Working Paper: CEPR ID: DP1935
Authors: Avinash Dixit; Gene M. Grossman; Faruk Gul
Abstract: We study political compromise founded on tacit cooperation. Two political parties must share a fixed pie in each of an infinite sequence of periods. In each period, the party in power has ultimate authority to divide the pie. Power evolves according to a Markov process among a set of political states corresponding to different degrees of political ?strength? for the two. The political strength of each party is a state variable, and the game is dynamic, rather than repeated. Allocations in an efficient, sub-game perfect equilibrium do not follow a Markov process. Rather, a party?s share reflects not only its current strength, but also how it got there in the recent history. We characterize the efficient division processes for majority rule and supermajority rule, and ask whether one regime allows greater compromise than the other.
Keywords: policymaking; compromise; risk sharing; political parties; dynamic games
JEL Codes: C73; D78
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| current political strength (P16) | division of resources (D30) |
| historical context (B15) | division of resources (D30) |
| recent history of power transitions (P39) | division of resources (D30) |
| popularity surpassing previous peaks (E32) | allocation shifts in favor of a party (D72) |
| supermajority regimes (D72) | less efficient outcomes (D61) |
| institutional rules (D02) | inequities in resource allocation (I14) |