Authors Christopher J. Devine and Kyle C. Kopko have focused their careers on studying the Vice Presidency, and have published the first detailed study on what, if any, impact a running mate has in a Presidential election. With Joe Biden set to announce his running mate, we’re happy to share the introduction to Devine and Kopko’s Do Running Mates Matter: The Influence of Vice Presidential Candidates in Presidential Elections.
In January 2015, there were no declared candidates for the next presidential election. It was just too early. Yet the veepstakes already had begun. Rumor had it that Ohio’s senior US senator, Sherrod Brown, was a leading contender for the Democratic vice presidential nomination. On paper, he was the perfect running mate: an experienced, two-term senator from a key battleground state, a vigorous campaigner, and a bona fide progressive who also could appeal to blue-collar swing voters. Many Democrats wanted Brown to run for president, in fact. But, like so many other credible candidates who doubted that Hillary Clinton could be defeated for the party’s nomination, he declined. So, speculation shifted to the next best thing: a slot on the presidential ticket, as Clinton’s running mate. There was one problem: Brown didn’t want it. Not at all. In his words: “I have zero interest in being vice president” (Terris 2015).
Who Wants to be the Vice President?
Brown’s proclaimed disinterest in serving as vice president was not surprising. Throughout US history, the vice presidency has been derided as a dead-end job for ambitious politicians whose talents are spent on helping the presidential candidate to get elected rather than helping the president to govern once in office. In the words of one former vice president, Walter Mondale, “The office is handmade for ridicule and for dismissal. In the nature of it, you always look like a supplicant, a beggar, a person on a string” (Woodward and Broder 1992, 196).
Vice Presidential Power(lessness)
The vice presidency owes its unenviable reputation—notwithstanding many informal expansions of power since the 1970s (see Goldstein 2016)—to its institutional design. Indeed, the US Constitution grants few formal powers to the vice president. These include, first, presiding over the US Senate (a power that Harry Truman’s vice president, Alben Barkley, last exercised with regularity; see Goldstein 2016, 22); second, casting tie-breaking votes in that body (rarely); third, in the presence of the US House and Senate, opening the sealed certificates containing each state’s votes in the Electoral College, and then overseeing the electoral vote count; fourth, and most important, assuming the office of president of the United States upon the president’s death (eight times in US history), resignation (once), removal from office, or incapacitation, and when a presidential election remains unresolved at the time that a new presidential term is scheduled to begin.
Dismissing the Vice Presidency
No one has been more critical of the vice presidency, or so eager to make jokes at its expense, than the vice presidents themselves. Take, for example, John Nance Garner’s dismissal of the vice presidency as being “not worth a bucket of warm spit”—or the favorite tale of several vice presidents, including Thomas Marshall (Milkis and Nelson 2011, 486), Alben Barkley (Smith 2008, 177–178), and Hubert Humphrey (Unger and Unger 1999, 255): “A mother has two sons; one goes off to sea, the other becomes vice president. Neither is heard from again.”
The first vice president, John Adams, famously called it “the most insignificant office that ever the invention of man contrived or his imagination conceived.” Lyndon Johnson, who served for nearly three years as John Kennedy’s vice president, said of his tenure: “I detested every minute of it” (Baker 2013, 60). Gerald Ford, who briefly served as Richard Nixon’s vice president, reportedly described this as “the worst eight months of [my] life” (60). And Nelson Rockefeller, who was nominated and confirmed to the office following Ford’s succession to the presidency, often referred to the vice president as mere “standby equipment.” He explained: “I did not want to be Vice President. I’m a doer by nature, an activist. And I always felt, and I told Dick Nixon that in 1960 when he asked me to [be his running mate, allegedly], that it was standby equipment and I just wasn’t cut out for it.” Rockefeller added: “I’ve known all the Vice Presidents since Henry Wallace. They were all frustrated, and some of them were pretty bitter.”8 Dick Cheney, one of Rockefeller’s successors, later would concur: “I’d never met a vice president who was happy.”
Conflicted?
Yet many of the same vice presidents and potential running mates who have dismissed the office as ridiculous at other times have betrayed a more conflicted—or, at least, a more nuanced—view of the vice presidency. In a follow-up interview to the one excerpted earlier, Nelson Rockefeller offered a very different assessment of the office:
“I totally disagree with John Nance Gardner [sic]. I think the office is a very important one, depending on the relation between the President and the Vice President and at least during the first 2/3rds–3/4ths of the time I was Vice President I’ve never been busier—heading commissions, undertaking special projects for the President and traveling at home and abroad. It’s a very useful function in terms of both ceremonial activities that relieve the President and which are interesting and important, plus, depending on the experience of the individual, the opportunity to use that experience to undertake assignments for the President.”
Dick Cheney, who initially resisted entreaties to run alongside George W. Bush in 2000, and bluntly recalled the unhappy fate of previous vice presidents, nonetheless said at the end of his two terms in the office: “I don’t regret it for a minute. It’s been a tremendous experience” (Malcolm 2008).
And then there is John McCain. After failing to win the Republican Party’s presidential nomination in 2000, McCain brushed off speculation that he would join the ticket as George W. Bush’s running mate. “The vice president has two duties,” he scoffed. “One is to enquire daily as to the health of the president, and the other is to attend the funerals of third world dictators. And neither of those do I find an enjoyable exercise.” Four years later, amid speculation that John Kerry would ask him to run for vice president on the Democratic ticket, McCain joked: “I spent seven years in a North Vietnamese prison camp, in the dark, fed with scraps. Do you think I want to do that all over again as vice president of the United States?” (Halbfinger 2004).
Yet in 1996, when Bob Dole reportedly weighed selecting him for the Republican ticket, McCain expressed a more sober, reverential view of the vice presidency. “John Nance Garner described the office as not being worth a bucket of warm spit, but I hold the office in higher regard than that,” McCain said. “It is certainly prestigious and would be a wonderful opportunity for some” (Pittman 1996). Then, in 2008, while once more—and this time successfully—seeking the Republican Party’s presidential nomination, McCain said in a debate that “the vice president of the United States is a key and important issue, and must add [sic] in carrying out the responsibilities of the President of the United States.”
Playing the Vice Presidential Game
It is a good indication of the vice presidency’s actual value that—jokes and public disavowals of interest notwithstanding—plenty of qualified individuals are willing to be selected as the running mate and undergo an intensive vetting process for that purpose. In fact, many ambitious politicians actively lobby for their selection, behind the scenes.
Take Dan Quayle, for example. George H. W. Bush’s decision to name Quayle as his running mate in 1988 came as a shock to nearly everyone, including many Bush campaign staffers. But Quayle had been plotting, along with two of his top Senate aides, to secure a slot on the Republican ticket since the day after Bush won New Hampshire’s Republican presidential primary, six months earlier (Woodward and Broder 1992, 15). This “sub rosa campaign” was designed to raise Quayle’s profile nationally, and with Bush in particular. In early 1988, Quayle began delivering more speeches in the Senate, issuing more press releases, and writing more newspaper op-eds, particularly on issues of national defense, than at any point during his previous seven years in office. Quayle also tried to make himself more visible to Bush and his inner circle by increasing contact with senior campaign advisers, more regularly visiting the vice president’s office in the US Senate, and taking a more vocal role at the weekly Senate Republican lunches that Bush attended, as vice president. To anyone witnessing these efforts, Quayle’s intentions were clear. As one Republican Senate colleague, William Cohen of Maine, said: “It looked like there was a game plan to get Bush’s attention because Quayle thought he had a shot [at being chosen as Bush’s running mate].”
Yet Quayle would not readily acknowledge his campaigning for the vice presidential nomination, even four years later while serving as vice president. He did so only reluctantly, after Washington Post reporters Bob Woodward and David Broder presented Quayle with irrefutable evidence of his efforts. Why play coy? Because, Quayle acknowledged, “You don’t run for vice president.” Rather, to avoid the stigma of failure if passed over—and, probably, to avoid alienating the presidential candidate by seeming overly ambitious—“you keep expectations down and do things as quietly and subtly as possible” (Woodward and Broder 1992, 15–16).
In fact, Quayle had designs on the vice presidency ever since being elected to Congress in 1976, at the age of twenty-nine. At that time, he explained in a 2002 interview, “Obviously you’re thinking about running for President or Vice President. It’s there.” But, he cautioned, “You don’t do it overtly. It’s something you have in the back of your mind and you set out a path to do it.” Quayle followed that path to the US Senate, from where, he observed, most recent vice presidents had come. He described this as a “stepping-stone” to the vice presidency, and ultimately the presidency. “I was positioning myself to eventually run for President,” Quayle explained. “Now, obviously, the Vice Presidency was a stepping-stone to that. I mean, that’s why people want to be Vice President [emphasis added].”
Joe Lieberman also wanted to be vice president, in 2000—but he, too, could not say so. “You’re not supposed to campaign for the vice presidential nomination,” he explained later. “You can’t even acknowledge that you are under consideration for it” (Lieberman and Lieberman 2003,). When Al Gore’s campaign asked to vet Lieberman for the nomination that year—a process so invasive that Lieberman likened it to “a colonoscopy without anesthesia” —the prospective running mate consented and then, with the help of a close friend, “developed a strategy not simply for surviving the vetting, but for pursuing the nomination.” That strategy included “reach[ing] out, in a very discreet way, to a very few people who we thought might be talking to Gore about this selection,” as well as key constituency groups within the Democratic Party. In all cases, “These contacts were to be made subtly, quietly.”
Lieberman also sought out advice from Chris Dodd, a fellow senator from Connecticut and his most trusted colleague. Dodd counseled, “You should make sure you do everything you can so you will never look back to this time and say, ‘If I had done just one more thing, I might have been the vice presidential nominee.’” But, Dodd acknowledged, “Of course I understand that you can’t go out and campaign for it” (Lieberman and Lieberman 2003, 14).
Even after being selected, Lieberman would disclaim any ambition to the vice presidency. Wondrously, he recalled someone telling him, “This is like the ministry. You’re called to the ministry, you don’t seek it” (Barstow with Seelye 2000). But, of course, Lieberman had sought the vice presidential nomination—enthusiastically and methodically. He wanted to be the vice president. He just could not say so publicly at that time.
Quayle and Lieberman are only two examples, but they illustrate what seems to be a well-known strategy among politicians aspiring to the vice presidency: publicly deny that you are interested—better yet, laugh off the idea as ridiculous—all the while privately working with associates and campaign contacts to cultivate interest in your candidacy. Then, after the selection is made and the election is over, you can publicly admit it: yes, of course, I wanted to be the vice president. I really wanted it, in fact.
And that brings us back to Sherrod Brown.
Zero Interest?
In July 2017, Sherrod Brown gave another interview to Ben Terris, the Washington Post reporter to whom he had protested two-and-a-half years earlier: “I have zero interest in being vice president.” A lot had changed since that time. In the summer of 2016—despite reiterating that April, “I’ve made it clear [that] I don’t really want the job” (Raju and Schleifer 2016)—Brown agreed to be considered for selection as Hillary Clinton’s running mate. He even participated in a vetting process that Connie Schultz—his wife, and a Pulitzer Prize–winning columnist—described as “excruciating.” Brown was a finalist for selection; in fact, according to two highly placed sources, he was the first runner-up to Virginia senator Tim Kaine, whom Clinton selected only after initially favoring Brown (Terris 2017; but see, e.g., Allen and Parnes 2017; Baumgartner 2016). Perhaps the biggest strike against Brown was that, if elected, his replacement in the closely divided US Senate would be appointed by Ohio’s Republican governor, John Kasich.
For a man who had expressed “zero interest in the vice presidency,” being passed over for the position should have come as a relief. But Brown was disappointed—devastated, in fact. “By the end,” he admitted, “I really wanted it.” Brown wanted it so badly that he envisioned, in Terris’s words, “liv[ing] out of a bus” and barnstorming across the Midwest throughout the fall campaign. Might this have changed the election’s outcome? In Terris’s article—titled “Sherrod Brown Thinks He Could Have Helped Democrats Win in 2016. But What about 2020?”—Brown answered cautiously: “I don’t pretend that my being on the ticket would have made [Clinton] win. I don’t know. I mean, if I had gone to Wisconsin and Michigan a lot, anything would have changed those two states.” But then, perhaps to cover this tracks, he added: “My wife thinks we would have won. She thinks we would have won Ohio.” Of course, Donald Trump won Ohio instead—helping him to become the next president of the United States. And Brown was left to wonder whether he could have made the difference in the Electoral College, apparently haunted by the regret that he did not get that chance.
As it turns out, Sherrod Brown really wanted to be vice president.
A Fundamental Tension
The preceding examples illustrate a fundamental tension in public opinion about the vice presidency as a governing institution. On the one hand, everyone is aware of the office’s institutional weakness and its beleaguered reputation. This makes it easy to dismiss the vice presidency as a joke, and to belittle or reject the prospect of seeking that office. Yet many of the news or opinion articles that repeat those jokes—including, almost invariably, a reference to Garner’s metaphorical “bucket of warm spit” (and, just as invariably, a caveat noting that he probably used more graphic language)14—or report a potential running mate’s lack of interest in auditioning for the role, ironically, if not hypocritically, also engage in fevered speculation about the vice presidential selection process or the running mate’s likely effect on the presidential race.15 And, as we have seen, many of the vice presidents or prospective running mates who have disparaged the office or foresworn interest in seeking it at other times have celebrated its significance or actively campaigned for selection behind the scenes.
We do not argue that every such dismissal of the vice presidency or the prospect of running for it is insincere. Indeed, it is entirely possible that Sherrod Brown’s initial denials of interest in the vice presidency were genuine, and that later he had a change of heart—or that, in Terris’s (2017) words, “Sherrod Brown never wanted to be vice president, until one day he did.” Nor do we contend that a vice president plainly contradicts himself when he is quoted both lamenting and praising his experiences in that office. For example, Nelson Rockefeller truly might have felt like “standby equipment” at some times and like someone serving “a very useful function” at others. Such inconsistency, we argue, is indicative not of insincerity but of a pervasive conflict—or a “fundamental tension”—in attitudes toward the vice presidency, among political elites and within the mass public generally. In short, the vice presidency’s public image—truth be told—is not that of an inconsequential institution or an immensely powerful one (at least by its design); rather, it is an institution characterized by obvious strengths and weaknesses, the relative weight of which is difficult to calculate and often context-dependent.
John Adams provided perhaps the most profound articulation of this truth when he said: “I am vice president. In this I am nothing, but I may be everything” (Milkis and Nelson 2011, 486). Here he was referring to the possibility of succession to the presidency. In the era of the “modern vice presidency” (Goldstein 2016), when the power of that institution remains constitutionally stagnant but informally expansive, the tension between viewing the vice presidency as powerless and powerful is that much greater.
Who Cares About the Running Mate?
This book is about vice presidential candidates, not vice presidents. But the two roles are intertwined—and not just because, barring extraordinary circumstances, serving as the former is a prerequisite to serving as the latter. It is also because the (perceived) weakness of the vice presidency, as a governing institution, has shaped perceptions of the vice presidential candidate, as an electoral institution. Given the vice president’s limited constitutional powers, and the remote possibility of presidential succession, running mates traditionally have been selected (by party conventions and then, starting in the 1940s, by presidential candidates) on the basis of electoral considerations, typically in order to provide geographic, demographic, or ideological “ticket balancing” (Baumgartner 2012; Baumgartner with Crumblin 2015; Goldstein 2016). Even so, most voters have had little incentive to weigh the running mate’s credentials when voting in presidential elections, at least not to the extent of casting a vote on that basis, since only the president is guaranteed to exercise substantial power once in office. Recent expansions of vice presidential power might have altered that equation somewhat, but—as indicated by the evidence presented earlier, and by the predominant caricature of vice presidents in popular culture as buffoonish incompetents—for the most part, the office’s unenviable reputation is engrained and enduring.
A Fundamental Tension (Continued)
It is, therefore, no mystery that the same fundamental tension that characterizes attitudes toward the vice presidency also extends to vice presidential candidates. On the one hand, political observers and practitioners recognize the limited powers of the vice presidency, and so they are duly cautious about overstating the running mate’s (likely) influence on presidential voting. Often, they do so by repeating well-worn aphorisms to downplay veepstakes speculation—referring to it merely as a “parlor game,” for instance. Or they cite the conventional wisdom that “Vice presidential candidates can’t help you, they can only hurt you,” and that “People don’t vote for a vice president, they vote for a president.” We do not argue that such folk wisdom is wrong, necessarily; in fact, some of the empirical evidence that we present in this book would tend to support these claims. Nor do we contend that expressing such skepticism is insincere or contradictory, on its face. But it is important to recognize that these sentiments often exist in tension with other statements or behaviors that directly express or clearly imply a perception that running mates are electorally significant and potentially helpful—even decisive. In chapters 1 and 2, we provide evidence of such tension in the attitudes expressed by presidential candidates and voters, respectively. For now, consider one prominent and consequential example.
When asked to describe his criteria for selecting a running mate in July 2008, Republican presidential nominee John McCain said: “First, you want to make sure you have a candidate that’s not going to hurt the ticket” (J. Mason 2008). In 1996, when discussing then-Republican nominee Bob Dole’s selection of a running mate, McCain expressed essentially the same view by saying that the selection process sometimes “brings you the person who might not necessarily help you the most, but hurt you the least” (Pittman 1996). At the same time, he indicated that a running mate might confer significant electoral benefits by suggesting that Dole select someone who appealed not only to Republican voters but also “to those that make the difference between winning and losing campaigns”—presumably, Independents.
Yet, when it came time to select his running mate, in 2008, McCain passed over the quintessential “do no harm” candidate, in Minnesota governor Tim Pawlenty (the runner-up), and instead accepted his campaign advisers’ recommendation to choose a risky but potentially “game-changing” vice presidential candidate, in Sarah Palin. McCain made this choice after meeting the Alaska governor only for the second time, and following a rushed seventy-two-hour vetting process (see chapter 1). This hardly fulfilled his top criterion: “mak[ing] sure you have a candidate that’s not going to hurt the ticket.” In fact, following her rocky vice presidential campaign, Palin was widely perceived as hurting the ticket in exactly the way that McCain previously had warned against—a perception validated by subsequent empirical analyses (e.g., Court and Lynch 2015; Elis, Hillygus, and Nie 2010; Knuckey 2012).
McCain illustrates the fundamental tension apparent in many people’s—including presidential candidates’—expressed views regarding running mates’ electoral significance. He repeatedly stated, and he genuinely might have believed, that the overriding principle of vice presidential selection is “First, do no harm.” But, as is the case with matters of public opinion more generally (Zaller 1992), McCain’s judgments on the matter seem to have drawn on a mix of considerations. Thus, when pressured by campaign advisers to adopt an alternative approach, he sampled from those conflicting considerations to arrive at a decision fundamentally opposed to his oft-stated conviction. In other words, no matter how confidently McCain stated this conviction in public, and when discussing vice presidential selection in the abstract, to some extent he also believed that it was possible—maybe just in these special circumstances—that a running mate could yield transformative electoral benefits, and that the chances of this happening in 2008 were good enough to justify picking Sarah Palin.
What to Believe?
It is ironic that McCain, of all people, would stand out as the presidential candidate who most obviously threw a “Hail Mary” pass, with his choice of a running mate, in order to win an election. Yet, because he at least entertained principles directly in conflict with his oft-stated, and quite possibly genuine, philosophy of vice presidential selection, we suspect that McCain is not a hypocritical outlier but an exemplar of the fundamental tension that characterizes much of public discourse regarding vice presidential candidates and their electoral significance.
Indeed, McCain’s is the same type of conflicted opinion that we see in the journalist (Chris Cillizza) who declares that “the vice presidential pick—viewed through the lens of history—has almost no broad influence on the fate of the ticket and, to the extent the VP choice has mattered, it’s been in a negative way”—yet, when ranking veepstakes contenders, regularly dangles the prospect of a decisive home state advantage (Devine and Kopko 2016, 14). Or the scholar (Stuart Rothenberg) who scorns veepstakes speculation as “a game” to be played at “cocktail parties or around the kitchen table” before condescending to remind readers that this year’s election actually is between two presidential candidates—yet in other writings, indeed in the same year, plays precisely that game by rating contenders largely on the basis of their ability to deliver swing states or the party’s base (13). Or the presidential candidate (George W. Bush) who writes, “I believe voters base their decision on the presidential candidate, not the VP”—yet also speculates that several other vice presidential finalists on his list might have delivered their home state in the general election (see chapter 1).
Like McCain, these actors state their convictions about running mates’ electoral effects definitively—with little, if any, hint of nuance. And, like McCain, it is entirely possible that they have expressed their convictions sincerely. (Bush, for example, did choose a running mate with little electoral appeal and no prospect of delivering a swing state.) But there is also good reason to believe that they, like many other actors in the electoral process that we analyze in later chapters, are more conflicted about running mates’ electoral significance than they let on. In reality, they probably entertain a mixture of views and therefore, under a particular set of circumstances (e.g., a close election, a popular governor, a divided party), might draw on considerations that conflict with their prevailing philosophy to determine that a (potential) running mate really could make a difference in the election—maybe this one time.
That is why one must be careful not to take the conventional wisdom about vice presidential candidates at face value, no matter how frequently or confidently it is publicly pronounced. As is true of the vice presidency itself, evaluations of the running mate’s electoral significance are complex, and often they reveal, on close inspection, quite a bit of internal conflict, especially when moving from the abstract to the particular.
In this book, we take nothing about running mates’ electoral significance for granted. Whatever our preconceptions about their effects on presidential voting—and it would be fair to describe us as skeptics—our commitment is to let the empirical data speak for themselves.
Do Running Mates Matter?
Recognizing the fundamental tension in public attitudes toward vice presidential candidates, described earlier, in this book we ask: Do running mates matter? That is to say, do they influence voting in presidential elections? And, if so, how? Do running mates influence voters, in general, or only targeted subsets of the electorate? Do they, in fact, “deliver” their home state or region? Affiliated demographic groups? Ideological allies? And, to the extent that running mates influence elections, is this because voters actually cast votes for (or against) a vice presidential candidate? Or is it because running mates help to shape voters’ perceptions of the presidential candidate who selected them, thereby exerting an indirect effect on vote choice?
To emphasize the practical significance of this research, and the manner in which we present it, our fundamental research question may be reformulated as follows: Do running mates do what people—namely, presidential candidates and voters—think they do, electorally speaking? Our objective is to answer this question, by testing perceptions of the running mate’s (potential) electoral significance against the relevant empirical data. To do this, we divide our research into two essential and integrally related parts. In the following, we briefly describe each chapter’s methods of analysis and its empirical results.
Part I: Perceptions of Running Mate Effects
Part I (comprising chapters 1 and 2) analyzes perceptions of the running mate’s electoral significance. In particular, we evaluate, first, whether, and to what extent, presidential candidates and voters believe that running mates influence vote choice in presidential elections; and, second, what criteria—particularly in terms of electoral versus governing considerations—they use to evaluate (potential) vice presidential candidates. The purpose of this analysis is to set the agenda for the empirical analyses that follow in part II. In other words, we seek to establish—based on systematic evidence, rather than mere assumption—that it is relevant to ask whether running mates matter in the first place. Moreover, we seek to identify the nature of these perceived electoral effects. Having done so, in part II we can test whether running mates matter in the way that people—particularly, those who select and elect them—think they do.
Chapter 1: (Why) Do Presidential Candidates Think That Running Mates Matter?
In this chapter, we use qualitative evidence from the 1976–2016 elections (i.e., the era of the modern vice presidency) to evaluate presidential candidates’ perceptions of running mate effects. In particular, we seek to determine whether, and why, presidential candidates think that running mates have the potential to influence election outcomes. We have gathered evidence for this analysis from a diverse range of sources, including public speeches and interviews, media coverage, personal memoirs, oral histories, and archival materials from presidential or other public libraries.
Our analysis indicates that most presidential candidates perceive vice presidential candidates to be electorally consequential—and, in some cases, determinative—but in public they downplay or deny such considerations, so as to focus on governing qualifications. In fact, presidential candidates in nearly all recent elections have publicly communicated a remarkably consistent set of selection criteria: that running mates must be qualified to serve as (vice) president, first and foremost; next, they must be personally and politically compatible with the presidential candidate; finally, as something of a bonus, they may provide a modest electoral advantage. But privately, or in subsequent public comments, many presidential candidates emphasize electoral considerations quite a bit more, and governing considerations less, than they do in public during the campaign.
While we cannot generalize across all selection processes, it is fair to say that most presidential candidates think that running mates matter, electorally speaking, and might even prove decisive in a close race. Such perceptions have the potential to influence the actual selection of a vice presidential candidate—and, in turn, who serves in office as vice president.
Chapter 2: (Why) Do Voters Think That Running Mates Matter?
In this chapter, we analyze public polling data on vice presidential selection, mostly from the 2000–2016 presidential elections. The purpose of this analysis is to determine, first, whether voters think that running mates influence their votes, or election outcomes more generally, and second, what criteria voters use to evaluate (potential) running mates, such that we might characterize the nature of their electoral appeal.
Our analysis indicates that voters have mixed, or conflicted, perceptions of running mate effects. Generally, survey respondents affirm the importance of vice presidential selection, in the abstract, and in many cases they report being more or less likely to vote for a given presidential candidate based on his or her choice of a running mate. Yet, at the same time, respondents rate vice presidential selection as less important than nearly all other electoral considerations; fewer than one in ten respondents say that a running mate ever has changed their presidential vote; and when given the opportunity to explain, in their own words, why they support or oppose a particular presidential candidate, few, if any, respondents cite the candidate’s choice of a running mate.
Also, voters seem not to have fixed criteria in mind when evaluating (potential) vice presidential candidates. However, they do prefer running mates who balance a given ticket by compensating for the presidential candidate’s perceived deficiencies. This is evident with respect to attributes of professional experience, in particular, but not demographic characteristics. Indeed, voters seem to value the running mate’s qualifications very highly, and for the most part they do not give credence to electoral considerations.
Part II: Evidence of Running Mate Effects
Part II (comprising chapters 3–5) tests perceptions of the running mate’s influence, as established in part I, against the relevant empirical evidence. Essentially, our objective in these chapters is to evaluate whether running mates matter in the way that relevant political actors think they matter. Each chapter examines a distinct—but not mutually exclusive—process whereby running mate effects might occur: first, by directly influencing vote choice among voters, in general (chapter 3); second, by directly influencing vote choice among targeted subsets of voters (chapter 4); third, by indirect means—that is, by influencing voters’ evaluations of the presidential candidates, which, in turn, directly influence vote choice (chapter 5). For this analysis, we use a diverse range of data sources and research methods. Indeed, the depth and breadth of our analysis of vice presidential candidates’ electoral influence are unprecedented in the political science literature.
Chapter 3: Direct Effects
In this chapter, we evaluate the running mate’s direct effect on voting, generally. First, we do so by providing descriptive statistics on presidential versus vice presidential candidate preferences, based on data from the 1968–2016 American National Election Studies (ANES). Next, using the same data, we estimate the relative influence of vice presidential versus presidential candidate evaluations on vote choice, via logistic regression analyses. Finally, and for the first time in the literature, we test the causal effects of dynamic changes in running mate evaluations (i.e., favorability ratings) on intended vote choice, as well as presidential candidate evaluations, over the course of a campaign. We do so using time series (i.e., rolling cross-sectional) data from the 2000 and 2004 National Annenberg Election Survey (NAES).
Our analyses indicate that running mates do, in fact, directly influence vote choice, but only to a limited extent. Indeed, vice presidential candidate evaluations have much less influence on vote choice than do presidential candidate evaluations. Furthermore, our vector autoregression analysis indicates that while running mates can influence intended vote choice during a campaign, in most cases their effects last only for a few days. This analysis also indicates that presidential and vice presidential candidate evaluations are interdependent (or endogenous) over time. Such direct evidence of interdependent intraparty candidate evaluations is important because it shows that running mates are not just shadows of the presidential candidate. Rather, voters view vice presidential candidates, in part, as a reflection on the presidential candidate—such that reevaluating the former may cause them to reevaluate the latter.
Chapter 4: Targeted Effects
Perhaps, then, running mates are most effective at influencing vote choice among particular groups of voters that presidential campaigns may wish to “target” in order to win an election. To evaluate this possibility, in chapter 4 we examine running mates’ “targeted effects” on vote choice among groups of voters with whom they share a salient geographic (i.e., home state or region), demographic (i.e., gender, religious), or ideological (i.e., liberal, conservative) identity. For example, we assess whether women were more likely than in other years to vote for the Democratic ticket in 1984 or the Republican ticket in 2008, both of which featured a woman running mate (Geraldine Ferraro and Sarah Palin, respectively). Our analysis draws on three distinct, high-quality data sources—the American National Election Studies (1952–2016), the National Annenberg Election Studies (2000–2008), and The American Panel Survey (2012–2016)—and estimates running mate effects by using a multimethod approach that includes linear or logistic regression analyses, for cross-sectional data, and an adaptation of Lenz’s (2012) three-wave test, for panel data.
We find little evidence of targeted running mate effects. For instance, cross-sectional data indicate that Catholics were no more likely to vote for the Democratic Party in 2016, when Tim Kaine was the vice presidential nominee, than in previous elections. And panel data indicate that the effect of Catholic identification on intended vote choice did not change from the period before versus after Kaine’s selection. Likewise, we observe no significant change in women’s voting behavior in response to the Ferraro or Palin selections. The only clear evidence of a targeted effect comes from 2012, when conservative support for the Republican ticket significantly increased following Mitt Romney’s selection of Paul Ryan as his running mate, and ultimately influenced vote choice. In all other cases, we see no such effects at any point during the campaign, or, at best, a temporary increase in support that fades away by Election Day.
Chapter 5: Indirect Effects
In chapter 5, we expand the scope of our analysis to include indirect running mate effects—or the effects of vice presidential candidate evaluations on presidential candidate evaluations, which, subsequently, influence vote choice. This, we argue, is the most realistic conception of running mate effects. Yet indirect effects have gone almost entirely unexplored in the scholarly literature, to date (two exceptions are found in Kenski, Hardy, and Jamieson [2010] and Romero [2001]; but see chapter 5, note 6). In this chapter, we provide an unprecedented analysis of indirect running mate effects, using data from the ANES (1968–2016), NAES (2000–2008), and Knowledge Networks (2008), and a multimethod approach comprising logistic regression analyses and structural equation models (cross-sectional data), vector autoregression (rolling cross-sectional data), and an adaptation of Lenz’s (2012) three-wave test (panel data). Moreover, we evaluate indirect effects based on a diverse range of candidate evaluations, including ones pertaining to ideology, experience, and various professional or personal attributes, as well as general favorability.
Our analysis of more than two hundred statistical models provides overwhelming evidence that running mates influence voters’ perceptions of the presidential candidate who selected him or her. We also present structural equation models demonstrating that running mate evaluations indirectly influence vote choice. In other words, running mates have indirect, as well as direct, effects on voting—although the former appear to be much stronger than the latter. To provide one example of such an effect, in 2008 respondents were significantly more likely to approve of John McCain’s judgment if they believed that his choice of a running mate, Sarah Palin, was ready to be president. Specifically, respondents who rated Palin as “extremely” ready to be president, versus “not at all,” rated McCain’s judgment 2.2 points higher on a scale of 0 to 10. This, in turn, decreased the respondent’s likelihood of voting for the Democratic ticket by 7 percentage points, according to our structural equation models. Nor are these effects limited to perceptions of judgment. Indeed, we find strong and consistent evidence that respondents’ perceptions of the presidential candidate—across a wide range of attributes relating to leadership skills, trustworthiness, and competence—are shaped by their evaluation of the running mate, in terms of overall favorability or experience.
Chapter 6: Why Does This Matter?
This book’s final chapter emphasizes the practical implications of our research findings by discussing several key takeaway points that may help to better inform future deliberations over vice presidential selection among political practitioners, journalists, and the public at large. We present these takeaway points as five recommendations to presidential candidates and their campaigns, when engaging in vice presidential selection. Our recommendations are as follows:
1. Pick someone who can be a good vice president.
2. Don’t just say it; mean it.
3. Ask whether the running mate will matter enough.
4. Don’t expect the running mate to “deliver” a key voting bloc.
5. Don’t just take our word for it.
In each case, we summarize the research findings that inform our recommendation and discuss their implications for vice presidential selection, presidential campaign strategy, and presidential administration. In the course of this discussion, we also consider the limitations of the present research, opportunities for future research, and the role that vice presidential candidates may play in the 2020 election.
So, What if Running Mates Matter?
To be sure, we are not the first scholars to analyze the electoral significance of vice presidential candidates. The existing literature on “running mate effects,” in fact, attests to the importance of a subject that easily can be mistaken as insignificant or even frivolous. And this literature provides valuable perspective—in terms of theory, methodology, and substantive conclusions—that helps to guide our research, while also raising questions that we hope to answer, with a greater measure of clarity, in the pages that follow.
Scholarly Literature
Numerous studies examine running mates’ effects on elections generally (e.g., Adkison 1982; Burmila and Ryan 2013; Devine and Kopko 2016, chap. 8; Grofman and Kline 2010; Ulbig 2010; Wattenberg 1984, 1995; Wattenberg and Grofman 1993), and with respect to specific considerations such as party identification (Court and Lynch 2015), ideology (Court 2012; Krumel and Enami 2017), demography (Jelen 2018), and geography (J. Campbell 1992; J. Campbell, Ali, and Jalalzai 2006; Devine and Kopko 2011, 2013, 2016, 2019; Dudley and Rapoport 1989; Garand 1988; Heersink and Peterson 2016; Holbrook 1991; Kahane 2009; Mixon and Tyrone 2004; Morini 2015; Rosenstone 1983; Schultz 2016; Tubbesing 1973). We review these literatures in the relevant chapters to follow. Also, many studies examine secondary matters that we reference in this book, including media coverage of vice presidential candidates (e.g., Ulbig 2010, 2013) and the vice presidential selection process (e.g., Baumgartner 2012, 2016; Hiller and Kriner 2008; Sigelman and Wahlbeck 1997). But the existing literature has two significant limitations.
First, many of these studies narrowly focus on one aspect of running mate effects, such as geography (e.g., Devine and Kopko 2016) or media coverage (e.g., Ulbig 2013). Second, other more comprehensive studies analyze vice presidential candidates, as an electoral institution, within the broader context of the vice presidency, as a governing institution, and with a predominant focus on the latter (e.g., Baumgartner with Crumblin 2015; Goldstein 2016). Such studies represent tremendous contributions to scholars’ understanding of the vice presidency, which we make no attempt to challenge or to significantly revise here. However, these studies do not provide a comprehensive, empirically driven analysis of running mates’ electoral effects, along the lines of what we present in this book.
In fact, no book to date has been devoted exclusively to the subject of vice presidential candidates’ effects on presidential voting, in general or along several dimensions at a time, such as ideology, geography, and demographics. Also, ours is the first study to systematically examine voters’ perceptions of vice presidential candidates’ electoral influence and their criteria for vice presidential selection (chapter 2). Finally, ours is the first study to analyze how voters’ perceptions of a vice presidential candidate’s characteristics (e.g., readiness to be president) influence their perceptions of the presidential candidate’s characteristics (e.g., judgment), as well as how these perceptions may influence vote choice (chapter 5).
Running Mates and the Vice Presidency
A clarification of terms is in order, also, before proceeding with this analysis. The title of this book, and our central research question, asks: Do running mates matter? We are not asking whether vice presidents matter. Goldstein (2016), in particular, has answered that question rather definitively, and in the affirmative. Although vice presidents have little constitutional power, since the inception of the “modern vice presidency,” under Jimmy Carter, they have wielded significant and growing power as a result of informal institutional changes. In particular, as Goldstein explains, this power comes from serving as a senior adviser to, and troubleshooter for, the president, with the support of extensive personal access and in-house resources. Indeed, many of the most recent vice presidents—including Al Gore, Dick Cheney, and Joe Biden—have played a major role in shaping administration policy on foreign and domestic matters, and in advancing the president’s agenda through their work with Congress and foreign leaders.
Unfortunately, we think, it is all too common—particularly among journalists—to treat the terms “vice president” and “vice presidential candidate” interchangeably. Needless to say (but we will, anyway), the two roles are different, and they coincide only when an incumbent vice president is seeking reelection. In fact, we see the study of vice presidential candidates, as an electoral institution, as quite distinct from the study of vice presidents, as a governing institution (although there is good reason to draw relevant connections between the two at times, in the same way that, say, studying judicial nomination and confirmation processes is connected to, but distinct from, studying judicial behavior). That is why, as noted previously, we explicitly characterize our work as a study of vice presidential candidates, not vice presidents. And it is why, in hopes of limiting confusion about our subject matter and research objectives, we emphasize that distinction by framing our title, research question, and much of the language to follow in terms of “running mates.”
This discussion also provides a useful reminder as to why it is important to know whether, and in what ways, running mates matter. Regardless of their electoral influence, running mates ultimately matter because, if successful, they become vice presidents. And, as the research cited earlier demonstrates, vice presidents are highly, and increasingly, influential actors in American government. If it is the case that presidential candidates (at least sometimes) misjudge the nature of running mates’ effects on presidential voting—perhaps by overestimating their ability to “deliver” a home state or an affiliated demographic group—then they might select someone who is unqualified, or at least less qualified than other credible alternatives, to serve as a partner in government and next in the line of presidential succession, simply because an electoral consideration tipped the scales. This may seem like a remote possibility, but—perhaps depending on one’s political views—it is not difficult to think of a time when it nearly happened or actually did (e.g., Dan Quayle, John Edwards, Sarah Palin).
To the extent that our research validates some perceptions of running mate effects, perhaps it will help to inform presidential candidates, their campaigns, and members of the news media when gaming out viable electoral strategies. But, to the extent that our research challenges errant or oversimplified perceptions of running mate effects, then perhaps it will help to divert attention away from illusions of electoral advantage and redirect it toward efforts to identify the person best qualified to serve as vice president.
