Community-level social capital and COVID-19 infections and fatality in the US

Francesca Borgonovi, Elodie Andrieu, S V Subramanian 22 July 2020



Evidence on the role social and civic capital on information acquisition and behavioural responses to the pandemic such as mobility changes is emerging (Bargain and Aminjonov 2020, Durante et al. 2020, Borgonovi and Andrieu 2020, Bartscher et al. 2020, Barrios et al. 2020), and so is evidence on the association between social capital and the spread of COVID-19 (Bartscher et al. 2020).

Social capital reflects the resources and benefits that individuals and groups acquire through connections with others. It involves both shared norms and values that promote cooperation as well as, crucially for disease spread, actual social relationships (Kawachi et al. 2008, Fukuyama 2000, Putnam 1993). It is therefore related but conceptually distinct from civic capital, conceived as “the set of values and beliefs that help a group overcome the free-rider problem in the pursuit of socially valuable activities” (Guiso et al. 2011). Social capital reflects both attitudes and dispositions that promote interpersonal cooperation as well as actual social connections and relationships within a community.

In a recent paper (Borgonovi et al. 2020), we argue that the two components of social capital played a different role in shaping variations across communities in the evolution of the COVID-19 pandemic. Relational social capital promoted a rapid initial spread of COVID-19 infections while cognitive social capital protected communities as the pandemic progressed, reducing case fatality because of a lower baseline incidence of pre-existing conditions and by promoting behavioural changes in response to the pandemic.


We used data from counties in the US to evaluate the association between social capital, measured through ‘the geography of social capital project’ and (1) the number of days between 22 January (the first day with available data) and the day when at least ten COVID-19 diagnoses were recorded (we run models with a threshold of 15 cases for robustness); and (2) cumulative case fatality between 22 January and 8 May, calculated as the total number of deaths attributed to COVID-19 divided by the total number of diagnoses.

Figure 1 illustrates the evolution of the weekly number of recorded infections in the US in counties with high (top 25%) and counties with low levels (bottom 25%) of social capital. The two vertical lines mark 17 March, when announcements on the importance of social distancing were first made, and the start of the period when many states implemented sheltering-in-place regulations. Figure 1 illustrates that at the start of the pandemic, more cases were recorded in counties with high levels of social capital than counties with low levels of social capital. The figure also illustrates how, following public announcements on the importance of social distancing case numbers grew less in high social capital than low social capital counties. The latest period, which reflects changes in cases following the relaxation of regulations to ensure social distancing, suggests a slowing of the pandemic in high social capital areas but an acceleration in low social capital communities.

Figure 1 Social capital differential in COVID-19 infections

Notes: Lines represent the log of the county average number of additional COVID-19 infections per 100 000 individuals. High social capital groups are areas within the top 25% in social capital and counties with low levels of social capital are within the bottom 25% in social capital

Sources: The geography of social capital in America and USA Facts website.

In our paper, we present results on the very beginning of the pandemic and identify social capital differentials in the speed with which COVID-19 reached communities and in case fatality. Results on disease spread control for the economic sector of the county, the share of population in the county above 65, population density and the number of beds in intensive care unit as well as for state fixed effects to account for factors could determine the intensity of close physical contacts in a community and, as a result, how fast transmission occurred. By including state fixed effects, we exploit within-state but between-county variations in infections and social capital, and therefore account for measurement error potentially due to differences in reporting standards across states as well as any other state level difference.  Results on case fatality control for the economic sector of the county, the share of population in the county above 65, the number of beds in intensive care unit, state fixed effects as well as mobility changes observed between March 16 and the week starting on April 20. Mobility is an important behavioural change that communities implemented in response and that may also signal the willingness of individuals to adopt other health protective behaviours.  

Results suggest that on average in the US it took a county 68 days from 22 January to reach at least ten COVID-19 cases. In counties with a higher level of social capital the propagation of COVID-19 was faster than in counties with lower levels of social capital: a change of one standard deviation in the social capital index was associated with a reduction of around three days in time elapsed until the county reached ten (or 15) cases. We estimate a cumulative case fatality rate of around 3.5% in the period between 22 January and 8 May 2020, on average, in the counties in our sample: i.e. for every 100 cases with a positive diagnosis of COVID-19 up to 8 May, 3.5 COVID-19 related deaths were recorded (see Tables 4 and 5 in Borgonovi et al. 2020). We also find that a difference of one standard deviation in social capital was associated with a larger difference in the case fatality rate of around 4 per 1,000.

Until a vaccine or effective treatments become available, the impact of the pandemic on the health of communities in the US will crucially depend on the willingness and ability of such communities to change their behaviours to reduce infections and to protect the most vulnerable (van Bavel et al. 2020). As national and local leaders across the world are relaxing limitations to freedom of movements, lifting shelter in place regulations, and reopening schools, social capital could play a role in protecting the health of community members. Personal behaviours such as washing hands frequently and adequately, wearing face masks when encountering others, avoiding close physical contact with others, isolating if one develops any symptoms can effectively reduce infections and transmission (Cowling et al. 2020).

Adopting such behaviours can be difficult and costly for individuals. Individuals living in communities with high levels of social capital might be more motivated and prepared to adopt such behaviours; in such communities, members may be less likely to adopt free-riding behaviours and may be more willing to sustain personal costs for the benefit of the community.


Bargain, O and U Aminjonov (2020), “Trust and compliance to public health policies in times of Covid-19”, IZA DP No. 13205.

Bartscher, A K, S Seitz, S Siegloch, M Slotwinski and N Wehrhöfer (2020), “Social capital and the spread of Covid-19: Insights from European countries”, IZA DP No. 13310 (see also the Vox column here).

Borgonovi, F and E Andrieu (2020), “Bowling together by bowling alone: Social capital and Covid-19”, Covid Economics 17: 73–96 (see also the Vox column here).

Borgonovi, F, E Andrieu and S V Subramanian (2020), “Community level social capital and COVID-19 infections and fatality in the United States”, Covid Economics 32: 110–126.

Durante, R, L Guiso and G Gulino (2020), “Civic capital and social distancing: Evidence from Italians’ response to Covid-19”, mimeo.

Cowling B J, S T Ali, T W Y Ng et al. (2020), “Impact assessment of non-pharmaceutical interventions against COVID-19 and influenza in Hong Kong: an observational study”, medRxiv 2020.03.12.20034660. doi:10.1101/2020.03.12.20034660

Fukuyama F (2016), “Social Capital and Civil Society”, IMF Working paper no. 00/74. 0.

Guiso, L, P Sapienza and L Zingales (2011), “Civic capital as the missing link”, J Benhabib, A Bisin and M Jackson (eds), Handbook of Social Economics 1: 417-480.

Kawachi I, S V Subramanian and D Kim (2008), “Social capital and health: A decade of progress and beyond”, in Social Capital and Health, Springer, pp. 1–26.

Putnam, R D (1993), Making Democracy Work: Civic Traditions in Modern Italy, Princeton University Press.

van Bavel J J, K Baicker, P S Boggio et al. (2020), “COVID-19 pandemic response”, Nature Human Behavior 1–12. doi:10.1038/s41562-020-0884-z



Topics:  Covid-19

Tags:  COVID-19, social capital, social norms, disease spread

British Academy Global Professor, Institute of Education - University College London

PhD student in Economics, King’s College London

Professor of Population Health and Geography, Harvard University


CEPR Policy Research