Abstract
Both the Republican and Democratic parties are internally divided. Each contains a party regular wing, which is interested in winning office and in the compromises necessary to govern. And each contains an ideological wing, which is interested in close adherence to the core coalition of the party. But the nature of the cleavage is very different within the parties. Among Democrats, the cleavage is mild, with most members belonging to the party regular camp, to the chagrin of ideologues, who are for the most part Bernie Sanders supporters. The cleavage among Republicans, though, is so deep that the party could not find a way to bridge it in the so-called invisible primary for 2016, creating an opening for Donald Trump, who is from neither camp.
It is fashionable to note that both the Republican and Democratic parties are “divided,” but it is only trivially true that divisions in the parties exist. Political parties are coalitions of actors who have interests that differ. What is less clear is just how the parties are divided. With diverse parties, there are many and diverse cleavages. Some, however, are more significant than others. In particular, the cleavages that matter are those that present challenges to the party leadership.
This article attempts to articulate one important way in which the modern parties are divided, which is different than most previous party cleavages. Whereas past cleavages have divided actors who want very different things from politics, today’s parties are divided between actors who mostly want similar things but who differ on how to pursue them.
There are coherent divides in both the more conservative Republican Party and the more liberal Democratic Party. Among conservatives, for instance, there are differences between social traditionalists and free-market libertarians—and within liberalism there are divides between anticapitalists and antiracists. These differences are important—and in many ways very important to the 2016 election cycle—but both parties have largely held together over such divisions. Partly this is because activists in each of the parties tend to agree about issues on both sides of the divides, even as they differ over priorities. What splits parties, though, are rifts between ideologically pure, less compromising members, and more pragmatic, moderate ones. This is a difference not about policy ends but about political means. Not about issues but about strategy. Not about what to do but how we should face the constraints of trying to do it.
Identifying these divides is complicated by the fact that they are not manifest in the same ways in each party. Political parties are complex and mostly disorganized institutions. V. O. Key—who focused on the party in the electorate, the party as an organization, and the party in government—did not know the half of it. An expansive definition of each party would include allied interest groups and activists, who may or may not agree with the formal party organization.
Here, I focus on two groups—elected officials and party activists—and in some cases, I compare them to what we know about the preferences of a third group, voters.
Parties are Coalitions
The prevailing view of political parties, and especially the two major parties in the United States, is that they are coalitions. 1 It can be useful to think of the members of these coalitions in different terms, but following Bawn et al. (2012), 2 I think the most traction comes from viewing them as comprising different social groups. Individual politicians of course represent more than one group, and many groups are divided. But the parties can still be seen as coalitions of, for example, bankers, religious traditionalists, and foreign policy hawks, among others; or of nonwhites, union members and atheists, among others.
But precisely because everyone is a member of more than one group and groups can be divided, it is trickier to nail down empirically what the exact coalitions are. Party platforms do not say “We are for the interests of Latinos,” or “We are for Wall Street.” The party’s agenda in Congress emerges over several bills, and members vary in how much they want their policies to pass and in how much they just want someone else’s policies not to pass. Candidates for many offices are often closely related to a single group, rather than representing the entire coalition (Bawn et al. 2015).
We can see the coalition being shaped during presidential nominations. The presidency is uniquely important, but it is also simply unique. In Congress, it is possible for a candidate to represent only a narrow group, and the coalition only emerges when the elected officials assemble in the legislature. Not so with the presidency, which involves players across the entire party. The eventual nominee will, in some respects, stand in for the terms of the party’s coalition.
Coalitions are assembled principally by party leaders (Karol 2009), and political activists pressure the leadership to adopt the positions they adopt. Partisan voters may diverge from the coalitions from time to time and force party leaders to react in some way, but the coalitions themselves are shaped by political elites—the party leaders and activists who sort out and articulate the issues around which the parties coalesce. Accordingly, my focus here is on those political elites—the leaders and activists in the upper ends of the parties.
Ideological coalitions in the two major parties
The party coalitions are constantly evolving. The current coalitions can be well described as “liberal” and “conservative,” but this was not always the case. Only a half-century ago, the parties each included a significant number of liberals and conservatives.
Party coalitions need not coincide with ideological clusters at all. At the end of the nineteenth century, the Republican Party included protariff businessmen, those with free-silver mining interests, and blacks. These interests were often wildly in conflict with one another, necessitating clever maneuvering by party leaders to hold the party together. In the 1940s and 1950s, the New Deal coalition of the Democrats brought together the northern union members, Catholics, Jews, and ethnic whites, but also southern white segregationists. Republicans included those with business interests as well as Protestants, women’s rights groups, and groups on both sides of the emerging conflict over civil rights.
Elsewhere (Noel 2012, 2013), I argue that this development began with the sharpening of a liberal ideological coalition in the early part of the twentieth century followed by a conservative coalition in response. These coalitions brought together the interests, ideas, and policy preferences that we today associate with liberals and conservatives. Liberals favor government economic intervention to encourage equality and labor interests; policies that advantage ethnic, religious, sexual, and racial minorities and disadvantaged groups; women’s rights; a multilateral and often less militaristic foreign policy; and a collection of many other positions. Conservatives favor free markets, business interests, a color-blind approach to race and ethnic issues, traditional religious and sexual norms, a foreign policy informed by American exceptionalism, and a number of other positions. I argue that it makes sense to think of these ideological movements as coalitions because they bring together potentially diverse actors, cementing their bonds with appeals to shared principles, values, and even symbols.
These ideological coalitions organized politics in a different way than the existing party coalitions did, and the two kinds of coalition often pulled actors in different directions. For example, most politicians from the South were both Democrats and conservative—especially, but not only, on racial issues. They would sometimes vote according to their conservative ideological coalition, joining many Republicans, and at other times vote with their party, sticking with Democrats. The New Deal Democratic coalition was held together with an understanding that southern Democrats would back labor legislation they disagreed with in exchange for northern Democrats not pursuing action on civil rights (Feinstein, Schickler, and Pearson 2010; Feinstein and Schickler 2008; Schickler 2016). Some votes were party-line votes. On other votes, a “conservative coalition” of southern Democrats and most Republicans voted against the liberals.
Today, the ideological coalitions are nestled comfortably inside the party coalitions. This transformation is behind most of what we observe as polarization. Scholars who argue that polarization is merely sorting (Fiorina 2005; Levendusky 2009) are right that one of the main things that has happened is that ideologues have sorted into their new parties. However, such sorting has led most actors, especially activists but even many voters, to adopt more ideologically consistent views (Abramowitz 2010; Abramowitz and Saunders 2005).
At the same time, the ideological core of the party is not enough to win elections. This is particularly true because, while voters are more ideologically sorted and more likely to express ideologically consistent beliefs than they used to, most are still not ideological. Even those who self-identity as liberal or conservative do not always hold beliefs associated with that label (Converse 1964; Ellis and Stimson 2012; Kinder and Kalmoe, forthcoming).
Partisan activists and elites, too, sometimes deviate from the ideological coalition for a variety of reasons. For a long time, Democratic Party leaders were hesitant to lead on gay rights, for example, even though this was the liberal position. The electoral needs of the party can override its ideological goals. It is thus wrong to say that the ideological coalitions fully coincide with the party coalitions. Rather, the party coalitions contain the ideological ones.
Ideologues versus regulars
It is common to describe the tensions within each party as between “moderates” and “extremists,” but this is not quite correct. “Extremists” are not necessarily further out on an ideological dimension. They are just much more consistent ideologically. Likewise, moderates are usually just less consistently liberal or conservative. They may, in fact, hold more extreme views on the issues that they care about, but do not hold the same views as other liberals or conservatives (Noel 2013, 73; Broockman 2016; Ahler and Broockman 2016). But any two moderates are often not the same. A prochoice, antiwar, anti-affirmative action, antiminimum wage voter is called “moderate,” and so is a prolife, prowar, black-lives-matter, fighter for fifteen. Some moderates are truly “in the middle,” but unless they hold few policy positions, they are better described as off-dimensional.
Seeing ideologues as simply the core of an ideological coalition and moderates as those who deviate from that core allows us to escape the pervasive ideological spatial model. This model has been very useful for political science, but it is also limiting. As other articles in this volume indicate (Bartels; Cohen, McGrath, Aronow, and Zaller), the most common prediction from the model’s application to elections—convergence on the median voter—does not regularly occur.
But a model of a party coalition with an ideological core would not make such a prediction. 3 It is likely that election-seeking politicians may need to deviate from ideological purity in many cases, especially when competing in a national electorate. But politicians in especially homogenous constituencies might do better by appealing to the largest block of ideologically consistent voters—ideologues.
There are four reasons why a politician, activist, or voter might deviate from the ideological core of his/her party. By deviate, I mean actively oppose elements of the coalition. Most people do not care passionately about every part of their party’s platform, nor its ideological core. But it is one thing to be indifferent on all but a few issues, something else to outright disagree.
1. They might simply not be ideologically sophisticated. Most voters and even many politicians are like this.
2. They might disagree with the ideology on one or another issue. It is common to hear that someone is “liberal on most things, but not on this.” Otherwise conservative business leaders, for instance, may depart from the conservative opposition to more relaxed immigration policy because they see the value in inexpensive labor.
3. They might find that some of the policies advocated by their ideological allies are not politically practical or implementable as policy.
A pair of columns by liberal blogger Matt Yglesias nicely illustrates this phenomenon. Yglesias writes on March 10, 2016, that he is persuaded by Bernie Sanders that universal free education is a good idea. The next week, he writes that he opposes Sanders’s plan for the same, not simply because it would be hard politically to implement, but because the current state of higher education in the United States has been built around a different model. The change would require too many other changes to that environment, which the plan has not worked out (Yglesias 2016a, 2016b). In other words, reasonable ideologues can differ about what the best policy is, even while agreeing on goals.
This distinction is similar to the tension often raised in the 2016 Democratic contest between Bernie Sanders’s and Hillary Clinton’s respective theories of change. Sanders is a revolutionary. Clinton is an incrementalist. It is common to call incrementalists moderates, but that is a function of our commitment to the ideological spatial model, which requires us to array preferences in a space.
4. They might be interested in winning office, and thus need to appeal to activists and voters who are in any of the first three categories.
Liberal ideological activists were ready to expand gay rights much earlier than most of the American public was. Democratic politicians who shared their goals were thus more cautious on the issue. After their loss in 2012, Republicans concluded that they needed to reach out to Latino voters (Barbour et al. 2013). Despite opposition from their conservative base, many Republicans have since looked for a way to reform rather than curtail Latino immigration. However, some election-minded politicians might not face this incentive. If your constituency is Texas or Vermont, you may lose more votes from your ideological core than you gain from those outside it.
All of these deviations from the ideological core could, in theory, create what looks like a thousand fractures, as those compelled to compromise do so on a variety of issues that change from context to context. There are important potential fractures crisscrossing both parties. However, the third and fourth types especially will tend to come from the same people. Some people are simply more inclined to find more workable policy solutions or to expand the coalition.
Thus, Republicans who oppose aggressive attempts to deport undocumented workers are (a) reflecting business conservatives who differ from social conservatives on this issue, (b) acknowledging that these approaches are unworkable, and (c) hoping to avoid alienating Latino voters.
We can call these people “moderates” if we like. But thinking of them instead as deviations from an ideological core allows us to acknowledge that politicians such as Jeb Bush, Paul Ryan, Marco Rubio, and John Kasich have very conservative records on many policy questions, and that politicians such as Barack Obama, Harry Reid, Nancy Pelosi, and Hillary Clinton have pursued very liberal goals. All of these politicians have been described as both moderates and extremists. 4
In this article, I describe them as party regulars or compromisers. They are similar to the pragmatists described by Polsby and Wildavsky (1968), but as evidence suggests, such pragmatists are increasingly idealistic as well (Layman et al. 2010). They are also sometimes called insiders or the establishment, which I discuss below.
Regulars are opposed by ideologues. Ideologues may often be more “extreme,” but they are also simply less likely to deviate from the ideological core. Politicians such as Ted Cruz, Rand Paul, 5 Rick Santorum, and Bernie Sanders are ideologues.
Ideologues versus outsiders
The regular-ideologue divide is similar to—although theoretically distinct, from—what is termed an insider-outsider conflict. Insiders are technically whoever is in power in the party. There are multiple definitions of an outsider. Sometimes, an outsider is a politician who is merely not powerful in the party, someone who might be a back-bencher in a parliamentary system. The Tea Party comprises (or claims to comprise) outsiders trying to get in.
But we also sometimes see another group of ultra outsiders—people who have held no elective office at all. Herman Cain, Ben Carson, Carly Fiorina, and Donald Trump are such outsiders. This makes the terminology vague. For the sake of clarity, I reserve the term outsider for these ultra outsiders.
For disruptive politicians who are still inside the party—people like Ted Cruz—the term ideologue is usually accurate. This does not mean that the ideologues have succeeded in taking control of the Republican Party. The House Freedom Caucus members still claim to be the outsiders. And they have to date not been successful in nominating a true ideological conservative for the presidency, nor in selecting one for the top leadership positions in Congress. But they are not outsiders in the same way that Donald Trump is. Trump is not representative of the ideological wing of the Republican Party. But Cruz is. In short, ideologues are often called outsiders, but I try to distinguish between the two concepts.
Measuring Compromise
It is particularly difficult to measure the willingness of politicians to compromise. Often, a compromise will look like “moderation” to most measures of preferences. It can be conceptually difficult to distinguish between compromise and sincere deviation from a pure ideological position.
Here, I focus on three arenas of compromise. First, I look at the action of elected officials, primarily by examining
Elected officials and nominate scores
Data on elites at the very top of the party hierarchy can be easy to come by. The familiar
The creators of The meaning of the second dimension has largely shifted from representing regional differences within the parties (e.g., between northern and southern Democrats) to intra-party divisions that are more subtle and less clear. One of these divisions appears to be an “insider vs. outsider” cleavage that pops up on votes such as raising the debt ceiling, domestic surveillance, and government funding bills.
The issues mentioned are a mix of issues of compromise and issues on which conservatives might disagree. Government funding bills almost certainly include a lot of spending that small-government conservatives would dislike, but such bills are the natural home of logrolls. Domestic surveillance represents another legitimate divide within the conservative ideology. The debt ceiling votes pit party regulars who want to maintain the integrity of the government’s past decisions against ideologues who need to take a public stand against government spending. 6
I have elsewhere argued (Noel 2013, 2014) that both dimensions are a combination of party loyalty and liberal-conservative ideology. The issues that are particularly ideological load on both dimensions, such that having high scores on both dimensions makes one conservative. Votes that involve many issues in a logroll, and especially procedural votes, cut the other way, such that Republicans are high in the first dimension but low on the second. This means that the second dimension captures the degree to which one favors his/her ideology over the procedural demands of the party, which can also be interpreted as the compromise dimension.
Activist survey
Activists are the labor force of political parties. Candidates need votes, but they also need volunteers and donors. Future party leaders often come from the ranks of activists. But the views of activists are often understudied. We have mass surveys for voters, and we can look at the behavior of politicians in office. Activists are a population that must be targeted as well. Surveys of convention delegates provide fruitful insight into activists, but the population is shaped by who is selected to attend the convention in any year.
To avoid these problems, I use a survey of political activists conducted by the Huffington Post via YouGov. In consultation with the author, the Huffington Post conducted three surveys of activists during the lead-up to the 2016 nomination process. The survey interviewed three separate samples of 500 Republicans and 500 Democrats on July 8–12, 2015, September 22–28, 2015, and January 14–20, 2016.
To take the survey, potential respondents cleared a set of filter questions to determine whether they were what I define as activists. Survey respondents said they had either done at least two of the following: (1) contributed money to a political candidate, (2) attended a political campaign event such as a fundraiser or rally, (3) done volunteer work for a political campaign, and/or (4) made phone calls to voters asking them to support a political candidate, or they reported having been at least one of the following: (1) a paid staffer for a political campaign or an elected public official, (2) a candidate for or someone who has held elected public office, or (3) an official in a political party (such as a local party chair or a precinct representative).
The activists in this sample thus had to clear a slightly higher bar than is often used to identify activists in mass surveys such as the American National Election Studies. Those who qualify through only the first set of criteria (about 62 percent) report having done much more than wearing a button or placing a yard sign. The second criteria (about 38 percent) are genuine politicians, albeit probably at the very bottom of the hierarchy.
The earlier waves of the survey asked a series of policy questions. They confirm the notion that, broadly speaking, there is considerable consensus on most issues within both parties. Figures 1 and 2 show the preferences of the activists on those issues in July 2015. The figures suggest a great deal of consensus on at least these issues for most of the activists. That is, most activists appear to be part of the ideological coalition at the core of their party.
Issue Preferences among Democratic Activists
Issue Preferences among Republican Activists
The biggest area of disagreement among Democrats and among Republicans is on trade. Significant proportions of both parties think major trade organizations are a bad thing, while many think they are good. And many in each party are also not sure.
Beyond trade, there are not many major divisions. On financial markets, for instance, there is no disagreement among activists. Neither is there really disagreement between the candidates. Officially, Clinton is just as concerned as Sanders about Wall Street. The claim has been that Clinton is not credible on these issues because of corporate influence. Clinton counters that she wants to be more realistic about the approach. In other words, she would compromise the pure goal of going after banks for a policy that is politically viable and technically workable.
The same is true of other issues that are important in 2016. The Democratic Party is unified on the Iraq War, gay rights, and abortion. On Iraq, Clinton has largely admitted that her support for the war was a mistake, while also claiming that she was motivated by practical considerations. The question for the other issues is not where do the candidates stand, but whether they will pursue the goals in the best way. It is unfortunate that there is no question here on racial issues. The closest is on immigration—here Democrats are very united.
In the July 2015 and January 2016 panels, all respondents were asked whether they thought it was better to have a politician who would: (1) “compromise to get things done,” (2) “stick to their principles, no matter what,” or (3) “not sure.” While this one item cannot capture all of the aspects of the compromise cleavage, it is a useful indicator. For sure, some people who say they value compromise still want the other side to compromise more, and some who say they want to stick to their principles have principles that make them compromise in practice. This is but one measure.
Figure 3 shows the distribution of compromise attitudes. There is very little change in either party between July 2015 and January 2016. Democratic activists are much more likely to want to compromise than are Republican activists.
Valuing Compromise among Democratic and Republican Activists
As we can see from Figure 4, those who value compromise do tend to be less ideologically extreme, in both parties, but only slightly. The activist survey includes very few people who consider themselves moderate, and fewer still who identify with the ideology of the other party. The difference between saying one values compromise versus sticking to principles is similar to the difference between saying “liberal/conservative” and “very liberal/conservative.” It may be that one’s willingness to compromise is even behind whether one includes “very” in their ideology answer. At the same time, a significant number of extremists also value compromise. Almost 40 percent of very liberal Democrats still favor compromise, and about 20 percent of very conservative Republicans do. About 40 percent of merely conservative Republicans want to stick to principles, as do almost 30 percent of merely liberal Democrats. Unwillingness to compromise when you are in fact more moderate (or only “merely” liberal/conservative) could mean being unwilling to compromise with the ideological core. There are not many respondents who seem to fit that description.
Valuing Compromise by Ideology
Ideology in the 2016 Presidential Nomination Contests
The presidential nomination is, as noted, an excellent place to look for evidence of partisan divisions, as different elements of the party back different candidates. What we know about the candidates—and what we can understand about why different actors back them—can yield insights into the conflicts within the party.
Both parties face the party regular-ideologue divide. In one party, the pragmatists have won fairly easily, while on the other side, the conflict has set the stage for unpredictable results.
The Republicans
The trouble for Republicans in 2016 is nicely captured by what happened on Capitol Hill in October 2015. While the preprimary phase of the presidential nomination progressed, the Republican Speaker of the House, John Boehner, felt the need to resign midway through his term. Boehner’s inability to keep the ideologues in the House in line meant he was unable to do his job.
The House then spent some weeks trying to figure out who could lead it. For a party supposedly united by (or in some accounts fundamentally shaped by) a common ideology, such fracturing is hard to explain. On most policy questions, the distance between Boehner and his principle backbenchers in the House Freedom Caucus was not enormous. But the gap could not easily be bridged.
The institutions in the House for selecting a leader are much better than those used to choose a presidential nominee. Paul Ryan, who was the eventual choice, was able to negotiate with the different factions, and party leaders could poll and meet privately with the relatively small number of actors with a stake in the decision.
Out in the country, the Republicans were dealing with the same divide. Cohen et al. (2008) argue that party leaders can and do coordinate around a single presidential candidate, enabling that candidate to win. The Republican Party failed to do that in 2016.
Figure 5 shows the distribution of party leader endorsements in the 2016 campaign in both parties. Following Cohen et al. (2008), I use only endorsements for the candidates that are made before the Iowa caucuses, which met on February 1, 2016. Ted Cruz and Jeb Bush led the pack, followed closely by Marco Rubio, John Kasich, and Chris Christie. The figure shows a deeply fragmented Republican Party. Indeed, this is the most fragmented a party’s leaders have been since Cohen and colleagues began collecting these data.
Distribution of Republican and Democratic Endorsements
The real lesson from Figure 5, however, is that there was no party favorite at all. Following the accounts of the invisible primary, some of the explanation is clear. While Bush was the top choice of many party regulars, he was despised by the ideologues, who viewed his positions on immigration, Common Core, and other issues as proof that he was not a real conservative. Meanwhile, Cruz, beloved by ideologues, was hated by party regulars, who saw him as a bomb thrower in his time in the Senate. One indication of this fault line is that Cruz drew from those holding offices in the states, mostly due to his many state legislators. Meanwhile, Bush did well among those in Washington. The state-national division reflects the ideologue-regular division. Ideologues have been more successful at the state level, where Tea Party activists have won local offices as well as a handful of offices in the national legislature. The movement has thus far only begun to crack national office.
If Cruz and Bush are favored by competing factions but unacceptable to the other faction, then Cohen et al. (2008) would expect that the party would find another candidate who could bridge the divide. Some saw Marco Rubio as such a choice. He is almost as ideologically conservative as Cruz, but he had spent a lot of time trying to build connections with party regulars as well.
We know something about the ideological locations of some of the endorsers. Figure 6 shows the
The behavior of the endorsers is predictable. More ideologically extreme endorsers—and those who are higher on the second dimension, which we saw above may capture the party regulars versus ideologues split—are more likely to endorse ideologues, while more moderate and those low on the second dimension support party regulars. Only two members of Congress endorse an outsider, and notably it is Carly Fiorina, the only outsider to have at least run for office before.
At the same time, there is some overlap between the two groups, suggesting that some members of Congress, mostly neither extreme ideologues nor extreme compromisers, did back a party regular candidate. In most cases, that candidate was Jeb Bush. Bush had a lot of appeal from the middle of the party, at least nationally. The intense opposition to Bush was not entirely from even the more ideological members of Congress, but these are the most insider of ideologues—those who have won national office. Meanwhile, Rubio had some diverse support, but not very much of it. The support for Cruz was minimal, consistent with the story that the party regulars did not trust him. Figure 6 captures the very top of the party hierarchy—those with elected office in Washington. One might infer from these figures that Bush was well-positioned to win the nomination. But Figure 7 shows the bottom of the party hierarchy, reporting the distribution of first-choice support among the activists in the Huffington Post survey. The top panel shows support in July 2015, before most candidates had dropped out. The bottom panel is from January 2016, right before the Iowa caucuses.
Support for Candidates among Republican Activists
Here, we see some evidence of the Cruz-Bush divide. Ted Cruz had the second-most support among these activists, but it was overwhelmingly from those who favor sticking to principles. Bush had significant support, but it was mostly from the smaller group of Republican activists who favor compromise.
In these data, it appears that Scott Walker was best positioned among the activists, with large degrees of support from both party regulars and ideologues. Marco Rubio’s support was only from the party regular wing.
Also notable is the high level of support for Donald Trump from among these activists. This sample is a part of the party, in the Cohen et al. (2008) and Bawn et al. (2012) perspective, albeit the very bottom rungs of power in the party. Trump’s support within most of the party is minimal, but he had some support here.
He solidified that support by January. With Walker out of the race, Cruz remained the choice of ideologues, Rubio the choice of compromisers, and Trump actually appeared to be appealing to both kinds of activist. If the only data we had were these activists, Cohen et al. (2008) might infer that the party did decide, and that they decided for Trump. Since we know that the more powerful elements of the party had not backed Trump at this point, and many persisted in opposing Trump until the very end, the inference should be that this is one more dimension on which the party failed to coordinate. Across these data, we find that:
Not one of those four candidates was able to reach more than minimally across the ideological-regulars divide. Cohen et al. (2008) would expect that the party leaders, seeing this situation, would try to find one of these candidates as acceptable. The data suggest that Rubio or Walker might have been such a candidate, but the party failed to coordinate on them.
Why? It is hard to prove, but it seems likely that part of the reason is that the attention given to Donald Trump made it hard to have the public conversation necessary to choose a champion. With Trump dominating the polls, media coverage, and the debates, it was hard to evaluate Rubio, Walker, and the other possibly unifying candidates. Party leaders were confident that Trump would fade in the same way that other outsiders tend to fade. When he did not, it was too late to coordinate.
The Republican Party is divided between regulars and ideologues. Trump is neither. As we saw above, Trump’s support among activists comes from those who would compromise and those who want to stick to principles. It is beyond the scope of this article to explain election outcomes, but this outcome has been at odds with the cleavages among the party leadership.
What Trump does indicate is that party cleavages need not be reflected among mass voters. Trump appeals to a group of largely alienated white voters. Many may self-identify as “conservative,” but they are not particularly ideological. This is why Trump’s ideological apostasies do not bother them.
It has been said 7 that conservatism is identity politics for white people. For more sophisticated conservatives, this identity politics is often expressed in terms of principles such as freedom, religious liberty, and respect for tradition (in the same way that sophisticated liberal ideologues adopt an ideology of egalitarianism and a disdain for hierarchy). But most voters are not sophisticated ideologues. Even most voters who use terms like liberty may not consistently apply such terms.
Trump also exploits the internal division on trade, noted in the previous section. He draws especially well among those who oppose trade organizations, as well among those who oppose immigration, although there is less variation on that latter issue. Neither the ideological nor the regular faction has exploited this issue as well as Trump has.
The Democrats
The Democratic Party began the nomination process with a similar cleavage to the Republicans. A group of ideologically uncompromising activists began to insist that more be done to combat Wall Street and the role of money in politics. This group found a champion in Senator Elizabeth Warren, but Warren declined to run.
A large part of Warren’s decision may have stemmed from Hillary Clinton’s overwhelming support from within the party. Refer back to Figure 5, which also shows the distribution of endorsements for Hillary Clinton on the eve of the Iowa caucus. In summer 2015, before any rivals had entered the race, Clinton’s share of support looked even more daunting. If the distribution among the Republicans does not fit the pattern that Cohen et al. (2008) found at all, the distribution among Democrats does. The party had decided for Clinton before anyone else had even thought about running.
Such overwhelming support from the party no doubt scared away a lot of candidates. As many as eighteen candidates were seriously considered among Republicans. Fewer than eight Democrats publicly tested the waters, including several who never had any traction at all.
One candidate undeterred by Clinton was Senator Bernie Sanders, technically not a member of the Democratic Party, but an independent who had caucused with Democratic legislators throughout his career.
Sanders nevertheless garnered minimal support among party leaders, as seen in Figure 5. Clinton’s support was exactly the kind of broad party support that Cohen et al. (2008) predict. In Figure 8, which shows
Support for Candidates among Democratic Activists
By January, Clinton’s support had not grown much. Clinton supporters are more likely to express their concern for compromise, while Sanders supporters are far more likely to say they wish to stick to their principles. The causal direction in this change is unclear. As the campaign wore on, clashes between the candidates centered on exactly this question. It is more than likely that some of each candidate’s supporters changed to reflect the argument being made by their candidate.
But the argument is real. Sanders does have a record of compromising on some issues. This is how he explains his support for gun rights, for instance. But on balance, the rhetoric of the campaign has been Clinton saying that she needs to compromise, sometimes with Wall Street or Republicans or others, to accomplish her goals, and Sanders saying that such compromise threatens a politician’s integrity.
There is also evidence that voters divide in this way. Dan Hopkins (2016) has shown that the one thing that best explains support for Clinton over Sanders is support for Barack Obama. Democratic voters who think that Obama betrayed them were in favor of Sanders, whereas those who were happy with Obama’s performance prefered Clinton.
This is not to say that the only division between Clinton and Sanders is on compromise. Clinton consistently earned the votes of nonwhites, while Sanders did much better among whites, but especially liberal whites. But from the perspective of policy demanders, the difference on racial issues between Clinton and Sanders was not huge. Sanders had a history with the civil rights movement, and after stumbling in his response to Black Lives Matter activists, he shifted his rhetoric on the issue. Meanwhile, Black Lives Matter has not been uncritical of Clinton—they have tied her to her husband’s incarceration policies in the 1990s. Race is an area where the cleavage is much more significant among voters than it is for party leaders. Without question, cleavages other than the compromise divide matter for the Democrats, just as they do for the Republicans.
That the Democrats have better contained their ideological faction does not mean it is not there, or that it does not play a significant role in the party. Sanders forced Clinton to discuss issues that she would not have otherwise. Clinton’s voting record is actually not that far from Sanders’s, but her campaign would not have been as far to the Left as it is had Sanders not run.
Implications
Conventional wisdom now holds that polarization is asymmetric (Mann and Ornstein 2012). Whatever is happening, it is more severe in the Republican Party. What is less clear is why this is. The lens of ideological coalitions and compromise sheds some light on the pattern. Both parties clearly have ideological cores. The Democratic primaries featured a debate over which candidate was the “real progressive,” and Clinton campaigned as a “progressive who likes to get things done.” The Republican candidates repeatedly attacked Donald Trump over his ideological heresies.
Both parties also clearly have compromisers. In the national party, regulars such as Paul Ryan and Jeb Bush or Barack Obama and Hillary Clinton are much preferred to ideologues like Ted Cruz or Bernie Sanders. Cruz, a perfect ideologue, struggled to get support from the more powerful party regular wing. Nonideologues Clinton and Trump have prevailed among voters. But the conservative ideological coalition has been much more successful in exerting influence within the Republican Party than the liberal coalition has been among Democrats. The psychological tendency for conservatives to be less likely to embrace complexity or compromise might explain why there are more ideological purists on the Right, but that would not explain the trend.
It is hard to believe that the Democrats are simply better at blocking ideologues. The party led the way in bringing intraparty democracy to presidential nominations. The McGovern-Fraser reforms were introduced after 1968 in direct response to the demands of an ideological wing of the party, and it is hard to see the modern Democratic Party institutions as particularly effective at blocking the participation of some groups.
But Democrats did, in this cycle, succeed at coordinating around a party regular candidate, however. Until recently, the conventional wisdom had been that Republicans were better at such coordination, with Democrats struggling in 1988, 2004, and 2008. Had they faced a similar problem in 2016, Sanders might look like Donald Trump.
It is possible that liberal ideologues are simply less interested in engaging. Certainly they have not been as active down-ballot as their Tea Party analogue, so far. But Sanders’s supporters were just as effective in grassroots warfare in the nomination process as Tea Party favorite Cruz; both candidates dominated the caucuses, which are particularly open to activists.
In short, we see very similar dynamics playing out in both parties, but with much less success among Democrats. None of this is to say that other issue cleavages do not matter. Racial issues, particularly as a matter of emphasis, remain important in the Democratic Party, and Sanders has also exploited the internal disagreement on trade. Republicans differ on trade as well.
One way in which voters on both sides of these or any other cleavage might become more satisfied is if the United States had more than two viable parties. Both the Sanders and Trump camps complained about how both parties’ nomination processes were not open to party outsiders, but both were compelled to run in a major party because only those parties have a shot at victory.
Even in multiparty systems, we often observe two major parties or alliances of parties. At the government formation stage, being biggest still has its advantages. These parties are made up of more party-regular type politicians, who take on the responsibilities of governance. Smaller parties pressure the major parties for policy concessions (Bawn and Rosenbluth 2006).
In a two-party system, this pressure comes from within the parties. All parties are about compromise, but parties with broad coalitions must be especially about compromise. One reason the Republicans appear to be more dysfunctional may simply be that they are overrun by ideologues, while our two-party presidential system requires party regulars.
Footnotes
NOTE:
I would like to thank participants at the ANNALS workshop at Vanderbilt University for useful comments, especially Larry Bartels, Gary Jacobson, Steve Rogers, John Sides, Michael Tesler, Lynn Vavreck, and John Zaller. Rachel Blum and Justin Rattey provided timely and useful research assistance. Natalie Jackson and Mark Blumenthal were helpful in arranging for and conducting the activist survey.
Notes
Hans Noel is an associate professor of government at Georgetown University. He is the author of Political Ideologies and Political Parties in America (Cambridge University Press 2013) and a coauthor of The Party Decides: Presidential Nominations before and after Reform (University of Chicago Press 2008).