The rise of the power of the Tea Party in the House of Representatives has raised numerous questions about legislative organization and party leadership. Using data from reforms in the 1970s that weakened the seniority system in the House, Andrew B. Hall explores one obstacle to centralizing power towards the party leadership and away from senior committee members. He argues that party members have electoral incentives to centralize power when they share an ideological vision, but that the senior members who lose power under this arrangement are similarly motivated to block these reforms.
Recent controversies over the filibuster, committee assignments, and the ability of the Speaker of the House to control renegade members have brought issues of legislative organization and procedural rules into the public’s mind, if only briefly. In general, the procedural rules and norms that the majority party adopts play an important role in determining what gets done (or does not) get done in the legislature. By adjusting these rules, the majority party can influence legislative outcomes and, in so doing, attempt to bolster its standing among voters.
The Conditional Party Government (CPG) theory of John Aldrich and David Rohde explores how the majority party centralizes the party’s ability to set the policy agenda (its so-called ‘agenda power’) in times when its members agree on policy goals. In so doing, the party is thought to enhance its brand by operating more effectively, sending a stronger signal to voters. While the benefits of this centralization are uniform, since every majority-party member enjoys the strengthened party brand, its costs can create an obstacle to its implementation. When centralization occurs, the costs are borne unevenly by the party’s most senior members because they possess a disproportionate share of agenda power when the House is decentralized, i.e., when the committee system and not party leadership is central. As a result, senior members are unlikely to hand their agenda power over to the party willingly. Any reform intended to centralize agenda power must overcome this opposition from entrenched senior members.
To test this theory, we looked at a set of reforms in the 1970s that fundamentally weakened the seniority system in the House and empowered party leadership, as represented by the Speaker and Majority Leader. Carl Albert, speaker of the House from 1971-1977, referring to the passage of these reforms, declared that “the seniority system…died that day.” These reforms roughly partitioned the House of Representatives into a “Weak Party Era” from 1945-1976, characterized by decentralized agenda power and a committee system organized by seniority, and a “Strong Party Era” from 1977-present, characterized by centralized agenda power, strong party leadership, and weak committees.
Our theory has two core predictions. First, the centralization of agenda power should lead to a mean increase in majority-party incumbent vote shares, representing the enhanced and improved party label. Second, centralization should produce a variance contraction in majority-party incumbent vote shares, i.e. the vote shares become more alike. This contraction occurs because incumbents previously distinguished from each other through their seniority and committee service are now more interchangeable party members, closer to being “foot-soldiers” of the party leadership. Figure 1 below plots the distribution of majority-party incumbent vote share across the two eras. As it shows, the strong-party era distribution is shifted to the right, relative to the weak party, and is also more concentrated. Majority-party incumbents receive a higher average vote share after agenda power is centralized, but are also more similar to each other.
Figure 1 – Distribution of Majority-Party Incumbent Vote Share by Era
By the same logic, we should also observe the value of seniority weakening in the strong-party era. Because members no longer accrue agenda power by ascending committee ranks in a centralized legislature, they are not necessarily able to bring the same value home to voters (who need not be aware of their incumbent’s seniority to enjoy the benefits it brings them). Figures 2 and 3 show how this “return to seniority” falls in the strong-party era. Figure 2 summarizes the association between seniority and vote share for majority-party incumbents in the weak-party era. There is a strong upward slope across levels of seniority. In Figure 3, on the other hand, we see that the relationship is relatively flat.
Figure 2 – Seniority and Vote share – Weak Party Era (1945-1976)
Figure 3 – Seniority and Vote share – Strong Party Era (1977- present)
We also conducted a variety of econometric specifications that produce the same conclusions. Using these estimates, we examined the implications of centralization for members of various levels of seniority. At low levels of seniority, the mean increase in majority-party incumbent vote share more than cancels out the estimated loss in vote share from the reduced value of seniority. More junior members thus have incentives to support the reform. But for senior members, the story is different. For members with 10 or more terms in a senior position, the loss of agenda power is more damaging than the resulting party “bump” is beneficial. These members are likely to oppose centralization.
We also looked for alternative explanations and marshalled evidence that other factors are unlikely to explain the effects we find. For one, minority-party incumbents see no such decrease in the return to seniority, nor do they receive a similar mean-increase in the strong-party era. We have also applied the analysis to state legislatures and the U.S. Senate as “placebo tests,” since these chambers did not undergo the same reform. Again, we did not find the same set of results. Finally, we tested and have shown the robustness of the finding to differing definitions for the exact year cutoff separating the strong and weak party eras.
Although the bulk of our work has centered on this empirical exercise, we believe the main contribution of the research is theoretical. Party members have incentives to centralize power when they share an ideological vision, but giving agenda power to the party leadership means taking it away from somewhere else. In the case of the House, this “somewhere else” is senior members. They are unlikely to go along with this change willingly.
We suspect that the disproportionate amount of power these senior members wielded in the weak-party era helps explain the long delay in implementing party-centric reform in the House, which did not occur until the 1970s even though many called for it as early as the 1950s. This obstacle to reform means that the “condition” of Conditional Party Government (that members share similar policy goals) is not a sufficient condition for centralization. In order for members to centralize power successfully, they must overcome the opposition of senior members. Perhaps this is why we see such reforms occurring after “wave” elections in which large cohorts of junior members are swept into office, like in the post-Watergate elections in 1974 and the Republican take-over of the House in 1994.
Many—perhaps most—studies of legislative politics focus either on the internal dynamics of the legislature or on the electoral environment outside the legislature. But the “electoral connection” that David Mayhew described implies that the decisions over the internal organization of the legislature are made with an eye towards their electoral consequences. Likewise, election results are in part a function of these organizational decisions. As such, there is a place for more research that explicitly links conditions inside the legislature to those in the electorate.
This article is based on the paper, “The Changing Value of Seniority in the U.S. House: Conditional Party Government Revised,” by Andrew B. Hall and Ken B. Shepsle, which was published online in The Journal of Politics in October.
Note: This article gives the views of the authors, and not the position of USApp– American Politics and Policy, nor of the London School of Economics.
Shortened URL for this post: http://bit.ly/19dqfVN
Andrew B. Hall – Harvard University
Andrew B. Hall is a Ph.D. candidate in the Department of Government at Harvard University and an affiliate of the Institute for Quantitative Social Science. His applied research studies the links between the design of American political institutions and the behavior of elected officials. For more information, please see andrewbenjaminhall.com. You can follow Andrew on Twitter @andrewbhall.