Be careful about what you wish for. Two generations ago political scientists cast a jealous eye toward differentiated, disciplined European political parties. The diffuse, centrist American parties seemed to preclude responsible party government. By the 1970s, political parties in Congress were becoming both more disciplined and more distinguished from each other, and that development has continued. It reflects the increasing differences in the party composition across congressional districts. Recent analyses also suggest that it is asymmetric, exhibited mainly by the increasing conservatism among Republican members of Congress.
The first two articles in this issue deal with two aspects of this change in the U.S. party system. Daniel M. Butler asks whether this asymmetric polarization in Congress reflects voting behavior in the electorate or is the product of party leaders. Using data on roll-call votes of incumbent members of Congress in the House of Representatives in 2001 and 2002 and data on the party identification of the voters in their districts, Butler shows that the degree of what he calls "roll-call extremity" among Republican members of the House is related to the increasing proportion of Republican party identifiers in their districts. There is no counterpart in the districts of Democratic members. He concludes that the asymmetric polarization in Congress is a response to the electorate and is not merely shaped by party elites.
The effect of divided government—of having a different party in control of Congress than of the presidency—has been a subject of accelerated research, ever since David Mayhew came to the surprising conclusion in his 1991 book that it made little difference in the passage of important legislation or the initiation of important investigations. David C.W. Parker and Matthew Dull reassess Mayhew’s conclusion so far as congressional investigations are concerned, using a large new set of data on congressional investigations between 1947 and 2004. They have a new standard of what constitutes "importance." Using data from the Congressional Information Service, they measure importance by whether the investigation dealt with violations of law, mismanagement, or abuse of discretion. Altogether their data include 1,015 investigations. Their conclusion is that since the reforms of the 1970s and the attendant changes in the party system, there were more investigations of waste, fraud, and abuse in the House of Representatives (but not in the Senate) during periods of divided government than when the same party controlled both branches. The consequences of increased partisanship, observed in many aspects of congressional behavior, appear to have had consequences for the frequency of congressional investigations as well.
The next two articles demonstrate once again the value of comparative research across legislative institutions. Jeffrey Lazarus and Amy Steigerwalt have given us an example of how institutional structure affects policy outcomes, even across chambers in a bicameral legislature. Legislative research has not only focused overwhelmingly on the U.S. Congress, but within Congress the focus has predominantly been on the House of Representatives. Lazarus and Steigerwalt offer four hypotheses on how membership in the two very different chambers of the bicameral U.S. Congress might affect pork-barrel politics. They test their hypotheses with data on the distribution of earmarks in the 2008 budget originating with senators and representatives respectively. They find that institutional and structural differences between the two houses do affect how earmarks are allocated. Seniority plays a greater role in obtaining earmarks in the Senate, leadership position is more important in the House. Electorally vulnerable members of the House receive more earmarks; senators close to reelection receive more earmarks in the Senate. Senators from the same state, even when of different parties, share influence in obtaining earmarks to a greater extent than members from the same state in the House. These differences are clearly attributable to differences between the two houses in their size, their election timetable, and their internal organization. In both houses members of the majority party have significant advantages over members of the minority.
James Coleman Battista offers a comparative analysis of 97 of the 99 state legislatures in the United States to explain the factors that determine why legislatures choose to compose their committees the way they do. He focuses on variation in committee informativeness, measured by the extent to which they are composed of members who are ideologically representative of the chamber as a whole. Battista demonstrates that chambers that have narrow partisan majorities have more informative committees. Other hypothesized influences on committee design, including the careerism of members, have only a small influence. Battista tests his conclusions about state legislative committees on the composition of committees in the U.S. House of Representatives and finds them applicable. He adds his voice to the suggestion of Squire and Hamm (2005) that comparative state legislative research can help us to understand institutional characteristics of Congress.
The two remaining articles in this issue deal with the increasingly important subject of Latino representation in U.S. legislatures. Latinos constitute a rapidly growing minority in the United States. While unsurprisingly there is a strong relationship between the proportion of Latinos in a state and their proportion in its legislature, Jason P. Casellas shows that there are other factors involved. On the basis of an examination of Latino representation in the 50 state legislatures in the United States between 1992 and 2004, he finds that the higher the rate of turnover of seats and the more liberal a state’s citizens, the higher the proportion of Latino legislators. He also shows that in states with a high proportion of Latinos in the population, term limits increase the proportion of Latino legislators and professionalization of the legislature decreases their proportion. Casellas tested these conclusions at the state level with data at the district level for seven states with the largest Latino populations and for the U.S. Congress. While the percentage of Latinos in a district is the strongest predictor of electing a Latino legislator, the influence of district composition varies across the states, being strongest in California and weakest in New York and in Congressional districts. The influence of district composition depends on rates of membership turnover and is weak where legislatures are highly professionalized. Casellas’s research adds to our knowledge of the factors that influence the entry of previously underrepresented groups into legislatures.
Once in Congress, Latino members face the challenge of effectively communicating both with their Latino and non-Latino constituents. Walter Wilson has examined the websites of all members of the 110th Congress to compare the communication styles of Latino and non-Latino representatives. He finds that there are surprisingly few differences and those that exist reflect the size of the Latino population in a district rather than the ethnicity of the member. Latino members do reach out to their constituents by including Latino perspectives on issues. However, their websites are not necessarily more accessible to Spanish speakers nor are Latino members more likely to express a position—on immigration, for example—that Latino constituents favor. Latino members of Congress seem anxious to be responsive to their entire constituency and not to appear partial to the support they have in their own ethnic group.
Legislative Research Center
Return to August 2009 Titles
Return to LSQ home page