---

A Précis on Legislative Leadership

 

 

 

Theories of legislative leadership have greatly advanced since Barbara Sinclair's Majority Leadership in the U.S. House (1983). In a field dominated by two extreme theories of leadership (one where leadership was distributed amongst many, if not all, office holders, and one where leaders were viewed as generals marching the troops to battle), Sinclair presented what would become the new consensus view -- a view of legislative leadership in which leaders not only exercise real power and influence, but also act as agents of their parties. To Sinclair, party leaders are responsible for building winning coalitions and maintaining intra-party peace, and her work has provided a foundation for contemporary theories of legislative leadership and legislative organization.

            In order to highlight the ways in which theories of legislative leadership have evolved in the past two decades, we first provide an overview of how scholars conceptualized leadership prior to Sinclair's work. After discussing these early views, we briefly describe Sinclair's work and emphasize the ways in which it has influenced current theories of leadership. In particular, we discuss two related theories of partisan leadership that stemmed from Sinclair's work -- the "parties as firms" and "parties as teams" approaches.

 

Pre-Reform Views of Congressional Leadership

Early theories of legislative leadership viewed leaders as "generals" (Follett 1896; Chiu 1928; Stephenson 1930; Gosnell 1937; APSA 1950; Gwinn 1957; Patterson 1963; Cohen and Taylor 2000). According to this formulation, leaders controlled the allocation of scarce resources (nominations, campaign contributions, votes, etc.) and enforced discipline among backbenchers. Because leaders essentially had the authority to hire, fire, and promote individual legislators (often controlling their careers, job placement, and income), these scholars argued that leaders were able to ensure compliance from their followers. This theoretical framework was often applied in the context of European party organizations (Cox 1987), and also used to describe local American party machines at the turn of the twentieth century (see Mladenka 1980, Royko 1988, Erie 1988, Hichborn 1959, Riordon 1994, McCaffery 1993, Milligan 1948, Holli 1969, and Stone 1996).

Another early view of legislative leadership was the distributed leadership theory. In this characterization, the legislative process resembled a "town hall meeting," in which legislators came together to make decisions consensually (Smith 1989; Smith & Deering 1984). As far as leadership is concerned, one interpretation of this theory implies that different legislators at different times act as leaders, while yet another version suggests that different legislators will be leaders for different policy issues (Arnold 1990; Fenno 1973; Kingdon 1973; Schickler 2001). Despite these various interpretations, scholars have applied this theory to ministerial portfolio allocation in Western Europe (Laver & Shepsle 1996), to the functioning of the Japanese committee system (Ramseyer & Rosenbluth 1993; Rosenbluth 1989), and to the German committee system (Saalfeld 1997).

 

The Post-Reform View of Leadership

Following on the insights of Jones (1968), Sinclair argued that legislative leaders' role is to unify rank-and-file party members who have their own, often conflicting, policy goals. In order to maximize party and individual successes, leaders work to keep "peace in the family" and build winning coalitions (Sinclair 1983). As Sinclair noted, these two goals can reinforce each other, but it is not unusual for them to conflict, making leadership a delicate balancing act. In order to overcome such difficulties, party leaders provide services to members, structure the choices that House members face, and attempt to include various party factions in the legislative process.

Sinclair's theory of legislative leadership also recognized that leaders face external constraints -- such as legislative rules and individual members' desire to be reelected -- that they must overcome in order to meet the demands of their party. Because of these constraints, legislative leaders must take individual members' preferences into account, while continuing to pursue the policy objectives of their party. Indeed, if leaders fail to unite their party, the party may factionalize and become unable to produce favorable policy outcomes. An additional consequence is that party leaders may be ousted and replaced, which forces leaders to be responsive to the wishes of the party. In short, Sinclair's theory provided a framework for thinking about the critical interdependence between party leaders and members: leaders must be responsive to party members and party members must rely on their leaders to deliver favorable policy outcomes (Sinclair 1981, 1983, 1989).

As a result of Sinclair's insights, a new perspective on party leadership emerged -- one in which party leaders are agents, and the party as a whole acts as the principal. Sinclair (1983) emphasized that parties delegate authority to leaders, who are then expected to solve the collective action problems that arise among the rank-and-file. This analysis has pushed scholars to think about legislative leadership in terms of classic principal-agent relationships and to apply insights from the industrial organization literature to political parties (c.f., Alchian and Demestz 1972).       

The basic analogy that leaders are agents of their parties, wielding delegated powers in the interests of their rank and file, raises many additional questions that later research has addressed (see, for example, Westefield 1974, Nelson 1977, Cooper and Brady 1981, Cox 1987, Stewart 1989, Maltzman and Smith 1994, Binder 1997, Gamm & Smith 2001). Consider just four of these post-Sinclair research issues. First, there is the question of when parties will delegate more and less power to their leaders. Aldrich and Rohde's (1998, 2000) theory of conditional party government offers an answer:  parties whose members are more in agreement on policy goals, and face an opposition party more sharply in disagreement, delegate more power to their central leaders. This hypothesis has been tested by Rohde (1991), who studies the congressional reforms of the early 1970s. This view has also been challenged by Schickler (2000) and Schickler and Rich (1997), who argue that changes in congressional rules do not correlate with changes in majority party homogeneity, arguing instead that it appears to be changes in the location of the floor median that drive changes in House rules. Interestingly, this debate has been unscathed by the vast literature on agency (e.g., Arrow 1971; Alchian and Demsetz 1972; Jensen and Meckling 1976; Rothschild and Stiglitz 1976; Holmstrom 1979; Myerson 1979; Teece 1980; Klein and Leffler 1981; Crawford and Sobel 1982; Fama and Jensen 1983; Guesnerie and Laffont 1984; Rogerson 1985; Hart and Holmstrom 1987; Tirole 1988; Holmstrom and Milgrom 1991) and delegation (see Kiewiet and McCubbins 1991, Lupia and McCubbins 1998, and McCubbins 2000 for surveys).

Second, there is the question of how leaders use their delegated powers.  In Aldrich and Rohde's initial statement of the conditional party government thesis, leaders were pictured as using their powers to increase the degree to which members of the majority party voted together on the floor (what they called party cohesion), thereby enabling the party to pull actual outcomes away from those preferred by the median legislator.  Challenges to this view have been registered, for example, by Krehbiel (1999, 2000), who asks why the median legislator would ever agree to a non-median outcome, and goes on to question the evidence adduced by Aldrich and Rohde for the connection between the conditions for conditional party government and the effects of party government (that is increased cohesion) (see Smith 2000 for an excellent review of this debate).

An alternative view of how leaders use their delegated powers is offered in the cartel agenda model (Cox and McCubbins 1993, 1994, 2002, 2003).  In this model, the power delegated to leaders is primarily the power to set the agenda.  While leaders may influence members' floor voting behavior, their main objective is to block certain issues from ever being considered on the floor, while ensuring that others are brought before the House.  From this perspective, even if the majority party can never secure a non-median outcome on the particular issue dimensions that advance to a final passage vote, it can still have a decisive influence on the overall direction of policy movement, simply by deciding which issues will be considered.  Challenges to this view have been registered by, e.g., Krehbiel and Meirowitz (2002), who suggest that the motion to recommit vitiates the majority party's advantages in setting the agenda, and by Schickler (2000, 2001) who argues that many of the changes in House rules seem to favor the minority party rather than the majority party.

Third, there is the question of how widely powers are delegated.  Should we think of a party's leadership as highly concentrated -- just the leader and his/her lieutenants?  Or should we think of it as widely dispersed, with key committee chairs, caucus leaders, whips, and others getting consequentially into the act?  Here too there are a variety of positions that remind us of the original contrast between the party leader as general versus the dispersed leadership model. Early versions of the conditional party government model seem to imply an analogy to the former view of parties, while the cartel in the cartel agenda model of party leadership argues that leadership is a collegium or team, taking a more dispersed view of party leadership.

Fourth, there is the question of the ultimate goal sought by party leaders.  Some models take a more intramural tact, focusing on the immediate policy gains to be secured by better-coordinated action (e.g., Rohde 1991).  Others place more emphasis on extramural goalsÑwinning the next election and the value of the party image in that endeavor (e.g., Cox and McCubbins 1993, 2003; Evans 2001, Doring 1995, Smith and Remington 2001).

There are many other issues that the party leader as agent metaphor raises. How are the agents kept in line (c.f., Kiewiet and McCubbins 1991 who discuss procedural arrangements and other "agency controls"; or McCubbins 2000 who discusses how the the dispersion of leadership creates checks and balances within the legislature that engender agency control)?  How are the rank and file kept in line (c.f., Cox and McCubbins, 1993, especially Chapter 7)?  How does the leadership actually define the collective good and motivate individuals to act for that collective good (Jones 1968, Sinclair 1983, Cooper and Brady 1981; Rohde 1991). How does the organization of the legislature serve to define and control these agency relationships (Kiewiet and McCubbins 1991; Cox and McCubbins 1993; Sala 2002)?

How we think about legislative leadership is part of how we think about legislative organization. This debate informs the debate about the legislative process, public law, and statutory interpretation. In turn, these debates inform how we think about our democracy. We have surveyed the major issues in the literature, emphasizing the analogies used in thinking about legislative organization and leadership. The analog of party leaders as agents, championed by Sinclair and others (Jones 1968; Cooper and Brady 1981; Shepsle 1985; Rohde 1991) stands out as one of the most important advances of the last half-century. Indeed, it is plain to see from our précis that Sinclair's work changed the terms of the debate in the literature on legislative organization and we will see the effects of this change in the debate for decades to come.

 

References

---

Gary W. Cox is Professor of Political Science at University of California, San Diego. In addition to numerous articles in the areas of legislative and electoral politics, Cox is the author of Making Votes Count: Strategic Coordination in the World's Electoral Systems (Cambridge University Press, 1997), which won for him the Woodrow Wilson Foundation Award for best book in political science, the Luebbert Prize for best book on comparative politics, and the Best Book in Political Economy Award in 1998. His dissertation, The Efficient Secret: The Cabinet and the Development of Political Parties in Victorian England, won the Samuel H. Beer dissertation prize in 1983 and was later published in a book (Cambridge University Press, 1987). Cox is the coauthor of Legislative Leviathan: Party Government in the House (University of California Press, 1993), which won the Richard F. Fenno Prize for best book in legislative studies; and coauthor of Elbridge Gerry's Salamander: The Electoral Consequences of the Reapportionment Revolution (Cambridge University Press, 2002). A former Guggenheim Fellow, Cox was elected to the American Academy of Arts and Sciences in 1996. His email address is gcox@ucsd.edu.

Mathew D. McCubbins is Professor of Political Science and holds the Chancellor's Associates Endowed Chair at University of California, San Diego. He is coauthor of Stealing the Initiative: How State Government Responds to Direct Democracy (Prentiss Hall, 2001); Legislative  Leviathan: Party Government in the House (University of California Press, 1993), which won the Richard F. Fenno Prize for best book in legislative studies; and The Logic of Delegation: Congressional Parties and the Appropriations Process (University of Chicago Press, 1991), which won APSA's Gladys M. Kammerer Award for best publication on U.S. national policy. He is editor of Under the Watchful Eye: Managing Presidential Campaigns in the Television Era (CQ Press, 1992) and co-editor of The Origins of Liberty: Political and Economic Liberalization in the Modern World (Princeton University Press, 1997); Elements of Reason: Cognition, Choice, and the Bounds of Rationality (Cambridge University Press, 2000); and Party, Process, and Political Change in Congress: New Perspectives on the History of Congress (Stanford University Press, 2002). His email address is mmccubbins@ucsd.edu.

 

---

Table of Contents| |Editor's Introduction| |Special Orders| |A Tribute to Julian Rothbaum| 
| 2003 Rothbaum Lecture| |News|  |Announcements| |Other Issues of Extensions |

---

| HOME | | Contact Us |
| Teaching & Research | | Public Outreach | | Congressional Archives | | Graduate Fellowship |

---