5. Cross-Sectional Analyses

5.1. Research Design
To test the H2a and H2b, we generate the regression model as follows:CAR = β0 + β1CIPHT + β2CIPHT × Conditioning_VAR + β3Conditioning_VAR + β4PRO_CASE + β5SIZE + β6ROA + β7CURR + β8R&D + β9LOSS + β10LEV  + β11OPCF + β12TURN + β13CEO_AGE+ β14CEO_COM + β15CEO_TEN + β16CEO_DUA + Week FE + Industry FE + Province FE + ε(3)  where Conditioning_VAR is a conditioning variable that moderates the association between continued increasing public health threats and accumulative abnormal return. All other variables are above-mentioned. To test H2a and H2b, Conditioning_VAR is in terms of the provincial information accessibility and provincial economic growth, respectively. We explain the detail proxies in the following sections.

5.2. The Conditioning Effect of Provincial Information Accessibility―Test of H2a
Regrading H2a, we investigate whether the continued increasing public health threats in decreasing the accumulative abnormal return is weaker in firms that located in the provinces with stronger information accessibility. We suppose that stronger information accessibility will decrease the investors’ risk assessment by mitigating the information asymmetry. We apply three proxies (High_WEB, High_MED, and High_MOB) to represent stronger provincial information accessibility. Here, High_WEB is an indicator variable that equals one if the provincial ratio of the number of websites per 100 enterprises to resident population is higher than or equal to the upper quartile value and zero otherwise; High_MED is an indicator variable that equals one if the provincial TV coverage rate of population is higher than or equal to the upper quartile value and zero otherwise; and High_MOB is an indicator variable that equals one if the provincial ratio of flow accessed to mobile internet to resident population is higher than or equal to the upper quartile value and zero otherwise. For the H2a, we substitute Conditioning_VAR in Model (3) with High_WEB, High_MED, and High_MOB, respectively, and expect the coefficient of the interaction term is positive. Table 8 shows the regression results on H2a. We find that the interaction terms of CIPHT × High_WEB, CIPHT × High_MED, and CIPHT × High_MOB are positive and significant, which support the H2a that the negative effect of continued increasing provincial public health threats on market reaction is less pronounced when the provincial information accessibility is stronger (in terms of higher websites rate, higher media coverage, and higher mobile internet rate).

5.3. The Conditioning Effect of Provincial Economic Growth―Test of H2b
Regrading H2b, we investigate whether the continued increasing public health threats in decreasing the accumulative abnormal return is weaker in firms located in the provinces with stronger economic growth. We suppose that stronger economic growth will decrease the investors’ risk assessment by enhancing the likelihood to have a positive outlook on the future economy. We apply three proxies (High_GRP, High_EMR, and High_URB) to represent a stronger provincial economic growth. Here, High_GRP is an indicator variable that equals one if the provincial ratio of the gross regional product to resident population is higher than or equal to the upper quartile value and zero otherwise; High_EMR is an indicator variable that equals one if the provincial employment rate in the urban area is higher than or equal to the upper quartile value and zero otherwise; and High_URB is an indicator variable that equals one if the provincial ratio of urban population to resident population is higher than or equal to the upper quartile value and zero otherwise. For the H2b, we substitute Conditioning_VAR in Model (3) with High_GRP, High_EMR, and High_URB, respectively, and expect the coefficient of the interaction term is positive. Table 9 shows the regression results on H2b. We find that the interaction terms of CIPHT × High_GRP, CIPHT × High_EMR, and CIPHT × High_URB are positive and significant, which support the H2b that the negative effect of continued increasing provincial public health threats on market reaction is less pronounced when the provincial economic growth is stronger (in terms of higher gross regional product rate, higher employment rate, and higher urbanization).