McCarty and Groseclose: The politics of blame

McCarty and Groseclose. 2000. The politics of blame: Bargaining before an audience. AJPS.

Overview

Most models of Congressional-presidential bargaining are two-player games--but in reality, both sides are playing before a third player, the electorate, and both sides have incentives to behave differently as a result. Under divided government in particular, Congress wants to pass things that it knows the president will veto (to make him look extreme), and the president might sign things that are worse (for his preferences) than the status quo, but which help him appear moderate.

The Model

Though Congress and the president may (or may not, but that's irrelevant) have perfect information about one another's preferences, vetoes and gridlock can happen anyway. The electorate lacks information about the president's (and maybe Congress's ?) preferences, so politicians want to send the public incentives to signal these preferences (for electoral reasons). Thus, Congress and the president might play the "blame game": Congress will pass a bill that it knows the president will veto, or the president will sign a bill that gives more than he wants to to Congress. "Thus, despite Congress and the president being completely informed, an uninformed thir party causes the outcome to be Pareto inefficient."

Puzzles Addressed

1. Vetoes occur more frequently under divided government and during election years
2. During divided government, a president's popularity drops after a veto.
3. Gridlock happens. Though Krehbiel (1996) argues that gridlock happens only when a (perfectly informed) congress and president cannot agree, Groseclose and McCarty show that gridlock might happen b/c congress and the president don't want to agree.

Findings

1. B/c he wants to appear conciliatory, the president might sign bills that he does not prefer to the status quo.
2. Congress will take advantage of this to get gains for itself.
3. In some equilibria, bargaining breaks down and Congress proposes a bill that it knows will be vetoed.
4. This can happen despite perfect information and the presence of possible Pareto improvements.
5. (Main contribution): When the president vetoes a bill, this implies (i.e. signals) that his preferences are more extreme than voters thought, leading to a drop in approval ratings.

Implications

It might be good to keep committee hearings, labor-management disputes, and other negotiations secret. Without an audience, the parties will bargain for a mutually acceptable deal. But with an audience, the parties have incentives to send signals to the audience--which might inhibit making any agreement.