# Ulric B. and Evelyn L. Bray Social Sciences Seminar

Wednesday, May 11, 2022
4:00pm to 5:00pm
Baxter Lecture Hall
Partisan Traps
Ethan Bueno de Mesquita, Sydney Stein Professor and Deputy Dean at the Harris School of Public Policy, University of Chicago,

Abstract: Why may opportunities for welfare improving (bipartisan) reforms be frequently ignored by policy makers? We propose that incumbents desire to signal to their ideologically aligned constituents that opportunities for bipartisan reforms are rare and instead most of policy making is zero-sum in nature. This is because in a zero-sum world, ideological alignment is electorally insulating. We capture this logic in an infinite-horizon model and characterize the resulting dynamics of partisanship and bipartisanship in policy making.

We show that in any equilibrium, the voter retains a partisan-aligned incumbent as long as the voter is sufficiently pessimistic about the frequency of bipartisan opportunities and replaces the incumbent with a partisan-misaligned challenger otherwise. When the voter is very pessimistic about bipartisan opportunities, she stays sufficiently pessimistic even if she observes a bipartisan policy. Anticipating that, the incumbent implements bipartisan policies whenever available. When the voter is somewhat less pessimistic, however, a bipartisan policy pushes her to elect the challenger. Anticipating that, the incumbent implements only partisan policies. So, for moderately pessimistic beliefs, players are stuck in a partisan trap: the incumbent never implements a bipartisan policy and remains in power forever.

The described features of the equilibria have interesting implications for the dynamics of beliefs and bipartisanship. First, if the voter starts off pessimistic, she will remain pessimistic forever. Second, a pessimistic voter is better off the more pessimistic she is, as bipartisan opportunities are only taken advantage of when she is sufficiently pessimistic. Third, if the voter starts very pessimistic but the state is high (or there are a few good'' draws), an initial period of bipartisanship will be followed by a partisan trap. Finally, for any non-degenerate belief, no matter how optimistic, there exists a finite sequence of bad realizations when bipartisan opportunities are not available after which the voter chooses an aligned politician over a misaligned politician. Once this happens, the players are stuck in a partisan trap forever. Hence, partisan traps occur with positive probability even if the voter starts with optimistic beliefs.