Seite - 559 - in VULNERABLE - The Law, Policy and Ethics of COVID-19

559Border Closures: A Pandemic of Symbolic Acts in the Time of COVID-19 closures (see Figures F4.2 and F4.3). A simple fixed-effects regres- sion confirms that enacting targeted border closures at any period during the outbreak is not statistically associated with fewer cases of COVID-19 (F(1,160) = 1.23, p = 0.268, within R2 = 0.0077) or fewer deaths (F(1,160) = 1.49, p = 0.223, within R2 = 0.0054).7 Targeted Border Closures Are Illegal In addition to their likely ineffectiveness, targeted border closures are also illegal. According to Article 43 of the International Health Regulations (see Box F4.1), countries are only permitted to enact “additional health measures” limiting travel during pandemics if such measures are 1) supported by science, 2) proportionate to the risks involved, and 3) respectful of human rights.8 These three con- ditions are intended to protect people and economies from needless harm and to avoid disincentivizing governments from alerting the World Health Organization (WHO) to new public health risks when first identified.9 The targeted COVID-19 border closures do not fulfil these three requirements.10 As described above, the science around targeted travel restrictions points against their effectiveness; research on previ- ous outbreaks of similar viruses suggests that these restrictions typi- cally delay their spread by only a few days. Second, targeted border closures are not proportionate to the risks involved, especially given that there are better strategies for addressing the COVID-19 outbreak, such as screening at airports, community surveillance, and public communication. Third, targeted border closures may violate human 7. In the fixed-effects regression model, each country is treated as an individual subject with repeated measures. A dummy variable was included to reflect whether a country had implemented a targeted border closure on any given day. Two statistical analyses were conducted, one with COVID-19 cases as the dependent variable of interest and a second with COVID-19-attributed deaths as the dependent variable. The F-statistic indicates model fit; neither model’s results are statistically significant. The very low within R2 value indicates that little variation in the dependent variables was explained by the enactment of targeted border closures across countries. All data were from: Max Roser et al, “Coronavirus Pandemic (COVID-19)” (last updated 26 May 2020), online: Our World  in  Data <https://ourworldindata.org/coronavirus-data-explorer> [Roser]. 8. World Health Organization, “International Health Regulations (2005)” (2005), online (pdf): World  Health  Organization <http://www.who.int/ihr/publications/ 9789241580496/en/>. 9. Habibi, supra note 1. 10. Ibid.