Volume XXXVII, Number 3
August 2012



The field of legislative studies is a cumulative effort in which current understandings are continually challenged to assess their validity. Through this process, more refined understandings emerge. The articles in this issue of the Quarterly evidence this practice.

The rise of an incumbency advantage in U.S. congressional elections during the 1960s is well established. What has been controversial is whether the incumbency advantage (the extra vote margin a candidate receives by virtue of holding the office to which he or she seeks reelection) results in electoral safety (a higher probability of reelection). Arjun S. Wilkins shows that these two measures are not the same because either can shift without shifts in the other. Wilkins examines U.S. House elections from 1900 to 2006, taking into account both the observed reelection rate and the unobserved rate including incumbent retirements prompted by predicted defeats at the polls. Using a conditional probability framework, he calculates the lower and upper bounds on the probability of incumbent losses. The most striking result of this analysis is that the substantial increase in incumbent electoral safety took place in the late 1940s and early 1950s, much earlier than the rise in the incumbency advantage. This finding, which is well-supported by the analysis, raises a number of interesting questions for legislative scholars. Potential explanations may be found in changes in the party system, changes in voter behavior, changes in member behavior, and changes in the media environment.

Legislative term limits have been in place long enough in a sufficient number of U.S. states to give scholars the leverage needed to study their impact on legislative behavior and policy outputs. Their possible effect on bond ratings is investigated by Daniel C. Lewis. These ratings matter, of course, because they influence a state's access to the bond market and the interest rates they must pay on money that is borrowed. Lewis examines whether the inexperience of lawmakers in states with term limits (using a nuanced measure of those limits) leads to fiscal problems. He finds that state bond ratings are negatively associated with term limits, with important consequences for government finances. While inexperience is not the sole culprit for the bond rating drops, isolating its contribution is useful. It is likely that this article will be part of a vigorous debate over the policy effects of term limits.

In recent decades most U.S. states have moved away from using multimember districts in their legislatures. The decision to stop using them in North Carolina in 2002 is used by Justin H. Kirkland to conduct a quasi-experimental study of legislator cooperation. Examining bill cosponsorships, he finds that when two members from the same party share constituents, their cooperation level increases. When two members from different parties share constituents, they institute cooperation. When members are shifted from multimember districts to single-member districts, their cooperative behavior largely disappears. A cross-sectional analysis of 2007 data from Maryland, New Hampshire, Vermont, and West Virginia produces findings consistent with those from North Carolina. These results suggest that multimember districts may alter legislative behaviors in ways that are not fully appreciated by those making choices about apportionment schemes.

Multimember districts, this time in the National Congress of Chile, are also the focus of the article by Juan Pablo Couyoumdjian and John Benedict Londregan. They investigate whether government largess in the form of discretionary agricultural loans flows more freely to more competitive two-member districts than to less-competitive districts, as swing-voter models would predict. Analyzing data on 1997 elections for both the lower and upper houses, they track funds distributed by the Institute for Agricultural Development. Interestingly, they find solid evidence that loans appear to be targeted to districts in both chambers where the government is competitive for gaining one of the two seats at stake, but not to districts where the government is competitive only for the second seat. They also report that the relationship is stronger for the upper house than for the lower house. But, as the authors note, these relationships need to be reexamined using a more complete roster of government transfers.

Occasionally a book is published with sufficient promise to make a broad contribution to the field of legislative studies to merit a review in the Quarterly. M. Steven Fish and Matthew Kroenig's, The Handbook of National Legislatures, which gathers information on 32 variables for virtually every national assembly in the world, is such a volume. A constructive assessment of it is provided by Scott W. Desposato. While lauding the Handbook's remarkable data-collection effort, Desposato focuses his comments on the potential problems legislative scholars might face in making use of these measures. Notably, he worries about variables that are missing, such as data on party systems, and coding decisions the authors made, such as opting to combine formal and informal institutions in several instances. The tenor of Desposato's criticisms, however, is positive, directed toward helping researchers think about how they can make use of this valuable resource. In their response, Fish and Kroenig admit to some of the limitations identified by Desposato and defend their work against others. More important, they help guide readers to make full use of the information they collected, directing them to the text on each country as a way to identify and incorporate the data nuances that serious comparative work on legislatures demands.

—Peverill Squire
   Senior Editor


 Return to August 2012 Titles

 Return to LSQ home page