Compliance with social distancing during the Covid-19 crisis

Paola Giuliano, Imran Rasul 18 June 2020



The Covid-19 crisis has confronted societies with unprecedented challenges to both their economies and their health systems. A key policy response in many countries has been the implementation of social distancing measures. In the following, we summarise recent evidence on the effects of social distancing based on real time data collection projects underway since the start of the crisis.

In April 2020, the European Economic Area (EEA) led an initiative to create a registry of research projects predominantly engaged in data collection with the goal of tracking economic outcomes in real time during the pandemic. The initiative has by now about 350 registered projects, covering many areas of economics including consumption and savings behaviour, labour markets as well as economic and political attitudes. In this article, we use evidence from completed papers in the registry and summarise the cross-project findings in the broad areas of economic and political attitudes and beliefs.1

If social distancing is key to slowing the spread of Covid-19, it is important to know what determines whether individuals will effectively adopt such practices. The strength of the social fabric of society but also the degree of trust in institutions are crucial elements in determining individual behaviour. Social capital has been shown to be relevant for contributions to public goods (Putnam 1993, Guiso et al. 2019), and the internalisation of negative externalities created by personal mobility is likely to be related to that. Higher levels of social capital should therefore be crucial for citizens' willingness to voluntary comply with social distancing measures.

Several papers using real-time data from different countries, have found evidence in supports of the importance of social capital and trust in institutions for compliance with social distancing during the pandemic. Durante et al. (2020) use data on individual mobility across Italian provinces to show that, after the beginning of the pandemic, mobility declined disproportionately in areas with higher civic capital.

Figure 1 Mobility and civic capital (differences between the top and bottom quartile), data from Italian provinces

Notes: The figure shows the difference in mobility between the top and bottom quartile in social capital for three periods: before 21 February (the day on which the first Covid-19 hot-spot was identified), between 21 February and 9 March (when the national lockdown came into effect) and after 9 March. Data on mobility are based on phone location tracking and record the number of movements of an individual during the day from home to work or to other places (shops, bars, restaurants, gyms, etc.) and vice versa. Social capital is defined as the principal component of three measures (the number of blood donations per 10’000 people, a survey-based measure of trust in others, and a measure of newspaper readership).

Source: Durante et al. (2020).

Bargain et al. (2020) use data on mobility and trust in government at the regional level in Europe. They find that high-trust regions decrease their non-essential mobility significantly more than low-trust regions. Further, lockdown policies are found to be more efficient in reducing mobility in regions with higher levels of trust. This is consistent with the idea that trust in institutions improves regulatory efficiency and compliance with rules and laws.

Figure 2 Daily mobility and political trust, variation across European regions

a) Retail and recreation

b) Work

c) Transit stations

d) Private residencies

Notes: The figure shows the authors’ calculations of mobility reduction based on Google mobility data and data on trust in politicians taken from the European Social Survey. Areas represent the 95% CI of average daily mobility across European regions (weighted by 1/number of regions in the corresponding country). “Trust” and “distrust” indicate regions, within each country, with trust level above/below the country average.

Source: Bargain et al. (2020). 

For the the US, Brzezinski et al. 2020 find that counties with higher income and education, and with higher levels of trust in science, are more likely to voluntary comply with physical distancing measures. However, the authors note that government action is endogenous to communities’ reactions. Hence, they examine the two-way interaction between physical distancing and policies, and find that government policies can amplify measures already taken at the community level. At the same time, restrictive policies are less necessary when the community acts independently in maintaining social distancing measures.

Barrios et al. (2020) also use individual and county level data for the US as well as data on European regions to investigate the importance of social capital for voluntary compliance with social distancing measures. The paper also studies differences in social distancing after US states began re-opening, finding that compliance is much higher in counties with high civic capital.

From a policy perspective, these studies have important implications: places with higher social capital are more likely to voluntary comply with social distancing measures. Hence, in these places, targeted policies can be as effective as across-the-board lockdowns but come at much lower economic costs.

Perhaps the most important element of compliance is the ability of governments to persuade people to internalize the externality they would impose on the community by not reducing their mobility. This is particularly important in democracies, where the ability to enforce lockdowns on a large scale is limited. Evidence from different countries indicates that political beliefs, coupled with differences in media consumption, have important implications for risk perceptions and compliance with social distancing.

For the case of Brazil, Ajzenman et al. (2020) find that pro-government localities are less inclined to follow social distancing measures if the president dismisses the risks associated with the pandemic. This relationship is even stronger in places with higher levels of media penetration.

The way in which people consume news and interpret facts is especially relevant for the US with its increasing political divide (Barrios and Hochberg 2020, Allcott et al. 2020, and Gadarian et al. 2020). The aforementioned papers all find that pro-Trump counties are less prone to keep social distancing,  a result of different risk perceptions emphasized by different news outlets.

Figure 3 Partisan differences in social distancing in the US

Notes: The figure shows the estimated coefficients for county partisanship on social distancing, calculated as the log number of point of interest (including restaurants, hotels, hospitals and many other public and private businesses) visits in a given county.

Source: Allcott et al. (2020).

To isolate the role of information, Bursztyn et al. (2020) study individuals watching two different shows on Fox News, one that warned viewers about the risks of the pandemic (Tucker Carlson Tonight) and one that instead dismissed the risks (Hannity). Carlson’s viewers changed behaviour earlier which impacted both the number of deaths and infections. Painter and Qiu (2020) examine partisan heterogeneity in response to state-level “stay-at-home” orders and show that such orders are more likely to be abided by Democrats when the governor is a Democrat. Grossman et al. (2020) document that governors’ recommendations – proxied by their tweets—preceded the issuance of “stay at home” orders, and that this had a significant effect on mobility, with the effect being stronger for Democratic counties. They also find that both Democratic and Republican counties are equally responsive to Democratic governors, where the same time Democratic counties are more responsive to Republican governors than Republican counties.

Overall, there is evidence that individual perception about the severity of the virus is strongly influenced by news consumption and whether information comes from someone with similar or different political leanings than themselves. Reliability of messages from political leaders, providing consistent information about the severity of Covid-19 could hence substantially affect the response to the pandemic.

We hope that these and new real time data collection projects will help shape the understanding of the importance of attitudes and beliefs as governments start to ease lockdown restrictions. Newly formed attitudes could play a pivotal role in shaping post-Covid societies.


Ajzenman, N, T Cavalcanti and D Da Mata (2020), “More Than Words: Leaders’ Speech and Risky Behavior During a Pandemic”, SSRN 3582908.

Allcott, H, L Boxell, J Conway, M Gentzkow, M Thaler and D Yang, (2020), “Polarization and Public Health: Partisan Differences in Social Distancing during the Coronavirus Pandemic”, NBER Working Paper 26946.

Bargain, O and U Aminjonov (2020), “Trust and Compliance to Public Health Policies in Times of COVID-19”, IZA Discusion Paper 13205.

Barrios, J, E Benmelech, Y Hochberg, P Sapienza and L Zingales (2020), “Civic Capital and Social Distancing during the Covid-19 Pandemic”, NBER Working Paper 27320.

Barrios, J M and Y V Hochberg (2020), “Risk perception through the lens of politics in the time of the COVID-19 pandemic”, Working Paper.

Brzezinski, A, V Ketch, D van Dijcke and A Wright, “Belief in Science Influences Physical Distancing in Response to COVID-19 Lockdown Policies”, University of Chicago, BFI, Working Paper 2020-56.

Bursztyn, L, A Rao, C Roth and D Yanagizawa-Drott (2020), “Misinformation during a pandemic”, BFI Working Paper 2020-44.

Durante, R, L Guiso and G Gulino (2020), “Civic capital and social distancing: evidence from Italians’ response to COVID-19”,, 16 April.

Gadarian, S K, S W Goodman and T B Pepinsky (2020), “Partisanship, health behavior, and policy attitudes in the early stages of the COVID-19 pandemic”, Working Paper.

Grossman, G, S Kim, J Rexer and H Thirumurthy (2020), “Political Partisanship Influences Behavioral Responses to Governors’ Recommendations for COVID-19 Prevention in the United States”, 2020, SSRN 3578695.

Guiso, L, P  Sapienza and L Zingales (2016), “Long-Term Persistence”, Journal of the European Economic Association 14 (6): 1401-1436.

Painter, M O and T Qiu (2020), “Political belief affect compliance with Covid-19 social distancing orders”, Covid Economics: Vetted and Real-Time Papers 4, 103–123.

Putnam, R (1993), Making democracy work, Priceton, NJ: Princeton University Press.


1 The original call for projects is here, and the list of currently registered projects, updated every few days, is here.



Topics:  Covid-19 Institutions and economics Politics and economics

Tags:  COVID-19, social distancing, attitudes and beliefs, institutions

Professor of Economics and Justice Elwood Lui Endowed Term Chair in Management at the UCLA Anderson School of Management

Professor of Economics, University College London


CEPR Policy Research