Although many important legislative theories have been devised with the U.S. Congress in mind, attempts to test them with cross-national data are rare. In this issue of the Quarterly, Shane Martin provides an important example of how this sort of work can be accomplished. Martin investigates the relationship between electoral institutions and the internal organization of legislatures. Building off the congressionally inspired idea that strong committees emerge from candidate-centered electoral systems, Martin posits that how legislators cultivate personal votes actually determines the importance of committees. He hypothesizes that where legislators can supply particularistic (or pork) benefits, committees are stronger; where legislators can only supply (nonpork) constituent services, committees are weaker. A test is conducted on data gathered on 39 national assemblies. The results unequivocally support the hypothesis. Thus, the ability to generate a personal vote does influence legislative organization, but in a more complicated fashion than heretofore supposed.
Ideal point estimators have come to be employed in a wide range of legislative voting studies. Cesar Zucco Jr. and Benjamin E. Lauderdale note that most ideal point estimators do not distinguish between ideological motivations and pressures brought to bear on lawmakers by political parties, executives, or others. Using survey data on Brazilian legislators, they introduce a Bayesian ideal point estimator that allows them to separate ideological motivations from what they term “government inducements” or vote-buying efforts. Analyzing Câmara dos Deputados roll-call votes between 1989 and 2009 (a period that encompasses five presidencies), they find that two dimensions emerge—one ideological, the other government and opposition. The government and opposition dimension has actually become the stronger dimension over the last decade or so. These results have two implications for legislative studies. First, in respect to our understanding of the Brazilian National Congress, studies from the 1990s that found voting was structured by ideology have now been superseded by evidence of the emergence of an additional government and opposition dimension. Second, legislative scholars should, wherever possible, incorporate contextual information into ideal point estimates; doing so promises a more nuanced understanding of legislator behavior.
Over the last two decades considerable energy has been devoted to studies of party influence on member voting in the U.S. Congress. James Coleman Battista and Jesse T. Richman extend a technique developed by Ansolabehere, Snyder, and Stewart to measure party influence on congressional roll-call voting to state legislatures. Using responses to the Project Vote Smart National Political Aptitude Test survey of state lawmakers to calculate ideal points, Coleman and Battista examine votes from 27 lower houses during the 1999–2000 biennium. They uncover detectable party influence on 44% of all roll calls and 69% of close votes. Importantly, however, they find considerable range on those figures across the chambers they study. Examining this further, they find that greater state diversity produces stronger party influence on roll-call voting, as does legislative careerism. Majority party size matters only on close votes, with narrower majorities inducing greater party influence. This article provides another example of how the study of state legislatures can provide important analytical leverage for testing congressionally inspired theories.
It seems reasonable to conceive of a legislature as a social network. Kathleen A. Bratton and Stella M. Rouse employ social network analysis to explain bill cosponsorship in the lower houses of nine U.S. state legislatures. They are interested to see the extent to which different connections that legislators have influence their decisions about which measures to cosponsor. The results of their nine exponential random graph models show strong support for the idea that lawmakers cosponsor bills introduced by ideologically similar colleagues, bills introduced by members from neighboring districts, and bills written by fellow committee members. Evidence also suggested that similarities in race, ethnicity, and gender, increase the probability of cosponsorship. There are, however, a fair number of behavioral differences across the nine chambers studies, suggesting that social networks are complicated and involve potential relationships and contexts that should be explored.
The basic dynamics of congressional elections are well understood. But Jamie L. Carson, Michael H. Crespin, Carrie P. Eaves, and Emily Wanless offer an interesting twist by examining the relationship between constituency congruence and candidate competition for the U.S. House. Exploiting GIS software, they calculate the population congruence between state legislative districts and congressional districts. Their expectation is that state legislators who represent larger proportions of House districts will be both more likely to run for the House and more likely to do well in such a contest. Using data from the 2004 and 2006 elections, their analysis, which includes the appropriate controls, confirms strategic behavior on the part of state lawmakers. They are more likely to run where congruence is higher, particularly when a seat is open. Increased congruence also pays off at the polls. All of this demonstrates that state legislators develop a personal vote and that it carries over to other elections.
Return to August 2011 Titles
Return to LSQ home page