Анонс статьи, с которой Торун будет выступать завтра, выглядит так:
Learning by trial and error can yield better policy outcomes. Realization of such benefits might depend on the type of accountability office holders are subject to. We model a policy environment in which an incumbent chooses between a ``safe'' option with a known payoff and a ``risky'' policy that yields higher or lower payoffs depending on the policymaker's type. We consider two types of elections: in ``non-competitive'' elections, voters learn about the incumbent's type and compare the expected payoffs from retention against dismissal and replacement by another representative agent; in ``competitive'' elections, an opponent campaigns on a ``safe'' or ``risky'' platform, voters evaluate the type of both incumbent and challenger, and choose between the two alternatives. Even when the expected payoffs are lower a representative agent will engage in ``risky'' policies in the first period. These efficiency gains are mitigated when politicians are subject to non-competitive election. First period investment in policy learning is (weakly) inefficient under competitive elections also. Despite this competitive elections can induce efficient policy learning that is due to the interaction between the incumbent and the challenger. Voters can use their votes to simultaneously select high performers and provide incentives to induce further efficiency gains. We analyze and provide new insights to incumbency advantage, show existence of an incumbency disadvantage, and explore the impact of term limits in our policy learning environment.
Обратите внимание, что время необычное, но так в этом году будет всегда. Заседание состоится в 16:40 по адресу: 101990, Москва, Покровский бульвар, 11, Высшая школа экономики, корпус Ж, аудитория Ж-419.