No disease link---there might be more cases, but democracies prevent deaths
Karabulut et al. 21 — Gokhan Karabulut is Prof. Dr. Department of Economics, Istanbul University. Klaus Felix Zimmermann is a German economist and emeritus professor of economics at Bonn University. Mehmet Huseyin Bilgin is professor of economics at the Department of International Relations at Istanbul Medeniyet University. Gokhan Karabulut, Klaus F. Zimmermann, Mehmet Huseyin Bilgin, Asli Cansin Doker, Democracy and COVID-19 outcomes, Economics Letters, Volume 203, 2021, 109840, ISSN 0165-1765, https://doi.org/10.1016/j.econlet.2021.109840. (https://www.sciencedirect.com/science/article/pii/S0165176521001178) WMK
3. Results In the Table 1, Panel A reports the baseline regressions includ- ing only the democracy indicators for the full sample (N = 128), and Panel B presents the estimates with all control variables (N = 99). The baseline equations show that democracy is positively re- lated to infections at the 1% significance level. When control variables are added, coefficients of the democracy variables con- tinue to be statistically significant at the 1% level except for column 3 where the Civil Liberties Index is used. In panel B, both temperature and population share of 65 and older variables’ coefficients are negative and significant for all five equations. Li et al. (2020), reach similar results for the temperature variable and (Zimmermann et al., 2020), for the older population group variable. Haischer et al. (2020) argue that people who are 65 and older are more likely to wear a mask, and thus the government’s stringency policies are mainly targeted towards this group. Using data from an earlier stage of the pandemic, Zimmermann et al. (2020) noted that the older age group has standard activities that make it less exposed to the virus. Both lines of argument would explain why the possibility of infection is lower for the older group compared to the younger population. Testing policy and doctor per 1000 variables are positively related to CP (ex- cept column 3 for testing policy), which is intuitive and confirm expectations. Finally, the Government Censorship Effort variable is negatively related to CP in general and significant in column 2 (Panel B). This indicates a weak tendency where more media control leads to higher infection rates since public attention to the disease is possibly smaller. The results of the regressions for the Case Fatality Rate are quite different. There is a negative relationship between all democracy measures and CFR. All coefficients are statistically significant at the 1% level except for column 5 where it is 10%. We observe COVID-19 to have a smaller effect on mortality for more democratic countries. Government Censorship has the largest coefficient at the 1% significance level and the sign of the coefficient is positive. This implies that a lower degree of censorship is associated with a larger case fatality rate. Beds per 1000 population and testing policy both have a negative relationship with the Case Fatality Rate. Therefore, more hospital beds and more tests may help to decrease CFR. GDP is also negatively related to CFR. This result is consistent with the results of previous studies (Liu et al., 2020; Zimmermann et al., 2020). On the other hand, the share of the population over the age of 65 is positively related to CFR at 1%; this means that once elderly people get the disease, they are more likely to die (Zimmermann et al., 2020). Results remain robust (see Online Appendix Table A5) for more detailed controls for the age distribution of the populations (estimates are not significant) and the inclusion of continent dummies with the exceptions of temperature (no longer affect- ing infection rates) and older age (somewhat weaker size and significance without changing conclusions). 4. Conclusion Democratic countries may react slowly in the short term but place a higher value on human life and health.