Costs and Benefits of Social Distancing

Principal Investigator: Pamela Giustinelli


Motivation The Covid-19 pandemic is confronting experts, policy-makers, and citizens with many dimensions of deep uncertainty and difficult decisions to make. Notwithstanding the global nature of the crisis, the responses of individual countries and their citizens to the emergency have not been uniform. The adoption, time of introduction, strictness, and expected duration of social distancing measures relative to the epidemic progression have varied greatly across countries, and to some extent also within them. Citizens, in turn, have displayed different degrees of compliance both within and across countries. While extremely draconian measures are not acceptable in western democracies, at the same time simple ‘nudging’ approaches do not appear particularly effective. Therefore understanding citizens’ perceived uncertainties associated to the Coronavirus emergency and its management by the public authorities, citizens’ perceived costs and benefits of social distancing, and how these vary across different groups of individuals and households, is of fundamental importance for policy.


Project summary We study the above issues by leveraging the most recent advances in survey design and econometric analysis of probabilistic expectations. We survey individuals in the UK, Italy, and the US regarding their perceived risks and benefits of complying (or failing to comply) with alternative social distancing measures implemented in their respective countries. Perceived risks and benefits of compliance (or lack thereof) include (higher/lower) subjective probabilities of: contracting Coronavirus, falling ill with Covid-19, being hospitalized, passing away, being fined, losing job, becoming socially isolated, and so on. Using these data, we document the extent and dimensions of people’s uncertainty surrounding the Coronavirus emergency.[*] Moreover, by combining these data with self-reported information on respondents’ compliance behaviors (current and expected) and with respondents’ characteristics (socio-demographics, health, familiarity and experience with Coronavirus), we estimate a model of compliance behavior that enables us to quantify the trade-offs people face when deciding whether to comply with social isolation measures, and if so to what extent. We are particularly interested in people’s perceived costs of social isolation; we provide estimates of monetary compensation required to be socially isolated.


Project status We have completed a draft of the survey for UK respondents, which we are currently piloting with UK workers on the MTurk platform. We are preparing survey versions for Italian and US respondents, which we also plan to pilot soon with Italian and US workers on MTurk. We will use the data from the pilot to revise the survey questions as appropriate. After that, we plan to field the surveys on larger and more representative samples of English, Italian, and American respondents.

[*]Respondents’ subjective probabilities are elicited on a numerical scale of percent chance, ranging from 0 to 100 percent, by means of a clickable slider (i.e., a slider whose cursor appears when the respondent clicks on it) in order to avoid anchoring. The survey protocol experiments with both point and interval expectations formats, thus also providing a methodological contribution. The point format, most commonly used in the survey expectations literature so far, requires that respondents express the likelihood of the event with a single number between 0 and 100 percent. The interval format enables respondents to express uncertainty about the likelihood of the event in question and it is particularly useful in settings where individuals may have imprecise subjective probabilities due to partial knowledge (deep uncertainty). 

06.04.2020 Bocconi Knowledge "Why People Comply, or not, with Social Distancing Rules"