In this volume, Gary W. Cox and Mathew D. McCubbins elaborate on the “cartel” theory of parties in the U.S. Congress that was first set forth in their earlier book, Legislative Leviathan (1993), and present a substantial body of new evidence in support of this theory. The central thesis of the new book is that a form of party government has been present continuously in the House of Representatives since the adoption of the Reed Rules in the 1890s.

In contrast to analysts who view party influence in terms of the ability to maintain unity in floor voting and enact a party program, Cox and McCubbins argue that party government and party influence in Congress primarily involve negative agenda power—the capacity of the majority party to keep off the legislative agenda those issues that are opposed by a majority of its members and that would evoke deep splits if allowed to come to a vote. In keeping with the earlier formulation of their theory, legislative parties are said to exist to protect party members' collective reputation with the voters and maintain (or win) majority status. The majority party acts like a cartel by seizing the most important offices in the legislature and distributing them among its “senior partners,” with the understanding that the procedural powers of these offices will not be used to advance legislation opposed by a majority of their party. When the majority party is unified over policy, it may delegate more authority to leaders to enforce party discipline and advance new legislation, but the exercise of this positive agenda power is the “superstructure” of party government, not its “bedrock.” Efforts at enforcing party discipline are said to be limited to “corralling a few votes on the margin” and are of secondary importance to controlling which issues get voted on (pp. 4–5). Thus, for Cox and McCubbins, the appropriate test of party influence in Congress is not how often party leaders can get members to vote in support of party measures, or whether newly passed legislation reflects the preferences of the median member of the majority party, but whether the majority party succeeds in blocking consideration of legislation most members of the party oppose.

The authors derive a series of implications from their theory and theories positing nonpartisan agenda-setting processes and marshal a formidable body of evidence to show that the majority party has consistently exercised negative agenda power in the House since the adoption of the Reed Rules in the 1890s. Drawing on roll call data reaching back to the 1870s, they show that after adoption of the Reed Rules, the majority party in the House has rarely been “rolled” (a majority of the majority party outvoted) on rules changes, procedural votes, or votes on final passage. They also develop a measure indicating the proportion of bills that move policy in the ideological direction favored by the majority party, and they show: 1) that a sustained increase in that proportion occurs after the 1890s, and 2) that the variation in this measure over the period from the 54th to 105th Congresses is explained better by spatial models, in which the majority party controls the agenda, than by nonpartisan models. The authors also examine the important question of how independent of the majority party the House Rules Committee became during the era of the conservative coalition (1937–60), concluding that the committee blocked liberal legislation but rarely advanced measures opposed by the majority of House Democrats. Here, Cox and McCubbins encounter the limitations of relying primarily on evidence from roll call votes; recent research by Eric Schickler and Kathryn Pearson has uncovered numerous additional cases during this period in which the majority party was rolled on issues on which recorded votes were not taken. Whether the form of party government posited by this theory was not at least partially displaced during these years remains an open question. The final section of the book briefly contrasts the cartel agenda theory with John Aldrich and David Rohde's conditional party government theory and presents some suggestive evidence in support of Cox and McCubbins's claim that increased majority party homogeneity tilts party government toward greater positive agenda control.

The authors make a strong case for the central importance of the majority party in setting the agenda in the modern Congress. But precisely how the use of negative agenda power constitutes the “responsible party government” referred to in the title of the book is a question that could have received more attention. If the bedrock of party government is a form of agenda control in which the majority party prevents consideration of proposals that would result in the passage of legislation reflecting the preferences of the median House member, party government would seem to consist primarily of the prevention of policy changes preferred by legislative (and possibly popular?) majorities. The sense in which this is responsible government is unclear.

Setting the Agenda represents state-of-the-art empirical political science in the rational choice institutionalist mode, exhibiting both the strengths and some of limitations of that approach. Cox and McCubbins's theoretical arguments are explicit and precise, and the implications of those arguments and competing explanations are subjected to careful empirical tests. However, by working within a spatial theory framework in which legislators are assumed to have fixed preferences, this approach inevitably misses important aspects of how parties and leaders influence legislative politics. As R. Douglas Arnold, Joseph Bessette, Joseph Cooper, C. Lawrence Evans, Richard Fenno, and other scholars have argued, leaders influence what happens in Congress not only by controlling the agenda and using selective incentives to corral votes but also by taking advantage of opportunities that arise to shape legislators' preferences. Still, Cox and McCubbins make a powerful case for viewing agenda control as a persistent and fundamental dimension of party influence in the modern U.S. Congress. Even readers who may not be persuaded that this mode of analysis captures the full range of contributions parties and leaders make to congressional politics will benefit from a close reading of this study.