Matthew Hindman

Party discipline ?

This conference paper (coauthored with Rodolfo Espino) looks at the "parties v. preferences" debate using an unusual data source: roll call voting data from the Arizona Territorial Legislature.  Party organizations were slow to arrive in Arizona; when they finally did emerge, they were weaker and more divided than in nearly any other area of the U.S.  The five decades of territorial voting data help disentangle the influence of partisanship, and suggest that neither side in the "parties v. preferences" debate has it quite right.  

Click here for a .pdf file of the paper.  Full citation and abstract below the fold.  

Citation: 

Espino, Rodolfo and Matthew Hindman. 2007.  "We Don't Need No Stinkin' Parties: Roll Call Voting in the Non-Partisan Arizona Territorial Legislature." Paper Presented at the Annual Meeting of the Western Political Science Association. Las Vegas, NV, March.

Abstract:

Scholars of legislative organization have long debated the importance of parties in structuring representatives' voting behavior.  Poole and Rosenthal (1985, 1991, 1997) find that voting in the U.S. congress is dominated by a single ideological dimension, and conclude that political parties are responsible for this fact. Other scholars had also focused on the ``bonding'' effects of parties. Yet critics have proposed alternative explanations, suggesting that legislators' preferences alone can produce similar outcomes (Krehbiel 1998), or that such findings are an artifact of data from the (highly partisan) U.S. Congress.

In this paper, we attempt to disentangle the influence of party by examining five decades of roll call voting from the non-partisan Arizona territorial legislature. Even without legislative parties, territorial legislators initially show voting patterns as highly structured than those in the U.S. Congress, with a single policy dimension dominant. Yet voting patterns become steadily less orderly over time, even after the emergence of territorial parties in the 1880s.  What Poole (2005) describes as ``basic dimensionality'' only reasserts dominance after 1900, as Arizona approaches statehood. The Arizona data suggests that legislative parties can provide long-term stability in voting behavior, but that they are neither necessary nor sufficient to create a low-dimensional policy space.