Volume XXXVI, Number 2
May 2011




Scholars have learned a great deal about the rules and procedures employed by legislatures in making policy decisions, but we know very little about the kinds of information lawmakers use in reaching their choices. In this issue of the Quarterly, Kevin M. Esterling combines data gathered from the transcripts of 72 congressional hearings on the Medicare program from 1990 to 2003 with data from a survey of 308 witnesses who testified in those hearings, to advance our understanding of the content of the information generated by the hearings process. Specifically, Esterling examines the conditions under which witnesses present falsifiable claims rather than resorting to nonfalsifiable rationales. Employing a latent variable model, he finds that witnesses frame their arguments to be persuasive given the political environment in which a particular hearing is held. When the subject of a hearing is thought to be only moderately contentious, testimony tends to emphasize falsifiable arguments. In contrast, when the topic is either non-controversial or highly contentious, witnesses lean more on nonfalsifiable claims. Interestingly, Senate hearings appear to produce more evidence-based testimony than House hearings. Easterling embeds his findings in the larger debate over deliberative democratic theory.

Approximately a third or more of state legislative elections are uncontested, meaning that one major party effectively fails to field a candidate. In recent years, some states have even witnessed uncontested rates of 75%. This lack of competition raises a number of interesting questions. The effect of uncontested elections on the behavior of incumbent lawmakers is examined by David M. Konisky and Michiko Ueda. Using data on bill introductions and voting participation rates in 1999–2000, they find that state legislators who enjoyed an uncontested race in their previous election missed up to 2% more floor votes and introduced 10% fewer bills than their counterparts who faced competition. These results are impressively consistent across chambers, levels of legislative professionalism, and existence of term limits. Although there is no difference in bill enactment rates between lawmakers who faced competition and those who did not, it may be that facing a challenger heightens an incumbent's sense of self-preservation, triggering a need to engage in activities that visibly demonstrate their commitment to their constituents.

Roughly 40% of the world's national assemblies are bicameral. In a substantial proportion of these two-chamber bodies the upper house is thought to be, at best, a junior partner in the legislative process, with considerably more power being vested in the lower house. The role played by upper houses in Belgium, Germany, Ireland, and the United Kingdom is examined by David Fisk. Fisk seeks to explain the roughly one in one hundred votes on which the government loses in the upper house. Analyzing an extensive data set on all parliamentary votes in the four national assemblies between 1993 and 2003, he finds that governments are more likely to be defeated when the opposition controls the upper chamber. Governments also experience more difficulty with upper chambers when political parties have less control over the nomination of candidates for the upper house. Intriguingly, the analysis indicates that when the upper chamber enjoys strong veto powers, the government has considerable incentive to anticipate difficulties in getting measures through, even when the upper house is controlled by a friendly majority. As Fisk notes, it will be important to test the dynamics he uncovers in presidential systems.

In American legislative elections, incumbents almost always win. The existence of an incumbency advantage is well documented in congressional and state legislative elections, but whether it exists at the local government level is a question that has received little attention. Jessica Trounstine remedies this omission by studying city council elections in Austin, Dallas, San Antonio, and San Jose from 1915 to 1985. Employing a regression discontinuity design, Trounstine finds that city council incumbents are 39 percentage points more likely to run and 32 percentage points more likely to win because they have served a term in office—figures that are only slightly lower than for U.S. House incumbents. This demonstrates the existence of an incumbency advantage at the local level. What is particularly interesting about this finding is that it implies that a resource-based argument for an incumbency advantage—incumbents win because of staff and constituent newsletters and the like—may be weaker than conventional wisdom holds because city councilors typically enjoy fewer, if any, such resources.

In every election, some relatively small proportion of congressional voters defects from their party's candidate to vote for the other party's candidate. Over the years, such defectors have received considerable scholarly scrutiny, with explanations for their deviations typically pinned on the power of incumbency and the influence of presidential approval. In a new twist, the impact of policy issues on the decision to defect is documented by Paul S. Herrnson and James M. Curry. Herrnson and Curry use data from the 2006 Cooperative Congressional Elections Survey to demonstrate that party ownership of salient issues can both push and pull voters, causing them to deviate from their normal behavioral patterns. Controlling for a host of theoretically important variables, the authors find that Republican identifiers who thought health care policy was important were considerably more likely to vote for the Democratic candidate. In the other direction, Democratic identifiers who thought terrorism and national security were the most significant issues were much more likely to defect to the Republican candidate. These results highlight the important role issues can play, at least under specific circumstances, in the congressional vote decision. The authors also detail the practical implications of their findings for candidates and campaigns.

Every year members of Congress announce that they intend to leave the institution. Such announcements almost always generate immediate speculation about the political implications of their departures, particularly whether the incumbent's party can retain the seat in the next election. In an analysis of open-seat contests for the U.S. House between 1996 and 2008, C. Douglas Swearingen and Walt Jatkowski III test whether the timing of such announcements affects the sort of candidates who enter the race, the amount of money the candidates raise, and most importantly, the outcome of the election. They find that the number of days between the departure announcement and the election does not exert a statistically significant effect on the quality of the candidates who run. But the amount of money the candidates can raise for the open-seat contest increases with the number of days between the announcement and the election, with the incumbent party candidate reaping the greater benefit. Given this relationship, it may not be a great surprise that the incumbent party's prospects for retaining the seat increase along with the number of days between an incumbent's announced decision to vacate and the election. Interestingly, however, there is no relationship with the margin of victory. The practical implications of these findings are straightforward: To increase its prospects for retaining a seat, an incumbent party needs to encourage its departing incumbents to announce their decisions earlier rather than later.       

                                                                                                                                                               —Peverill Squire
                                           Senior Editor


 Return to May 2011 Titles

 Return to LSQ home page