Economic Connectedness As a Metric of Bridging Social Ties and Its Association with COVID-19 Infections in Arizona

Background: Underserved communities in Arizona have been disproportionately affected by the COVID-19 pandemic due to limited resources for testing and meeting basic needs. Zip code tabulation areas (ZCTAs) with high numbers of racial/ethnic minorities and low-income residents were particularly vulnerable. Social capital, which refers to the social networks that underpin the exchange of COVID-19 related (dis)information, is believed to be relevant to the risk of COVID-19 infection. Social network theories suggest that individuals with bridging social ties - connections with people from different social identities and backgrounds - will have better access to novel information, resulting in increased knowledge to lower the risk of COVID-19 infection. Economic connectedness (EC), a metric produced by subscribers of Facebook living in each ZCTA, measures community-level bridging ties. Our study aims to examine whether ZCTAs with higher bridging social ties can explain disparities in COVID-19 infections between October 2020 and November 2021.

Methods: The sample included 194 ZCTAs in Arizona which made available of biweekly COVID-19 case counts from October 21, 2020 to November 25, 2021 and the ZCTA-level EC. EC measured the bridging social ties of lower-SES subscribers living in each ZCTA, indicating the share between high-SES Facebook friends and low-SES Facebook friends. Covariates were obtained from the five-year estimates of the American Community Survey in 2020 and included the share of female residents, residents over 65 years old, racial/ethnic minority groups, residents using public transportation and residents uninsured. Also, we used average household size, and median per capita income. Linear OLS regression analyses were conducted using SPSS 28.0 (IBM) with an alpha level of 0.05.

Results: The average EC for ZCTAs in Arizona was 0.86, indicating that low-SES individuals had less than 50% of their Facebook friends with high-SES. EC was higher in ZCTAs with more residents identified as White (r=0.45, p<0.05) and Asian (r=0.47, p<0.05), while lower EC was reported in ZCTAs with more residents identified as Black (r=-0.20, p<0.05) and American Indian (r=-0.35, p<0.05). Adjusted analyses showed that ZCTAs with higher EC had fewer biweekly COVID cases (p<0.05). Even after applying the Bonferroni correction to prevent the risk of Type 1 error, significant correlation between EC and COVID-19 cases between October 2020 and April 2021 was still observed (p<0.001). The standardized coefficient of the association between EC and the biweekly COVID-19 cases increased up to -0.44 (p<0.001) near January 2021 when Arizona was passing the COVID-19 surge, followed by a steady decrease to -0.27 in November 2021.

Conclusions and Implications:Our findings highlight the significance of bridging social ties that vulnerable groups possess and demonstrate how they can be indispensable resources necessary to cope with the threats posed by COVID-19. Analyzing bridging social capital through the measure of EC can provide a useful framework for identifying communities where health disparities are present. Multilevel programs targeting social networks of residents are inevitable to address disparities in access to care during pandemic.