Robust Technology Regulation

Abstract

We analyze how uncertain technologies should be robustly regulated. An agent develops a new technology and, while privately learning about its harms and benefits, chooses whether to continue R&D. A principal chooses among dynamic mechanisms to shape R&D choices in different states. A regulatory sandbox mechanism comprising of a zero marginal tax on R&D up to a hard limit is (i) robust: it delivers optimal payoff guarantees when nature chooses the agent's learning process and preferences adversarially; (ii) dominant: it outperforms other robust mechanisms evaluated at any learning process and agent preference; and (iii) important: absent a hard limit, worst-case payoffs can be arbitrarily poor and is induced by weak but growing optimism generating excessive risk-taking. If regulators also learn, an adaptive sandbox optimally incorporates new information while safeguarding against the worst-case. Our results offer optimality foundations for existing policy as well as guidance for future policy. 

Keywords: Mechanism Design, Information Design, Regulation, Robustness, Learning, AI, Sandbox