10 July 2020 – Policy Brief
Executive summary
Digital proximity tracing (DPT) can complement traditional contact tracing because it allows contacts in anonymous crowds to be alerted. However, even the best technological solution needs to contend with the human factor.
We assess the incentive structure faced by users of decentralized DPT apps. We show that DPT is impossible to enforce by decree or material incentives. Its success therefore hinges on widespread voluntary cooperation. Compliance breaks down into ‘passive’ actions (downloading the app and carrying it around) and ‘active’ actions (entering an alert when tested positive).
We show that ‘passive’ adoption generates considerable private benefits that many may underestimate – especially by offering information to guide their behavior towards vulnerable relatives and friends.
‘Active’ compliance is closer to a typical social dilemma, where private costs must be weighed against societal benefits. We argue that the costs are likely to be negligible for most users, while the social benefits (saving lives) are potentially large.
Therefore, public information while being transparent about risks should make clear that adopting the app has considerable private benefits. Moreover, it could be stressed that the social benefits increase more than proportionately with the population share of app users.
One informational nudge could be for government, in its regular Covid-19 data releases, to communicate the share of new infections (i) that were detected thanks to an app signal received, and/or (ii) that led to an app signal sent by the infected person.
Date of request: 6/7/2020
Date of response: 10/7/2020
In response to request from: NCS-TF economics group
Comment on planned updates: Update planned as soon as an important body of new evidence becomes available.
Expert groups and individuals involved: Document written by economists from University of Lausanne and EPFL (E4S Center): J.P. Bonardi, M. Brülhart, J.P. Danthine, A. Saxena, C. Thöni, M. Thoenig, C. Zehnder (with important input from S. Suggs, E. Bugnion and M. Salathé)
Contact persons: M. Brülhart