The usual suspects: do risk tolerance, altruism, and health predict the response to COVID-19?

We survey college students during California's stay-at-home order to test whether compliance with social distancing requirements depends on primary preferences and characteristics that affect their marginal benefit from doing so. We find a quarter of students violated the order. Yet, neither risk preference, altruism, nor preexisting health conditions were predictive of compliance. Our findings raise doubt about the efficiency of minimally enforced social distancing policies, as well as commonly assumed motivations for compliance. Our results also imply that those with preexisting health conditions may not voluntarily comply, resulting in higher health care congestion than otherwise expected.