Partners and Rivals: Representation in U.S. Senate Delegations

INTRODUCTION

Pilfering laugh lines is a public speaking practice that goes back at least to Demosthenes. Most perpetrators escape unscathed, but occasionally one does not. A month ago in New York City, Senator Alfonse D’Amato heard his Democratic colleague, Senator Daniel Patrick Moynihan, respond to a laudatory introduction with this line: Lyndon Johnson used to say, I wish my mother and father could have heard that introduction. My mother would have believed it, and my father would have enjoyed it. Three weeks later before a different audience, Senator Moynihan used the same opening with success and left before Senator D’Amato arrived. The Republican began his remarks, obliviously: I wish my mother and father could have heard that introduction. My mother would have believed it, and my father would have enjoyed it. The audience collapsed with laughter. One member said later: I think D’Amato still believes he was a smash.

(Francis X. Clines and Warren Weaver, Jr.,

Who Laughs Last)

COMPETING FOR good opening remarks at constituent events is just one of many ways in which senators from the same state present obstacles to each other in building successful Senate careers. The United States Senate is one of the few legislative institutions in American politics that has two legislators for the same geographic district. ¹ From a representational standpoint, we should expect to see senators from the same state representing similar issue areas and economic interests in their legislative portfolios. From an electoral standpoint, we might also expect to see senators addressing similar issues—if they are perceived to be popular among voters—because they each seek a majority voting coalition in the state. For example, senators from Nebraska, a heavily agricultural state, should both be expected to address farm interests. Senators from the same state should also be expected to respond equally to an overriding policy concern in that state; for exam ple, both senators from Texas should make immigration a primary focus of their legislative agendas.

However, what we actually observe is that senators from the same state do not build legislative agendas based on similar issues and interests; on the contrary, they build very distinct and separate legislative agendas on which they base much of their reputation as senators, regardless of same- or split-party affiliation. We are therefore faced with the following puzzle: Why do senators who represent the same state, but who never compete for the same Senate seat, adopt distinct representational agendas?

The concept of dual representation—two legislators sharing the same geographic constituency—is a vital and underappreciated component of legislative behavior in American politics.² From the very first day they enter the Senate, senators face the constraint of having another colleague from the same state who has similar institutional opportunities to represent that state. Each senator’s decisions about voting, committee selection, bill sponsorship, amendments, media strategies, political contributions, and campaign behavior are made in the context of sharing the political stage with the other senator from their state. As will be shown throughout this book, senators from the same state join different committees, focus on different policy areas, and address different economic interests through bill and amendment sponsorship. Moreover, senators also adopt contrasting governing styles that are products of their own personal style, the political expectations in their state, and the existing reputation of the other incumbent senator.

The central argument of this work is that the incentive to differentiate within a Senate delegation is a combination of electoral incentives and institutional forces that push senators in contrasting directions. Taking advantage of the particular blend of party, economic interests, and political culture within their state, each public official seeks to develop an identifiable niche, which helps attract media coverage and build the long-term reputation necessary for a successful political career. The behavior of each senator therefore shapes the range of state interests and opinions that are incorporated into federal policy-making. This goes to the heart of democratic decision-making. In assessing how well a senator represents his or her state, we often examine an individual senator’s voting record and try to match up that behavior with objective state interests or opinions. Inevitably, we conclude from such a test that there are constituents whose interests and opinions go unheeded by the senator. When we do that, though, we are setting up the senator for failure because no legislator can possibly address all the interests and opinions in a single state. Most important of all, a senator is not supposed to represent the majority constituent opinion on every issue because there is another senator who can also provide representation to the state. If we continue to measure Senate representation by examining just one-half of a Senate delegation, we will always produce incomplete results.

As an alternative, this book proposes to assess Senate representation from the perspective of the two senators from each state, each of whom works individually to craft a reputation and seek reelection, but who, when examined as a collective unit, attend to a wide spectrum of interests and opinions in their states. In doing so, this work moves beyond conventional approaches to measurement of representation by demonstrating how senators work within a multidimensional framework. This approach is key for those who believe that partisanship is the clearest signal of accurate representation. If scholars look only through the lens of partisanship, they can construct a majority constituent interest in each state when, in fact, no such majority exists. Subsequently, they might mistakenly conclude that constituents from a state where both senators share the same party affiliation would only be represented in the Senate if they shared that party affiliation.

In reality, Senate representation is multidimensional: constituents have a range of interests and political opinions, based on their partisan views, economic status, and geographic location, which may overlap on one dimension but rarely overlap consistently on all three at any given time. Senators react to the varied structure of constituent interests by constructing legislative agendas that appeal to a cross-section of the state, and which go well beyond partisanship. Senators from the same state do not oppose each other directly in an election, but they nevertheless perceive the need to establish contrasting reputations to secure reelection and therefore have a strong incentive to rest the foundation of their support on different combinations of constituent interests. To do that, senators have to take concrete steps, both in the state and in the Senate, to adopt separate portfolios. A theory of representation that takes this behavior into account by incorporating the multidimensionality of constituent interests, and the unique character of the two-member structure of Senate delegations, can provide a more comprehensive understanding of democratic representation in American politics.

Additionally, studying the Senate in this way can yield fruitful comparisons to the behavior of legislators in other political systems that have multimember districts. Even though there are some important differences, namely, that senators do not compete directly in a simultaneous election as legislators in other systems do, there are potential similarities in the way that candidates and legislators shape their portfolios. We would expect parties that run a number of candidates in a given district to differentiate those candidates along some dimension of representation—either ideology or governing style or issue selection.³ Applying the work of comparative politics scholars to explain behavior in the U.S. Senate can shed light on how and why legislators who share the same geographic constituency make starkly different choices in issue selection, governing style, and ideology, as candidates and as elected officials.

In many respects, representation in the Senate is stronger than often thought. In the world of contemporary politics, enormous public cynicism about representative democracy dominates political discourse; ordinary citizens believe that public officials are out for themselves and not representing voter interests. Although public discontent can be attributed to a number of sources, part of citizen unhappiness may be a function of academic studies that show weaker levels of representation than actually exist. If we redefine the notion of representation to include dual representation, we find that senators are doing a much better job representing voters than is commonly believed.

The remainder of this book sets out to demonstrate key features of this argument. Chapter 1 reviews the origins of dual representation. Going back to the founding period of American history, I discuss how the dynamics of competition and cooperation between same-state senators are an unintended consequence of compromises made at the Constitutional Convention. For all their foresight, the framers underestimated the extent to which state legislatures might use the Senate as a means of ensuring that the maximum number of state interests would be represented at the national level. By sending senators with contrasting economic and political viewpoints to the Senate, states could achieve a broader scope of representation at the national level than if they had chosen to treat their Senate delegation as a single entity. The Senate behavior that we witness today, when we have direct Senate elections, mirrors the ancestral behavior of state legislatures in this regard. Senators from the same state, both in campaigns and in the Senate, now do what state legislatures used to do in differentiating themselves in order to build distinct reputations among constituents.

Chapter 2 examines how senators build their reputations and shape their careers in the Senate. Every day, senators make decisions that affect the way they represent their states as well as their overall reputation among their colleagues and constituents. At different points throughout a congressional session, a senator’s decisions may take on more or less importance, depending on the issue at hand and the legislative arena in which he is operating. A senator casts a roll-call vote, which may or may not accord with the party’s or state’s interests, in committee and on the Senate floor. At the same time, a senator may be deciding whether to sponsor major legislation, or to cosponsor bills introduced by colleagues. Likewise, a senator may decide to offer an amendment to a chair’s mark in committee, or to a bill once it reaches the Senate floor. Every senator makes these decisions within a context that incorporates the decisions that his state colleague makes about the very same potential universe of issues and concerns.

To measure the extent of overlap in legislative areas between two senators from the same state, an index of similarity is constructed as a function of the diversity of state interests, seniority, same- or split-party affiliation, and prior competition in state elections. In general, senators from larger states, which tend to have a more diverse set of interests, will overlap less than senators from small states do because the wider spectrum of state interests makes it easier for them to construct different legislative portfolios. As chapter 2 also shows, senators who share the same party affiliation face greater electoral and institutional incentives to differentiate in these areas of legislative behavior than senators who have opposite party affiliations. Given that senators from the same party and the same state will look very similar on their roll-call voting records, they take great pains to contrast themselves in all the other legislative arenas in which they operate.

Chapter 3 is an in-depth content analysis of local media coverage of senators from the same state. In this chapter, I argue that because the average senator is unlikely to attract a lot of national press attention, it is a more productive and efficient strategy to gear his efforts toward the local media. Unlike most House members, individual senators are likely to receive some minimum level of coverage for their legislative actions in particular issue areas. However, because two senators from the same state share the same media audience, they face constant competition in their efforts to attract individualized media coverage. This competition for visibility provides same-state senators with a strong incentive to develop contrasting and distinct representational agendas, and it pushes them in more divergent directions than we might expect of two legislators who represent the same geographic constituency.

The results support the argument that the need for publicity to build visible reputations in the state drives senators from the same state to diversify their legislative portfolios. In general, the amount and content of local media coverage of a senator reflect choices of committee assignments, issue specialization, geographic ties, and general governing style. There is a direct and strong correlation between committee assignment and media coverage across almost all issue areas. Therefore, the decisions that a senator makes to join a different set of committees from that chosen by a state colleague pays off in terms of individualized media coverage. Moreover, patterns of media coverage discourage encroachment into the other senator’s territory because when one senator receives publicity in an issue area, the other senator is less likely to be covered in that same issue area. Lastly, a senator’s effort to differ stylistically from his state colleague among the press also reaps rewards because the impressions that the elite media develop are filtered down into everyday newspaper coverage.

In Chapter 4 the main focus is on how senators use reputations to cultivate their home-state audiences. Because the general level of familiarity with elected officials is so low, and party affiliation as a predictor of vote choice has declined, senators face greater uncertainty about potential support among constituents than they did thirty-five years ago. This electoral uncertainty exerts pressure on senators to establish broad reputations, based on party, legislative portfolios, and governing styles, as a means of marketing themselves to their constituents. Given the sheer difficulty of penetrating the many layers of information available to most voters, senators and their staff consider even the vaguest impression among constituents to be a sufficient return for the cost of the resources that are devoted to building a reputation.

Chapter 4 uses both local media coverage and public opinion to test for constituent recognition of senators’ legislative portfolios, and to test for awareness of differences between two senators from the same state. To this end, the results provide support for senatorial differentiation in two ways. First, senators’ choices about legislative work are reflected in local media coverage about them, which means that senators’ reputations are based on reasonably accurate information. Second, differentiation among senators from the same state is recognized by constituents and reflected in their specific reasons for liking and disliking their senators. In sum, the efforts by senators from the same state to use their legislative work to establish unique reputations in their states, controlling for their partisan and ideological affiliation, appear to succeed.

Chapter 5 explores the extent to which senators’ efforts to differentiate succeed in attracting distinct sets of constituents; in particular, it provides an overview of the geographic and demographic patterns of electoral support for senators from the same state. This chapter includes a model of geographic representation, which shows how senators start from a particular region in their state and build their core group of support by expanding to other parts of the state. When two senators expand their bases into each other’s home territory, they come into direct competition with each other for the support of constituents in that region. Such support is not mutually exclusive, but county-level electoral returns indicate that senators from the same state, and the same party, do not attract equal levels of support in the same counties, controlling for the quality of the challenger. In addition, building on existing demographic analysis, this chapter reveals that senators from the same state attract votes from voters in different ethnic and economic class groups, even when they are from the same party.⁴

Chapter 6 examines the patterns of support from economic interests in the state. Constituents’ economic interests can be delineated by their profession, for example, doctor or farmer, or they can be employees of a major industry in the state; in either case they are usually represented by organized groups or political action committees (PACs). Senators try to address the economic interests of their constituents in their legislative portfolios. Because the benefits that senators secure for these economic interests are often narrowly concentrated and easily identifiable, this strategy can be a highly efficient means of attracting both electoral and financial support in the state. Electoral support for a senator that stems from targeted economic benefits may overlap with the senator’s partisan or geographic coalition, but in some cases, voters who might not be otherwise disposed favorably toward the senator choose to weigh his efforts to help the state’s economic interests more heavily than party affiliation or geographic association. Financial support will come from those who benefit from the senator’s efforts, either individually or as part of an industry, in the form of sizable campaign contributions to the reelection effort. In this way, the economic dimension of representation, depending on the diversity of the state’s economy, can serve to expand the range of interests for senators to choose from when constructing their legislative portfolios.

Chapter 6 also combines work on how two senators court economic interests in their state with current findings on the determinants and effects of campaign contributions on legislative behavior. First, I construct an alternative measure of demand behavior by using legislative activities other than roll-call voting as indicators of senators’ attentiveness to state interests. Second, I use patterns of campaign contributions by state-based PACs as a test of whether state economic groups distinguish between two senators from the same state. If senators are succeeding in distinguishing themselves in the area of state economic interests, then we should see state-based PACs contributing more to one senator than the other, depending in large part on which senator appears more willing to address their specific interest. The results indicate that state-based PACs do not uniformly contribute the same amount of money to each senator from the state, and some PACs only contribute to one senator and exclude the other. Furthermore, PACs clearly appear to be investing in one senator over the other, but not investing based on party affiliation. Rather, a PAC’s contribution decision appears to vary with a senator’s institutional position and willingness to address its specific economic interest.

Chapter 7 argues for expanding the framework within which we study Senate representation to incorporate the effects of two-person Senate delegations. The comprehensive model of Senate representation presented here leads to a richer and more accurate understanding of individual Senate behavior. This chapter also suggests that the U.S. Senate has more in common with legislatures in other countries than previously believed. Scholars rarely draw parallels between U.S. legislatures and legislative systems with multimember districts, but viewing the Senate as a multimember-district institution makes it possible to do so. In one way, then, this book may add to the growing body of work that bridges the divide between the study of comparative politics and American legislative institutions.

RESEARCH DESIGN AND METHODS

This book relies on qualitative and quantitative data sets that are drawn from a number of sources. The sample of senators that is used includes all senators who held office during the years 1987–1992, with additional references to senators who held office in subsequent years. Because this study explores the multiple facets of Senate representation, data were collected to measure legislative behavior, media coverage, public opinion, and campaign contributions, using the senator as the unit of analysis. Indicators of legislative behavior include committee assignments and positions, roll-call votes, bill and amendment sponsorship, and bill and amendment cosponsorship. (The names of the Senate committees used here conform to the names used during the time period 1987–1992.) Patterns of media coverage are measured by the number and content of newspaper articles about senators taken from local newspapers in a sample of ten states. Public opinion of senators is taken directly from the open-ended like and dislike questions in the 1988-1990-1992 pooled Senate Election Study. The data on campaign contributions were taken from Federal Election Commission records for the same time period. A description of each specific data set is included in the chapter in which it is used, and the definition of the variables that are used in the quantitative analysis is included in appendix A.

The qualitative data used in this book come primarily from personal interviews with Senate staff members on legislative agenda formation. The interviews were conducted over the course of August 1991 to January 1993. To measure elite opinion of senators among members of the media in their states, I sent a mailed survey to newspaper editors and reporters; the bulk of the information from respondents is used in qualitative form, and a copy of the questionnaire is included in appendix B. Lastly, selected newspaper articles were collected as part