Menu Adam R Brown

Notes navigation: Browse by titleBrowse by authorSubject index

Cox, Kousser, and McCubbins: Party Power or Preferences? Quasi-Experimental Evidence from American State Legislatures

Disclaimer. Don't rely on these old notes in lieu of reading the literature, but they can jog your memory. As a grad student long ago, my peers and I collaborated to write and exchange summaries of political science research. I posted them to a wiki-style website. "Wikisum" is now dead but archived here. I cannot vouch for these notes' accuracy, nor can I say who wrote them.

Cox, Gary W.; Kousser, Thad; McCubbins, Mathew D. 2010. Party Power or Preferences? Quasi-Experimental Evidence from American State Legislatures. Journal of Politics 72 (July): 799-811.

Consider why some bills can get through a legislature but others can't. Perhaps (1) legislator preferences are all that matters; liberal legislators vote for liberal bills and against conservative ones. Perhaps (2) majority party cajoling also matters; if a strong majority leader can persuade his caucus to vote for a bill, it passes. Perhaps (3) the majority party's agenda powers also matter; if the Speaker or the Rules Committee can kill bills without a vote, then bills that might split the majority on their way to victory will never see the light of day.

Proponents of these three views have engaged in a long and productive debate for years. Mayhew (1974) and Krehbiel (1998) have pushed the first view. "Conditional party government" is a variant of the second view. And Cox and McCubbins have pushed the third view over and over.

In their recent article in JOP, Gary Cox and Mat McCubbins team up with state politics guru Thad Kousser. By looking at state legislatures instead of at Congress, they can look at quasi-experiments where agenda powers changed without a big simultaneous change in legislative membership. Consider:

Given how hard Cox and McCubbins have pushed the "agenda control" argument in previous work--what they call "cartel theory"--you might expect them to uncover MASSIVE effects in these two quasi-experiments. Well, not quite. But they do find effects. Here's the key table:

Table 1, Roll Rates and Policy Movement on Contested Roll Calls
Click to enlarge

So for each state, the "treatment" group (with agenda controls) is the left column, the "control" group (no agenda controls) is at right. There are fewer "majority rolls" when there are agenda controls--that is, bills that pass without a majority of the majority party's support. And there are more "minority rolls" when there are agenda controls--that is, bills passed without a majority of the minority party's support.

Comment and Criticism

Okay. So agenda control does matter then. Take away the majority party's agenda powers, and the minority will fare slightly better. But what's really striking in this table is how poorly the minority performs (and how well the majority performs) in BOTH columns. Even without a majority "cartel" controlling the agenda, it sucks to be in the minority.

Remember, there are three possible explanations for why some bills pass and others don't. The first is that legislator preferences alone determine it all. The second adds majority party cajoling. The third adds the majority's formal powers. Clearly, these formal powers matter. But it seems from this table that the majority's informal powers (cajoling, whipping) must matter even more than the formal ones. If not, then what explains the huge disparity between majority rolls and minority rolls when agenda powers are taken away?

Cox, Kousser, and McCubbins want to show that agenda control matters. They did. But without meaning to, they also showed that persuasive leaders and party cohesion matter even more.

Research by the same authors

Research on similar subjects

Tags

Cox, Gary W. (author)Kousser, Thad (author)McCubbins, Mathew D. (author)American PoliticsExperimentJournal of PoliticsLegislaturesMedian VoterPartiesParty GovernmentState Politics (U.S.)

Wikisum home: Index of all summaries by title, by author, or by subject.