 0.547*** − 0.193*
(0.162) (0.0619) (0.135) (0.111)
× stay at home − 0.597*** − 0.278*** − 0.0688 − 0.110
(0.186) (0.0800) (0.121) (0.143)
Other cities 0.00121 − 0.00159 0.00167 − 0.00108 0.00129 − 0.00142
wt. = inv. dist. (0.000852) (0.00167) (0.00114) (0.00160) (0.000946) (0.00183)
Wuhan 0.00184 0.00382 0.00325* 0.00443 0.00211 0.00418
wt. = inv. dist. (0.00178) (0.00302) (0.00179) (0.00314) (0.00170) (0.00305)
Wuhan 0.00298 0.00110 − 0.00187 − 0.000887 0.00224 − 3.26e-07
wt. = pop. flow (0.00264) (0.00252) (0.00304) (0.00239) (0.00254) (0.00260)
Average # of new cases, 2-week lag
Own city 0.0345 − 0.0701 − 0.0103 − 0.0818 0.0396 − 0.0533
(0.0841) (0.0550) (0.0921) (0.0523) (0.0804) (0.0678)
× closed management − 0.367*** − 0.103 − 0.259** 0.0344
(0.0941) (0.136) (0.111) (0.222)
× stay at home − 0.294*** − 0.102 − 0.124* − 0.162
(0.0839) (0.136) (0.0720) (0.212)
Other cities − 0.00224 − 0.00412** − 0.00190 − 0.00381** − 0.00218 − 0.00397**
wt. = inv. dist. (0.00135) (0.00195) (0.00118) (0.00177) (0.00129) (0.00192)
Wuhan − 0.00512 0.00197 − 0.00445 0.00231 − 0.00483 0.00227
wt. = inv. dist. (0.00353) (0.00367) (0.00328) (0.00348) (0.00340) (0.00376)
Wuhan 0.00585*** 0.00554*** 0.00534*** 0.00523*** 0.00564*** 0.00516***
wt. = pop. flow (0.00110) (0.000929) (0.00112) (0.00104) (0.00109) (0.00116)
Observations 8064 8064 8064 8064 8064 8064
Number of cities 288 288 288 288 288 288
Weather controls Yes Yes Yes Yes Yes Yes
City FE Yes Yes Yes Yes Yes Yes
Date FE Yes Yes Yes Yes Yes Yes
The sample is from February 2 to February 29, excluding cities in Hubei province. The dependent variable is the number of daily new confirmed cases. The instrumental variables include weekly averages of daily maximum temperature, wind speed, precipitation, and the interaction between wind speed and precipitation, in the preceding third and fourth weeks, and the inverse log distance weighted averages of these variables in other cities. Additional instrumental variables are constructed by interacting these excluded instruments with variables that predict the adoption of closed management of communities or family outdoor restrictions (Table 10). The weather controls include weather characteristics in the preceding first and second weeks. Standard errors in parentheses are clustered by provinces. *** p < 0.01, ** p < 0.05, * p < 0.1