The story so far has been one of political ambition, the electoral costs and institutional benefits of switching parties, and the willingness of legislators to take a risk by jumping to the other side. Switching parties is a personal decision made after weeks, months, or even years of serious consideration by individual legislators who must ultimately decide what is in their best interest; the proverbial buck stops with them. The decision, however, is not made in a vacuum. Rather, the party switcher is part of a larger social system, namely the legislature, where such a career-defining moment is not lost on party leaders or rank-and-file colleagues. A member's decision to switch parties will affect those actors as well, and in turn, those actors can influence a legislator's thinking. Instead of being passive bystanders, it may be reasonable to surmise that individuals such as party leaders will play an active role in legislators’ decision to switch parties. The next two chapters provide a glimpse into those dynamics with evidence from semistructured interviews with elected officials and their staff members.

In this chapter, I turn to the behavior of party leaders, rank-and-file members, and former party switchers, and ask the following questions: Do party leaders help lower the costs of party switching? To what extent do other members interact with switchers or potential switchers, both before and after a party switch? What goes on “behind the scenes” that, unlike committee assignments or electoral returns, may not be readily observable? And how does this advance our theoretical understanding of elite decision making generally and party switching specifically? My contention is that answers to these questions will provide even more depth to our understanding of legislators’ decision to switch parties.

The better-known cases of party switching may be found at the national level (e.g., Arlen Specter, Jim Jeffords), but elite party switching is not confined to the halls of the U.S. Congress. Following the 2010 midterm elections, for instance, a number of state legislators – many of them conservative southern Democrats – decided the time was right to bolt to the Republican Party. Those cases of party switching were widely reported in the media as an indication that on the heels of its historic landslide victory, the GOP had the wind at its back. Theoretically any sitting politician at the national, state, or local level can decide to switch partisan stripes. These numerous opportunities for party switching at all levels of government in the United States can then be leveraged to inform our theoretical understanding of the phenomenon. With a much wider pool of potential party switchers, states legislatures represent another arena in which I can test some empirical implications of my theoretical framework.

From the perspective of testing theories of elite decision making, the many cases of party switching at the state level offer at least three distinct advantages. First, we can examine whether the theory and findings from the congressional analyses are generalizable to other settings across time and space. Second, the data allow for more precise estimates if only because of the larger number of cases. Third, as one of the main launching pads for a national career (Berkman 1993; 1994; Canon 1990; Maestas, Maisel, and Stone 2005; Maestas et al. 2006; Schlesinger 1966), state legislatures are an incubator for ambitious politicians with national aspirations. Party switching among state legislators therefore has implications that can spill over to the national stage because it directly affects one of the main pipelines of viable congressional candidates.

This chapter serves a very specific goal: to further our understanding of party switching by testing one implication of the theory in the state legislative arena, namely the role of progressive ambition in fueling party switching. The central motivation for this chapter, then, is to test a causal mechanism of the theory with systematic data from an entirely different setting. In short, I examine whether the progressive ambition hypothesis finds support at the state level. In addition, this chapter offers detailed descriptive accounts of specific instances of party switching at the state level.

On April 28, 2009, U.S. Senator Arlen Specter announced that he was leaving the Republican Party and seeking reelection as a Democrat. His decision, coupled with the adjudication of the 2008 senatorial election in Minnesota in favor of comedian-turned-politician Al Franken, meant that Democrats would, for the first time in more than thirty years, have the sixty votes required by Senate Rule XXII to invoke cloture and thwart Republican filibusters. Specter's switch gave rise to a predictable torrent of reactions. The news “thrilled” President Obama, as he vowed to give the senior senator from Pennsylvania his “full support.” Many Democratic senators and other party bigwigs, such as Pennsylvania Governor Ed Rendell, immediately embraced Specter. The positive reception from the Democratic Party brass was met with equally intense scorn from the Republicans. Republican National Committee chair Michael Steele, for instance, accused Specter of “put[ting] his loyalty to his own political career above his duty to his state and nation,” while Senator John Cornyn, the chair of the National Republican Senatorial Committee, accused Specter of “political self-preservation.”

According to a press release issued by Specter's office, ideology was a key factor behind his decision, with his “political philosophy more in line with Democrats than Republicans.” And, according to news reports, Specter realized he was in danger of losing the upcoming Republican primary in 2010. A poll taken shortly before the switch showed the five-term senator trailing conservative stalwart Pat Toomey by more than twenty points (Rasmussen Reports 2009); six years earlier, Toomey had come within two percentage points of defeating Specter. It appeared, in early 2009 at least, that Specter would stand a better chance of getting reelected were he to avoid the Republican primary and instead run as a Democrat.

Specter's decision came after weeks of prodding not only from Senate Majority Leader Harry Reid, but also from Vice President Joe Biden. The two longtime colleagues often rode the train back to their respective states together, and the vice president now saw a great opportunity to bring Democrats closer to the sixty-vote mark in the Senate while striking a blow to the image of the Republican Party as a mainly conservative, southern party anathema to a “reasonable” moderate such as Specter.

These quotations are an apt synopsis of the main theoretical assumption I make in this book. Having witnessed a number of Democratic officeholders defecting to the Republican Party during the 1990s, former Democratic representative (and political science professor) Glen Browder made his observation after Republicans had approached him about possibly switching sides. What struck Browder was the extent to which many party switchers failed to be embraced immediately by their new party, as well as the animosity felt by former co-partisans they left behind. As a politician from Alabama and a scholar of southern politics, Browder had a front-row seat to appreciate the dynamics surrounding party switching, and he noticed that incumbents who switched parties were often attacked from both sides.

The idea that switching parties entails significant costs is not the dominant paradigm in the literature. To the contrary, the received wisdom is that politicians incur benefits at the polls by switching into the “correct” party in terms of their constituency (Aldrich and Bianco 1992). This intuitively makes sense, as it is reasonable to expect reelection-seeking politicians to bolt to the other party when doing so increases their chances of getting reelected. Policy motives also figure prominently in explanatory models of party switching (Castle and Fett 2000). Yet most politicians, even those in electoral jeopardy or who are cross-pressured ideologically, do not switch parties.

Political parties are essential to the operation of electoral politics and representative democracy. Candidates affiliate with parties, voters support candidates on the basis of this affiliation, and legislative battles are often fought along partisan lines. Because politics is a team sport where the players choose the side with which they affiliate, understanding the factors that lead politicians to affiliate with one party or another is a question of great theoretical and practical import.

Political ambition is also at the heart of representative democracy. By virtue of their desire to win (or retain) office, politicians tend to behave in predictable ways, engaging in actions that increase their chances of electoral success while shying away from behaviors detrimental to their electoral prospects. Reelection, therefore, provides significant constraints on the choice of actions by ambitious politicians.

Yet reelection is not the only objective sought by politicians. Obtaining and yielding power, effecting public policies that reflect one's preferences, and moving up the political ladder are some of the goals that go beyond the narrow reelection motive. As I argue throughout the book, legislators’ decision to switch parties cannot be explained fully by a singular focus on reelection. Crossing the aisle is as much about progressive and intrainstitutional ambition as it is about reelection. This is not to say that legislators switch parties with the intent of losing votes. Rather, what I assume is that switching parties entails significant costs and that politicians may be interested in more than the upcoming reelection. I then derive some predictions regarding elite behavior, which I put to the test with original data from Congress and state legislatures. I also show, empirically, that party switching does come at a significant cost at the ballot box, and I provide both quantitative and qualitative evidence to that effect. It is no coincidence that very few legislators choose to leave their party voluntarily. As it turns out, the act entails much more than donning the other team's gear. The decision to cross the aisle is a complex one, and by focusing so much on the reelection motive, our extant models did not provide a full picture of this decision.

Between 2005 and 2009, I conducted interviews with thirteen members of Congress or their staff. In addition, I interviewed a state legislator who switched parties and a former county legislator who ran against a future congressional party switcher. I made attempts to interview three types of members of Congress (or their high-level staff): party switchers, party leaders, and potential switchers.

I considered a potential switcher any member with at least two confirmed reports of a potential switch or wooing efforts from the other party (from sources such as The Almanac of American Politics, Roll Call, The Hill, or print newspapers). These potential switchers were primarily southern Democrats and northeastern Republicans. While a reliance on media reports could potentially lead to some members being wrongfully included in the pool of nonswitchers, I relied on credible outlets and looked for information that could be verified in at least two sources. Moreover, none of the nonswitchers I interviewed denied having been approached to switch parties or having thought about it at some point.

My relatively low response rate (25.5 percent) precludes any kind of meaningful statistical analysis on the data collected from the interviews. Also, I do not claim to have a representative sample. My interviews are meant to provide a behind-the-scenes look into the role of leaders and other members in the decision to switch parties, a dynamic that cannot be easily observed from the outside. They are also critical in establishing the validity of some of my theoretical assumptions. Finally, they serve to supplement the quantitative portions of the book rather than provide additional data used to test some hypotheses derived from the theory.

Each interview was semistructured and lasted an average of about fifteen minutes. I promised each respondent anonymity and that I would treat their responses as “not for attribution.” A few respondents agreed to waive anonymity and be interviewed on the record. Nevertheless, I keep all interviews anonymous in order to maintain the confidentiality of all subjects. With a small pool of potential subjects, identifying some interviewees may lead to other subjects being identified simply by process of elimination. I also use the masculine pronoun to identify interview subjects throughout the book. Readers should not assume that this reflects a subject's gender.