The Ideological Positioning of Legislative Leaders

The classic academic debates on the ideological positioning of legislative leaders highlighted the contextual underpinnings of a party's leadership choice. Comparing legislatures with vastly different partisan contexts, Patterson (1963) found evidence that ideologically extreme members were selected to lead in highly partisan legislatures (e.g., the Massachusetts and Wisconsin legislatures) while ideological middlemen were selected in less partisan legislative settings like the Oklahoma House and the U.S. Congress.⁴

Even Truman's (1959) classic formulation of the “middleman hypothesis” held that the emergence of middleman congressional leaders was due, in part, to “the depth and persistence of the cleavages in both parties” (205).

Considering the selection of a legislative leader as a strategic choice party members make about how best to advance collective goals suggests the importance of broader contextual factors that are likely to vary by legislative party and over time. In some contexts where deep intra-party divisions predominate and where relatively large, but internally divided, majorities control a legislative chamber (like the Oklahoma House or the “Textbook Congress”), party success hinges on the ability of leaders to bridge intra-party gaps; the need to straddle the divide between the party's ideological camps and to build broad—sometimes bipartisan—coalitions counsels choosing middlemen. By contrast, in contexts where intra-party divisions are less prominent and partisanship is high (like Patterson's more partisan legislatures and, arguably, the contemporary Congress), the ability to advance the more or less unified goals of party followers takes center stage, making legislative parties more likely to select leaders closer to the party's edge. These leaders can then sharpen inter-party divides, make the party's case to friendly interest groups, the press, and the public, and negotiate inter-party differences by “counter-punching” across the partisan aisle, all tasks that are arguably better performed by more extreme partisans than by moderates (King and Zeckhauser 2002).⁵

Do contemporary parties, particularly those in the 110^th Congress, seek “middleman” brokers or extreme counter-punchers? Contrary to the middleman expectation, in recent years, the legislative parties frequently have selected leaders from their more ideological wings. Table 1 reports the ideological positioning of top House and Senate leaders from the 107^th to the 110^th Congresses as measured by the percent of a leader's party's membership to his or her ideological extreme.⁶

Whereas House leaders like J. Dennis Hastert (R-IL), John Boehner (R-OH), Jim Clyburn (D-SC), and Rahm Emanuel (D-IL) and Senate leaders like Tom Daschle (D-SD), Harry Reid (D-NV), and Bill Frist (R-TN) were selected from near the middle of their parties, it is just as likely that party leaders come from the ideological extremes. Only little more than one-quarter of the Democratic Caucus was more liberal than Nancy Pelosi at the time she was selected speaker just as other House leaders like Tom DeLay (R-TX), Dick Armey (R-TX), and Dick Gephardt (D-MO) represented ideological leadership selections. Similarly, on the Senate side, the 110^th's top Senate Republicans Mitch McConnell (R-KY) and Trent Lott (R-MS) are both from the party's more conservative wing; and, some Senate leaders like Don Nickles (R-OK), John Kyl (R-AZ), and Dick Durbin (D-IL) rank as some of the most ideologically “extreme” members in their chamber's parties. If Speaker Pelosi's ideological positioning stands at odds with the “middleman” expectation, she is not unique in this regard.

Table 1 points up a third category of leaders: those selected well on the moderate side of their parties' caucus or conference. This, as Peabody (1976) noted, is an unusual occurrence: “With only a rare exception or two, a potential candidate cannot deviate far from the mainstream of his party's ideological orientation if he hopes to become a leader” (470). While recent party officials, like Democratic Caucus Chair Martin Frost (D-TX) and Republican Conference Chair Deborah Pryce (R-OH), were in the moderate portion of their parties when selected, the only true moderate among the top party leaders of either party in the 110^th is Democratic Floor Leader Steny Hoyer (D-MD), who is positioned with 68.8% of the caucus to his left.⁷

Placing recent leaders (those from the 107^th to 110^th Congress) in the middleman, extremity, and chamber moderate categories, the legislative parties have been just as likely to have “extremity” leaders as they are to have “middlemen.” There are exactly 11 examples of each on the House side with six chamber moderates and 12 examples of each on the Senate side with no chamber moderates. Thus, if the evidence does not point to a complete displacement of “middlemen” selections with “extremity” choices, it does suggest that would-be leaders from both parties' extreme wings are increasingly likely to win intra-party contests and to serve as top leaders. But these examples of “extremity” leadership selection and new developments are not so much disconfirmation of the prior middleman hypothesis as they are reminders of the contextual underpinnings of the leadership choices that parties make and how dramatically the party context in Congress has changed in recent decades. Republican gains in the South and the diminished influence of Northern moderates have made the Republican Party more conservative and the loss of Southern conservatives has made the Democratic party more dominated by Northern (and Western) liberals. Those changes have made the United States Congress both more partisan and more ideological, contributing to the intra-party preference homogeneity of both parties and the polarization between the parties (Rohde 1991; Sinclair 2006). While the U.S. Congress in the Textbook Era—overwhelmingly Democratic with deep internal divisions in the majority party—led to the selection of moderate leaders, increasing partisanship has made the contemporary Congress more like the partisan state legislatures that Patterson (1963) found selected “extreme” and “highly partisan” leaders precisely because “the two parties were relatively cohesive” (405). As congressional Democrats and Republicans became more internally homogeneous and polarized, the need for “middlemen” lessened while the need for “counter-punching” in the inter-party context increased. However, the inherent organizational weakness of American parties and, perhaps, the narrowness of contemporary House and Senate majorities may keep some middlemen on the premises, albeit in diminished numbers.

It is important to note that extremity selections would yield even deeper divisions between the two parties' top leaders given this increased inter-party polarization and intra-party homogeneity; indeed, even the respective parties' middlemen would be more ideological and divided than in the past. Figure 1 displays data on the ideological distances between Democratic and Republican floor leaders in the House and Senate, respectively.⁸

There were relatively great distances on the House side in the first four decades of the twentieth century. When FDR's New Deal coalition came unglued in the 1938 election and the Conservative Coalition gained its first major victory, House Democrats were obliged to create a counter-coalition—the Austin-Boston alliance—led by Sam Rayburn and John McCormack who could straddle the party's divisions. It was that alliance (especially with Joe Martin on the Republican side) that contributed to the lessening of the inter-party leadership differences in the House.⁹ Remaining low throughout the Textbook Era and into the 1970s, inter-party differences initially rose again with the 97^th Congress, the first Congress of the Ronald Reagan presidency, as Republicans replaced floor leader John Rhodes (R-AZ) with the considerably more conservative Bob Michel (R-IL), and these differences continued to rise as Democratic leader Jim Wright's leftward drift that began with his 1976 victory over Burton continued through the 99^th Congress (just before succeeding to the speakership unopposed).¹⁰ While Dick Gephardt's (D-MO) 1989 election as majority leader pushed the two parties even further apart, the largest differences emerged after the Republican takeover pitted an even more liberal Gephardt against Republican Majority Leader Dick Armey of Texas, whose position on NOMINATE's 1^st dimension made him the most conservative floor leader since Bertrand Snell (R-NY) was minority leader in 1938. Somewhat surprisingly, the pitting of Nancy Pelosi and Tom DeLay in the 108^th and 109^th Congresses was (albeit only slightly) less polarized than the Gephardt-Armey match-up, perhaps suggesting that leaders' tactics as well as their ideologies help set the partisan tone of a given Congress.

The ideological distances between Senate floor leaders follow a similar but not identical pattern. Relatively high levels of inter-party disagreement from 1920 to the mid-1930s gave way to a drop after the 2^nd New Deal with a spike both at the end of the 1940s (though briefly) and a steep decline in the closely balanced Senate of the 1950s. Despite a period of increased inter-party difference in the 1960s—the pairing of Democratic leader Mike Mansfield (D-MT) and Republican leader Everett Dirksen (R-IL)—the low levels of inter-party difference in the 1970s House are mirrored in the Senate, particularly when moderate-liberal Republican Hugh Scott (R-PA) succeeded Dirksen as Republican leader and continued as moderates Robert Byrd (D-WV) and Howard Baker (R-TN) became their parties' floor leaders from the 95^th to the 98^th Congresses, 1977–1985. If choosing middlemen and cross-party cooperation was predominant in Senate leadership in the Textbook Era and especially in the 1970s, that, too, has given way to more polarization in the post-Reagan era. And, since the late 1980s, there has been a steady increase in the polarization of the two parties' Senate floor leaders reaching, in the current match-up, its highest point in over half a century.

Finally, the broad similarities between the House and Senate represented in Figure 1 are noteworthy in and of themselves. Despite the fact that the four congressional parties are independent of one another and have developed different internal selection patterns, the ideological distances between both parties' floor leaders are, at least in broad contour, similar in the House and Senate following (and perhaps doing their own part to reinforce) the same broad partisan directions of American national politics over time.¹¹

Partisan Patterns of Competition and Conflict

Although stability (that is, the tendency for leaders to remain in their positions) is the predominant characteristic of congressional leadership selection, there are occasional revolts against sitting leaders and, when vacancies occur, legislative parties exhibit varying degrees of consensus and conflict to fill their posts. Prior studies of such patterns of competition and conflict in leadership races have differed as to whether there were patterns across the parties and chambers or party-specific factors that best explained leadership choice. Whereas Canon (1989), for example, found “institutionalizing” patterns across parties and chambers, Nelson (1977, 920, 932) warned against neglecting “the impact of party identity” and “the party variable” in predicting patterns of competition and conflict in leadership selection. Are there institutional patterns across parties and chambers or are there distinct partisan patterns reflecting the different calculations associated with majority and minority status and the varying institutional contexts and strategic choices faced by Democrats and Republicans in both chambers?¹²

Again viewing leadership selection as a strategic choice parties make, we would expect the four legislative parties to exhibit different tendencies toward conflict and consensus in leadership selection. Nelson (1977), for example, argued that House Democrats had established orderly succession patterns precisely because of the potential for their deep internal divisions to be exacerbated by open conflict, and he predicted that “As the Democratic party becomes more philosophically united, and as the ancient regional conflicts recede, it is likely that the Democratic leadership will no longer need their controlled succession system and that the membership will no longer tolerate it” (939). By the 1980s, Nelson's predictions were already coming true as the Democrats' patterns of succession weakened and conflict was on the rise. And, intra-party competition in Democratic leadership contests skyrocketed during their 12 years in the minority, though conflict was diminished somewhat in Republican ranks. If Canon (1989) revealed emerging patterns across chambers and parties in the 1980s, extending analysis into the 1990s revealed that the two parties were “moving in different directions with Republicans slightly mitigating their tendencies toward conflict and Democrats becoming more prone to intra-party conflict over time” (Harris 2006, 200).

Organizing for the 110^th provides new support for these observed differences between Democrats and Republicans, majorities and minorities (see Table 2). If Nelson's predictions of increased Democratic conflict in the 1970s and 1980s were correct and that conflict was exacerbated during their 12 years in the minority, the return to majority status in 2006 seems to agree with Democrats. On the House side, Democrats demonstrated far more stability than conflict in their late 2006 leadership selections. Deftly clearing the way for successful DCCC Chairman Rahm Emanuel (D-IL) to become the consensus choice for caucus chairman (even though Representative John Larson (D-CT) had been elected caucus vice-chairman earlier in the year) and watching as Jim Clyburn (D-SC) succeeded without opposition from caucus chair to the party's whip post, House Democrats also advanced Minority Leader Pelosi to the speakership without even the hint of intra-party challenge, thus returning Democrats to that pattern of succession from floor leader to speaker that had existed in the party for most of the twentieth century.¹³

For their part, Senate Democrats exhibited a great deal of stability: Harry Reid (D-NV) and Richard Durbin (D-IL) continued as leader and whip while Patty Murray (D-WA) emerged as a consensus choice to replace Barbara Mikulski (D-MD) as Conference secretary. In all, Senate Democrats have not had a contested race since Tom Daschle bested Christopher Dodd (CT) by one vote after George Mitchell (ME) retired in 1994 (Kelly 1995).

This recent harmony notwithstanding, there is little reason to expect that House Democrats have returned to the stability of the Austin-Boston years. If Pelosi advanced unopposed to the speakership, her overall rise in the leadership was not without opposition. She bested Steny Hoyer 118 to 95 in a titanic 2001 battle for minority whip; and her elevation a year later from whip to minority leader was slightly challenged as she rolled (177 to 29) over Harold Ford (TN) who had argued that Pelosi was too liberal to be the House Democrats' top leader.¹⁴

While the races for the 110^th portrayed a great deal of stability and accommodation among House Democrats perhaps reflecting their satisfaction with leaders who had returned them to majority status after 12 years, there nevertheless was one contested race. While Democrats rarely challenge floor leaders who become speaker, they will contest for the floor leadership and that pattern resurfaced when there was open competition for the majority leader's spot—in John Murtha's (D-PA) challenge to Hoyer. Murtha's late entry, general ideological placement well to the right of most of the caucus, and Hoyer's existing place in the leadership as whip led to an easy (149-86) victory for Hoyer. Still, the undercurrent of personal animus between Pelosi and Hoyer leads one to wonder if the Murtha challenge was evidence of the kind of conflict that prior Democrats avoided though Republicans (and perhaps some contemporary Democrats) seem willing to court. Indeed, Pelosi's difficulties in getting Murtha chosen over Hoyer echoed the inability of newly elected Minority Leader Gerald Ford's (R-MI) efforts to displace Leslie Arends (R-IL) as Republican whip in 1965; in any event, it was a message to Speaker Pelosi (as Arends' victory was to Ford) that there were limits to her authority in the caucus.

The Republicans, it seems, continue to be more prone to competitive leadership selections and to revolt. The 2006 leadership elections displayed a great deal of intra-party conflict for leadership posts: with Tom DeLay's resignation from the House, an early February race where John Boehner (R-MO) and John Shadegg (R-AZ) challenged acting Majority Leader Roy Blunt (R-MO) to replace DeLay was but one of several competitive Republican races in 2006. Such open competition, and particularly against a sitting whip and acting majority leader, was, in part, a response to Republicans' dire political situation as it was shaping up in early 2006 (Ota 2006a; 2006b; 2006c). After the election returned House Republicans to minority status for the first time since 1994, Republicans' higher tendencies for intra-party conflict coupled with minority-party frustrations led to even more conflict. Conservatives Mike Pence (R-IN) and John Shadegg (R-AZ) led unsuccessful “revolts” against Boehner and Blunt, respectively, arguing that Republican losses were to be blamed on the party's infidelity to conservative principles. Deborah Pryce's (R-OH) narrow re-election in her Ohio district and the likelihood of an internal challenge led her to vacate the Conference chair post producing two of House Republicans' three “open competition” races (one for Pryce's old job and the other for the chair of the Republican Policy Committee which was vacated by Adam Putnam [R-FL], who sought and eventually won the Conference post). If things were somewhat more tame on the Republican Senate side with Mitch McConnell (R-KY) succeeding unopposed from whip to floor leader and three other “consensus choice” selections for lower party posts, there was one late surprise “open competition” race as Trent Lott (R-MS) beat Lamar Alexander (R-TN) for whip (again by one vote) to return to the leadership after his 2002 resignation as floor leader.

In all, it seems that intra-party conflict for leadership posts is alive and well in both parties (the Democrats' recent harmony notwithstanding) and the predictability of leadership selection has been lessened. Since the Republican takeover in 1995, only one instance of an established pattern of succession—Pelosi's moving from floor leader to speaker—has occurred and there have been no “challenges to an heir apparent” since 1969 (largely because there have been few heirs apparent for any leadership post).¹⁵

By the same token, getting knocked off the “leadership ladder” seems less permanent than in the past. Not since the establishment of the leadership ladders of succession (see Brown and Peabody 1992) had former leaders knocked off the ladder returned to leadership posts, but both John Boehner's selection as Republican floor leader in 2006 (after having been out of the leadership since being beaten by J. C. Watts [R-OK] in 1998 in one of many Republican “revolts”) and Trent Lott's return as Senate minority whip provided two such instances of a remarkable renaissance of a legislative leadership career.

Conclusions: New Roles, New Leaders

What does this recent spate of leadership selections reveal about followers' expectations of their leaders and the emerging prerequisites of the key party posts in Congress? In addition to dealing with intra-party factions and the party's overall strategic context, parties' choices of their legislative leaders are also about “casting” the right actor to fulfill the many roles that leaders perform throughout a Congress. If middlemen were particularly appropriate and effective for a bargaining era and bipartisan Congress, new leadership roles, including leaders' public, media-oriented roles (Harris 1998; Sinclair 2006) might advise selecting different, more ideological and aggressive leaders. In addition, demands for fundraising are increasingly important in leadership selection. Indeed, Pelosi's advance to the speakership as well as Boehner's rise to the Republican floor leadership (Green and Harris 2007), aided in large part by prodigious fundraising, exemplify significant changes in the campaigns legislators wage for leadership positions. If the narrowness of recent congressional majorities somewhat maintains the value and importance of “middlemen” and chamber moderates who can bridge intra- and inter-party divides, the twin needs to “go public” and to court a party's donor base (which is often extraordinarily partisan) increase the need to select leaders from the party extremes as well.

Moreover, as party leaders (middlemen and extremity leaders alike) represent their polarized parties, their increased control over agenda-setting and legislative issue-framing define debates both in Congress and in the broader public. Thus, ideologically extreme party leaders and the more general polarization of the House and Senate parties may not only be the result of a partisan legislature, they may also be reciprocal, contributing factors to partisanship as well. To the extent that this is true, match-ups between Pelosi and Hoyer, on the one hand, against Boehner and Blunt, on the other hand, as well as the very conservative Senate Republican leadership team of McConnell and Lott (particularly when pitted against outspoken liberal leader Dick Durbin) suggests that America's deep partisan divisions are likely to continue to be reflected in and, perhaps, exacerbated by what takes place in Congress. Indeed, if Harry Reid was almost the exact middleman among Senate Democrats, he certainly has led the 110^th in an aggressive manner, repeatedly challenging the Bush administration on Iraq and other issues. In the contemporary partisan-polarized Congress, even the middlemen counterpunch.