Building Blocs: Groups and Contested Party Transformations
What do you do if you belong to a small, unpopular group that wants something from the government?
For many political scientists, the answer is obvious: join a political party. Parties are commonly understood as coalitions of groups that, by banding together, can win elections and thereby gain the power to enact the policies they prefer. If a new group—even a small and unpopular one—helps a coalition win an election, it will be rewarded with a share of the policy benefits.
But what if powerful groups in both major parties oppose the group's program? That was the situation faced by African Americans in the 1930s. The Republican Party, their nominal ally, had done little for them in decades. The Democratic Party, though sponsoring the New Deal, was not especially eager to spread its benefits to racial minorities. Its entrenched and powerful southern wing was adamantly opposed to government benefits or civil rights for African Americans. The Democratic Party of Franklin D. Roosevelt was little more promising as a coalition partner than was the Party of Lincoln.
Christian conservatives faced a similar problem in the late 1970s. They believed that abortions were legalized murder but were unable to get either party to take up their cause. The Democratic Party had become home to feminism and Republicans were the party of mainline Protestants, business people, and educated suburbanites, many of whom supported women's right to abortion. As a consequence, GOP officeholders viewed the "Christian Right" as an albatross that would sink the party.
But both African Americans and Christian Conservatives nonetheless found a way to make political parties the solution to their political grievances. Civil rights organization leaders set aside longstanding grievances with white labor unions, became their allies in the internecine fights of the Democratic Party, and managed to wrest leadership from the party's southern wing. At the 1948 national party convention, delegates passed a civil rights plank that so infuriated southern opponents of civil rights that many of them bolted to form a third party. When Democrats won the presidency without them in that year's election, most Southerners returned to the fold, but their ability to block civil rights reform was much diminished. The party continued on its path to supporting civil rights legislation, culminating with the Civil Rights Act of 1964 and the Voting Rights Act of 1965.
As with the civil rights coalition, the first important step for anti-abortion Christians was to overcome fissiparous tendencies among potential allies. As late as the mid-1970s, the religious denominations that eventually came together as the Religious Right were numerous, independent, and more apt to feud with one another on theological matters than to make common political cause. The Catholic Church was historically opposed to abortion, but its voters were mainly in the prochoice Democratic Party. A handful of Catholic operatives sought to push GOP candidates to toward anti-abortion stances, but lacked connections to voters sharing their position. But by the early 1980s, anti-abortion groups had largely overcome their coordination problems. Focusing first on Republican nominations for president and moving on to state and Congressional nominations, they became regular players in the Republican Party by the 1990s. As one indication of their influence, all but one of the Republican presidential nominees from 1980 to 2016 began their careers pro-choice but changed to pro-life as they began nomination campaigns.
These two cases of party transformation illustrate a general process. It starts with individuals and groups unable to accomplish their goals working alone, gains traction with the formation of broader coalitions, and ends with new groups in the core membership of a party coalition. This book follows the path of party transformation from its disorganized beginnings to eventual success, where success is defined by a group's ability to reliably secure the nomination of presidential candidates committed to its goals.
No scholar can understand the dynamics of American politics without understanding the two cases of party transformation described in this book. The cases are paradigmatic examples of how political parties organize neglected political grievances into core voting issues. As important, the cases highlight the role of nonparty groups in using parties to advance their agendas.
To be sure, economic issues remain at the heart of the American party system, as they have since at least the New Deal. But social issues have changed party platforms, demographic bases, and electoral maps. This book explains why and how this happened.
More precise definitions are in order. Political Parties include officeholders, candidates for office, officeholder employees, and paid party employees (such as party chairs). In V. O. Key's terms, I will study parties as party in government and the party organization, but not the party in the electorate. The groups who seek to influence government through political parties are considered separately from "parties as organizations" and "party in government" because they have different incentives. Following Karol, I define groups as "self-aware" collections "of individuals who share intense concerns about a particular policy area." Organizations are groups with formal and legal arrangements. The civil rights movement, the Tea Party, and the National Association for the Advancement of Colored People (NAACP) are groups, but only the NAACP is an organization. Although some organizations are called "interest groups," their interests need not be material. For example, members of Christian conservative groups or the Anti-Saloon League gained little in the way of material benefits. They may take positions on one narrow issue or many major issues of their day.
Party transformation takes many forms, but this book is concerned with what issues a party stands for. Since presidential nominees are the most significant national representatives of the party, I measure transformation by the issue positions of successive presidential nominees. As E. E. Schattshneider wrote, "He who can make the nomination is the owner of the party." Transformations need not be periodic or limited to extraordinary periods in our history, but parties must nominate presidential candidates committed to the new issue line-up for decades; they haven't really changed if they backslide one or two election cycles later.
Eric Schickler has recently shown that other actors in the civil rights transformation, including voters, state parties, and congressional votes, changed before control of presidential nominations changed. But no group is fully integrated into a party until it has a secure role in the selection of party nominees for the country's top office. Using presidential nominations as the measure of when party transformation has occurred, as I do in this book, therefore gives a more valid estimate of when parties change has occurred.
Cultural conservatism refers to government support for traditional or religious moral values. In public opinion surveys, principal components analysis shows that issues like abortion rights, pornography, school prayer, and gay rights are highly correlated with each other and not economic issue positions. On this basis, I treat them as a bundle of related "cultural issues." Cultural liberals either oppose government support for traditional or religious values or promote tolerance for less traditional values.
Civil rights and cultural conservatism show how previously marginal groups made their way into a major party coalition and thereby transformed it against the wishes of many party politicians. I offer original, empirical evidence to evaluate how politically effective coalitions form and groups choose the allies they do, and whether party agendas parties are primarily responsive to organized groups or politicians.
The main alternative to group-centered change is the idea that politicians change what their parties stand for. Politicians are responsible for party transformation if they change what parties stand for in response to electoral opportunities or initiate new group alliances that do so. Transformations are not complete in this scenario until voters accept a transformed party's new agenda. In contrast, groups are responsible for party transformation if they persuade reluctant politicians to change or displace holdouts with new candidates committed to their goals. This book shows that group efforts are the most consistent—though not the only—influence on parties in the contested transformations it studies. Groups exercise this influence mainly because they play a disproportionate role in mobilizing voters and delegates during nominations, when most voters are inattentive.
The book shows not just that groups change parties, but why groups make the changes they do. By doing so, it disaggregates parties into their true raw ingredients—the selection of group issue positions and allies. When a group sets out to change a party against other factions in the party, it needs to find allies committed to party change. Since one group is seldom powerful enough to change a party on its own, it needs to maintain good relations with other groups with their own agendas. Usually, this means adopting new issue positions or broadening its focus. Groups evaluate coalition allies and parties not just on common issue positions and ideologies, but the need to grow their membership, raise funds, and compete with other groups for prestige and influence. Once they form a coalition, they tend to support mutually acceptable candidates and platforms.
These points can be digested into three major claims
- The critical first step in party change is for groups to form a viable coalition committed to a common agenda they want the party to adopt.
- Groups, not politicians, are the most important force for transforming parties
when other party factions resist change.
- Groups force parties to change primarily by gaining influence over nominations.
Therefore, party change reflects the groups composing parties and their role in nominations.
This book does not attempt to understand all party transformation—only contested party transformations. In such cases, some party factions oppose the new entrants, if not majority factions. In some transformations, there is little or no disagreement between existing and new party factions. Since both approve the change, there is no telling whether groups or politicians are more responsible. For example, federal attempts to address pollution constituted a new issue in the 1950s, but federal spending and business regulation did not alienate any major Democratic constituency at the time. When transformations are contested, the relative influence of groups and parties come to light.
Claim 1: Party Change Starts with Coalitions of Groups
To help readers identify the relevance of my claims to contemporary debates, let me start by fleshing out the claims. A description and justification of the evidence I use will follow. This book is based on the idea that groups are, in the common phrase, the life of the parties—not only the building blocks of party coalitions, but their most important moving parts. By mobilizing followers and influencing nominating contests, groups pull politicians closer to them than to median voters.
Claim 1 delves deeper than other group-centered theories of parties. In their treatment of parties, Cohen, Karol, Noel, and Zaller admit they examine the activist origins of party change but not where "new activist agendas came from." I argue that party change reflects the agendas groups construct with allies. Using historical evidence, we can trace the positions of parties back to the positions that groups introduce, emphasize, or drop to work in a coalition. Groups stitch together a new agenda with needed allies to defeat their opponents in a party, mostly independent of politicians.
The first step in contested party transformations is for groups to form a viable coalition committed to a common agenda. Nominations are competitive affairs in a large country with only two parties. Even large national organizations like the Congress of Industrial Organizations (CIO) and National Education Association (NEA) can succeed only if they can gain the help of allies. When deciding whom to ally with, groups weigh such factors as issue overlap, the approval of their own members and staff, and the ability to reach new audiences. In viable coalitions, groups broaden their original agenda to accommodate their allies. Whether they succeed in transforming parties depends largely on choosing carefully and making the proper accommodations.
Agreeing on a single nominee for office can be difficult in the best of times. When two or more groups in a party are fighting for the direction of the party, the need for allies is dire. Stalwart incumbents are generally well networked with important groups and party activists that avoid party disruptions. Transformative groups need to defeat not only stalwarts, but politicians hoping for compromise under a big party tent. Even if party change seems inevitable, a politician may hope that a straddle will bring in votes from two inconsistent blocs during the current election cycle. In addition to other considerations, changing the party in the name of a coalition suggests broad demand for change and not simply one "special interest group's" agenda.
Working with allies requires groups to compromise, adapt, or add to their agendas. To earn enough good will, groups must not simply take positions but offer meaningful support such as campaigning, lobbying, and voter mobilization. Naturally, a group will disrupt its own organizations less if it allies with a group that already agrees on one or more issues, even if they agree for different reasons. Schickler calls such issues "common carriers," "policies that a multiplicity of interests support for sometimes dissimilar reasons." Groups in a coalition may bundle common carriers together with their more distinct priorities. If a coalition can present a united front, it shows politicians that they risk losing a broad array of groups unless they adopt the whole bundle.
Groups often choose allies who best complement their strengths and weaknesses. Allies are more desirable if they can message supporters they would not otherwise be able to reach. Political outgroups usually seek to mobilize people who are not motivated by policy concerns alone. One way to expand their base of support is to recruit voters from captive audiences motivated by nonpolitical concerns, like union members and churchgoers. A more socially acceptable group can also lend legitimacy to a less popular group. Some allies make up for a group's deficiencies better than others.
Societal changes beyond a group's control also influence the desirability of allies. Laws protecting the right to unionize made unions a far more formidable partner for civil rights groups than they were for Populists decades earlier. For conservatives in the 1970s, demographic and cultural changes caused the broadcasters and churches they targeted to flourish. Religious broadcasting expanded as television networks changed their rules.
Groups need to worry about whether adopting a new ally or taking a new issue position will alienate members of their own organization. A prospective partner may cause a group to lose donations or members. If a group changes its position or focus to obtain a partner, rival organizations can grow at a group's expense. Rivals for group leadership may also criticize an alliance as a way of competing for power with the current leader. Strolovitch argues that groups on the left betray their ideological commitment to universalism to help their most advantaged subgroups. Less-advantaged groups need the resources of their more-advantaged subgroups, so that, for example, civil rights groups will work hardest on behalf of well-off African Americans, women's groups will do so for heterosexual women, and so on. I found that less-advantaged members could sometimes overcome their disadvantages by compensating with volunteers. In other cases, ideological purists may be crowded out by those willing to compromise to get ahead. In the United Auto Workers (UAW), Walter Reuther changed from a socialist to an anticommunist liberal because socialists and communists blocked his advancement in the union.
Groups must be sensitive to the focus and language of a potential ally. Past conflict or clashing positions may create severe obstacles to working together. In contested transformations, groups need brokers to forge ties with other groups with their own histories and approaches to politics. Brokers are typically familiar with the sensitivities of both camps. Conservative strategists who formed the "New Right" found a broker with conservative Christians in Ed McAteer, a Colgate-Palmolive salesman with an extensive network of evangelical Christians who trusted him. Schlozman (2015) points out that groups need brokers who understand them to forge ties with party insiders. But it might not be in a politician's interest to unite conflicting groups if they have less ability to pressure the politician separately. Groups often need to find brokers other than officeholders to unite them.
Heaney and Rojas point out that when the Democratic Party wanted to shift attention away from the Iraq War to other priorities in 2006, it withdrew brokers and other resources from the antiwar movement, reducing it to a relatively extreme splinter faction. The more a transformative coalition relies on party loyalists as brokers, the less it can muster challenges to its party. As once-disruptive groups become more fully integrated into the party and dependent on its resources, their members may lose leverage. For most participants in the antiwar movement's "party in the street," party identification trumped movement identification. For civil rights activists and cultural conservatives, this was not the case.
Common ideology is seemingly an obvious reason for groups to trust each other. Noel's research on pundit ideology argues that a new consensus among pundits anticipates and constrains new party positions. If we define ideology as a set of issue positions fit together by pundits, it might restrict groups, parties, or both. Alternatively, ideology could be a product of the agenda groups construct, which then constrains both parties and pundits. Three possible formulations are:
- Ideologies directly constrain parties regardless of groups in the party.
- Ideologies constrain group positions, and groups constrain party positions.
- Ideology is the result of groups building coalitions, and ideology constrains party positions.
This book provides evidence against the first position, showing that politicians respond directly to group pressure. Groups could still be influenced in turn by ideology, however.
According to the second position, ideology constrains both candidates and interest groups. By looking at whether groups prioritize ideology over other conflicting considerations, we can weigh their relative importance. I generally find that ideology plays a limited role. Groups can tailor their interpretation of a fluid ideology to their material or social needs. For example, evangelical Christians could downplay biblical exhortations to be stewards of the Earth in order to work effectively in a coalition with opponents of environmentalism. Even if they offer an ideological rationale for opposing environmentalism, it could be an afterthought rather than good-faith deliberation.
The third position holds that ideology is a product of group agreement, assigning a small role for the pundits usually thought to define ideology. In this formulation, pundits rationalize positions that groups had already conjoined instead of being the agents who define ideology. Liberalism might entail civil rights and economic liberalism because the NAACP and CIO worked together, largely for nonideological reasons. As Kersch has written, it is only in hindsight that parties appear to be following a linear ideological path toward a natural or logical endpoint. While the third position is consistent with my findings, I do not explore how pundits adopt ideologies. This book merely argues that ideology is not a strict constraint on groups, group leaders, and coalition-building efforts.
Instead, preexisting ideology is one of many influences on a group's choice of allies. Ideological groups, who usually lack captive audiences, can still serve an important role in preserving alliances. The number of voters motivated by ideology alone is small, but groups like Americans for Democratic Action (ADA) and the American Conservative Union (ACU) can offer volunteers and intellectual ammunition to groups with more concrete interests. In doing so, they create an incentive for other groups in a coalition to adhere to an ideology even when group interests change. Over time, they can help convince people that some issues naturally fit together. More self-interested allies need to weigh the loss of ideological group support if they change positions for other reasons. Liberal groups, for example, helped ensure that civil rights remained part of the Democratic Party agenda even as rifts emerged between labor unions and civil rights groups in the 1960s.
Looking over past coalitional transformations, I find several reasons to be skeptical of ideology as a primary reason groups work together. First, the interests of two or more groups in working together are often compelling enough that they might have allied without any ideological motivations. We cannot test counterfactual motivations, but group leaders set aside documented ideological commitments when they conflict with their own interests, member interests, or the interest of allies. Second, groups ignore good ideological fits when institutions prevent them from capitalizing on them politically or organizationally. Third, purveyors of ideology sometimes attest to the limited impact of their ideology on their readers and practical political groups. This does not mean that it is in a group or party's interest to present their issues as a smorgasbord of disparate issues; ideology can better inspire followers than narrow interests.
Claim 2: Groups, Not Politicians
During contested party transformations, groups enjoy several advantages over vote-maximizing politicians. First, politicians who want to counteract group influence usually need the support other groups or unorganized voters. If they cannot find support among rival groups, voters generally suffer debilitating collective action problems. Second, groups provide a resource for politicians by offering to mobilize their members on their behalf. If politicians accept group issue positions, they can use their own resources elsewhere while groups do considerable work for them. Third, politicians have more to lose than groups by changing the direction of a party.
As Olson realized long ago, intense minorities have incentives to put much more time and effort into working for their public policy agendas than dispersed and less motivated majorities. It takes more work to organize a larger number of people, and activating weakly motivated people is obviously more difficult. Financially, unions have much more at stake than the average voter in labor laws, and Christian school parents have more at stake in tuition tax credits. Olson also pointed out that groups who can mobilize political followers in the process of providing nonpolitical benefits are especially powerful, since they bring votes and volunteers that are otherwise uninterested in political appeals. For example, unions can disseminate political messages while providing job benefits and church groups can do so in the process of providing community services.
Candidates who rely on their own campaigns or formal party organizations need more time and money than competitors who can outsource campaign work to preexisting groups. Such groups coordinate with other groups with shared interests to provide campaign mobilization to the right candidate and withhold it from others. It is therefore in a politician's interest, ceteris paribus, to commit to groups who can provide it. One study finds that the Republican Party developed its own electoral apparatus, the Republican National Committee (RNC), because it lacked the kind of assistance the CIO and its successors provided to the Democrats. The Democratic Party did not develop a worthy counterpart to the RNC until decades later, after unions suffered great attrition. Parties and candidates still develop their own mobilization efforts to target particular groups, but building such an organization is harder than letting existing groups do it.
Politicians may also be less familiar with the norms of groups whose leaders have intimate ties to the communities that they represent. To obtain support, candidates have to demonstrate a reasonable level of congruence with not only the immediate agenda of the group, but that of other groups in their coalition. Opponents to the war in Iraq were able to mobilize rapidly by recruiting from existing networks of groups who respected each others' work. If groups withhold support from candidates who do not support their allies, candidates have an incentive to adopt the agenda of a group's coalition wholesale.
Finally, groups have less to lose and more to gain by antagonizing incumbent party politicians. In some cases, groups sit out or threaten to sit out an election in order to obtain concessions. It can be a rational strategy for future-minded groups: heads, the politician caves in; tails, the politician ignores them and loses the election. Groups will then have an easier time nominating loyalists in the next election when the party has no incumbent running. They might even raise more money, railing against the opposition party in power or telling donors how they helped defeat half-hearted partisan candidates.
Politicians attempting a transformation need to be more willing to sacrifice near-term electoral prospects than they usually are, since party transformation can divide a coalition during elections. Politician and formal party organization employees hurt their reputations by losing the current election, and threatening an incumbent during a primary season may cost the party during the general election. The media may interpret election losses as a repudiation and ignore the longer game. Any particular party official or officeholder might lose his or her party's esteem before reaping the benefits of party transformation. Politicians tend to refuse a group's advances if it conflicts with other important groups, which is why new groups need to counteract the incentives for inertia.
Hypothetically, politicians might seek new groups to support them to augment their votes, thus transforming their party. But office-seekers will avoid augmenting their base when a new group's goals conflicts with that of existing groups. Stable alliances become more difficult to rearrange as groups become more accustomed to working together. Politicians risk a tense confrontation and the opposing party may pick up the pieces from an alienated partner. Schattschneider realized that a party can have "a vested interest in the old lineup in which it confronts familiar antagonists already well identified in old contests." Even minority party politicians often benefit from maintaining existing alliances when they can obtain a prominent position within the minority. An old saying around Washington, D.C., holds that "the minority leader still rides in a limousine." Vote-maximizing politicians are especially unlikely to risk changing a lineup for a less popular group than the one they wish to displace.
The idea of groups as the fundamental components of parties has a long lineage in political science, but has been neglected for some time. Schattschneider called organizations the manifestations of interests and interests and the "raw materials" of politics, but even he did not document the steps on the assembly line, and concludes that the refinery owners (party politicians) rather than the raw materials (groups) dominate the process. For more than a generation, scholars deemphasized groups even farther. Realignment scholars, for example, viewed party change primarily as the "reordering of public preferences" by politicians. All the while, sociologists paid close attention to the internal dynamics of social movements and organizations, but seldom studied political parties; only recently has the sociology of parties enjoyed a resurgence.
The main alternative to group-centered transformation has been candidate-centered transformation. In Downs's pioneering theory of politics, politicians maximize votes and aim for nothing more than gaining office. Downs specifies that political parties are the vehicles of politicians and not "agents of specific social groups," implying that politicians transform parties when they need to upgrade their vehicles. A leading contemporary treatment of parties, John Aldrich's Why Parties? continues to view parties as the electoral tools of ambitious politicians. Although he is aware of the role of "benefit seekers," most of his casework and quantitative evidence stress the role of candidates and politicians. Carmines and Stimson espouse a candidate-centered model of issue "evolution," emphasizing the role of politicians in changing partisan voter opinion. Prominent party leaders attempt to increase their vote margins by changing their stance on issues to win new groups. When they win, other politicians follow suit, and voters change their party identification with the politicians. Candidate-centered theories of party transformation are generally voter-centered theories that postulate that politicians take the positions they do in order to win votes.
A new generation of political scientists has conceptualized parties as coalitions of groups. Bawn, Cohen, Karol, Masket, Noel, and Zaller emphasize that voter ignorance creates "blind-spots" in which parties can satisfy intense party activists because moderate voters usually ignore politics, especially primaries. Karol describes party change on civil rights and abortion as party "coalition maintenance" and "coalition incorporation," respectively. As with the others, he argues that politicians are responding to external events or changes in interest group positions, but ascribes a larger role to politicians in coalition maintenance. DiSalvo and Layman, Carsey, Green, Herrera, and Cooperman also view groups as sources of change who extend party conflict to new issues and nominate loyal candidates, but disagree with the group/party distinction. DiSalvo, for example, writes that Bawn et al. set "up a misleading choice about whether politicians or groups are at the center of a party." However, the distinction is important for deciding where the incentives and strategies for party transformation originate. Without this distinction, it is harder to understand the rotation of party outgroups and ingroups into the proverbial smoke-filled rooms and which groups remain in the room when the decisions are made. Sometimes politicians disagree with the interest groups in a party. While pursuing the Republican nomination in 2016, Texas Senator Ted Cruz gained many endorsements from evangelical leaders, but not national politicians. For the purposes of studying party transformation, we need to know how outgroups can gain influence until they become ingroups, and treating groups separately from politicians adds clarity to the discussion.
Claim 3: Groups Change Parties by Influencing Nominations
Nominations are the primary arena in which groups transform parties. By making sure that parties nominate candidates committed to their agendas, parties are more likely to work for their goals. Whether the nominating system consists of conventions or primaries, voters pay little attention. This provides motivated groups with the opportunity to play an outsized role in who is nominated, since they can provide money, volunteers, and publicity that dispersed and inattentive voters will not provide. Before the McGovern-Fraser reforms required parties to nominate the winners of state primaries and caucuses, starting with the 1972 presidential elections, this mainly consisted of influencing convention delegates; after these reforms, it meant influencing primary elections.
Dispersed majorities are unlikely to research candidates during general elections and even less likely to research them during nominating contests in either system. In presidential elections, primary voters often jump aboard the bandwagon of candidates who won earlier contests. Lacking the party cue to distinguish one candidate from another, they distinguish primarily on the basis of prior victories. Groups who have a major stake in the direction of a party do not usually sit idly and wait for convention delegates or early state primary voters to make up their minds. They contribute volunteers, money, prestige endorsements, and expert advice to candidates who commit to their agenda. Politicians know that pressure groups are more likely to scrutinize and publicize their activities than unorganized voters. They might still refuse a group's advances if it will lose the interest of mainstream voters, but only to the extent that voters would be aware of it. Groups nudge politicians to be as extreme in their commitment as they can be without being noticed by more moderate (but typically inattentive) voters. Politicians backtrack on unpopular commitments in general elections, but on balance, I argue that they are more responsive to groups than median voters. In the general election, median voters can only vote against their party's optimally extreme nominee by supporting an optimally extreme nominee on the other side. This arrangement enables groups to transform a party's agenda in ways against the wishes of more numerous but unorganized voters.
Groups prefer candidates with real commitments in the first place, though recruiting good candidates is usually difficult. Instead of constantly looking over the politicians' shoulder, groups hope to nominate someone who would work for them without supervision. As the CIO PAC founder Sidney Hillman said in 1943,
The Democratic Party is very open to the proposition of giving our groups a great deal of say right in the party—not merely on policies, but a discussion of the kind of people they are going to nominate before they nominate them. I think if we have real leadership we can work out, especially with the Democratic Party, some satisfactory arrangements so that we do not really have a choice between two evils. After my trip I have seen some of the top leadership of the Democratic Party, and I think there, too, there is a desire to discuss with us before instead of after. You know what happened in New Jersey where, because of the AFL and the CIO and the Brotherhoods working together, we have forced our nominee as the gubernatorial candidate.
Groups scrutinize politicians in office as well as candidates during campaigns, but some important deals, compromises, and lost opportunities may occur in "off the record moments." This explains why Tea Party groups preferred Christine O'Donnell over Mike Castle in the 2010 Republican Primary for an open Senate seat in Delaware. Castle was more likely to win the general election but O'Donnell was more likely to follow the Tea Party's agenda. This does not necessarily mean that ideal nominees will be able to pass a group's policies in office, which depends on circumstances greater than the preferences of an officeholder. But they are more likely to fight for it under a favorable context.
Of course, electability is one factor groups consider. Different groups and politicians value electability more than others. But on a continuum representing the tradeoff between commitment and electability, groups are more likely to choose a place closer to the commitment end; for politicians, it is the electability end. Groups can still have influence in politics even if their preferred party loses a particular election, while a single lost election can damage a politician's career trajectory.
Politicians would have to challenge fellow partisans during nominations to avail themselves of the most opportune moments for party change. Generally speaking, politicians in the same party refuse to do this because they value collegial relationships in small capitol communities. Even when politicians want their party to change, they seldom endorse challengers to their fellow incumbents. Politicians thereby deprive themselves of a powerful tool for party transformation available to groups. In an important exception, President Franklin Roosevelt attempted to purge conservative Democrats in the 1938 midterms, but his effort was an unqualified failure that thwarted his future success in domestic policy. In another rare case in which a sitting politician supported primary challenges to fellow partisans, South Carolina senator Jim DeMint was sharply criticized for raising $7.5 million for the midterm elections of 2010 and using it to support conservative challengers to incumbent Republicans. One congressional aide anonymously reported DeMint was cowed by the party. "If on Nov. 3 there are two or three seats in Democratic control that otherwise would have been Republican victories," he said, "then that anger will come back up to the surface and there will be consequences," without specifying the consequences. Apparently stirred by these criticisms, DeMint only targeted Democrats in 2012.
Nominating the right candidates is the primary mechanism for party transformation, but we need not make the strong claim that it is the exclusive mechanism. Groups may also threaten to withdraw support during a general election, holding their nose for an even worse officeholder in hopes of teaching their party a lesson. In nominating conventions, groups also signal a national direction for a party by passing party platforms reflecting their group agendas. Truman and many other political scientists have dismissed platforms as nonbinding words that appease interest groups without offering them real policies. In fact, platforms offer interest groups a way to vocalize goals that may be too ambitious for the current political climate. Research also indicates that platforms are not mere parchment trophies. Most critically, Pomper and Lederman find that between 1944 and 1978, parties fulfilled platform pledges at a rate between 50 and 100 percent. Another study finds that platforms alter the way politicians frame issues and create expectations among issue publics.
My focus on nominations, rather than on general elections, fits squarely with the claims made in Cohen, Karol, Noel, and Zaller's The Party Decides. In their account, motivated "intense policy demanders" in one party decide on a nominee that satisfies major party factions before primaries take place, signaling other party members with endorsements. Policy demanders need to find a nominee that can win the general election, but "they cede as little policy to voters as possible." One difference in my account is greater attention to the differences between politicians and other intense policy demanders. Recently, the Tea Party has supported candidates like Ted Cruz, Marco Rubio, and Christine O'Donnell over candidates with more endorsements from politicians. Groups and politicians usually accommodate each other in periods of stability, but sometimes the party's direction is contested. The Tea Party challenges are one example of the need to maintain the distinction even after a party transformation has already taken place.
Critics of the theory like Wayne Steger argue that in many nominating contests, party activists fail to generate a consensus behind one candidate ahead of time. In the transformations covered in this book, groups are intensely active when they are trying to transform a party, but mainly consolidate their gains after the party is on the right trajectory. They still demonstrate their power by vetoing unacceptable candidates even if they do not actively agree on their first choice. President Lyndon Johnson's change on civil rights and President Donald Trump's on abortion rights are just two examples of candidates who flip-flopped to avoid group approbation. One of the coauthors of The Party Decides, Karol, emphasizes the role of adaptation rather than displacement. That is, parties mainly change when party politicians change their minds and adapt to new positions in the party, with little need to nominate new candidates. However, one of the reasons they adapt is the risk of primary challenges by groups.
Most recently, Schlozman argues that "anchoring groups" trade independence for "ideological patronage" by providing parties with a reliable stream of money and manpower. Such groups, he says, will not get access to parties if party elites doubt that they provide votes on net. But nominations can force candidates to take losing general election positions, since politicians need to win nominations before they run in general elections. Contemporary Republicans have provoked government shutdowns despite their enormous unpopularity to avoid primary challenges from aggressive conservative groups.
Even if minority factions did not prevent parties from being more responsive to majorities, many majority coalitions are possible, depending on which issues are grouped with which other issues. The United States has scores of independent issues of varying popularity, and no dominant or natural majority exists. The Democratic Party could have been competitive in the 1940s and 1950s either by continuing to rely on the overwhelming support of southern whites and picking up northern votes when conditions were favorable, or by turning to blacks and labor as its core groups. Or it might have bundled support for the New Deal, civil rights, and conciliation with the USSR, as Progressive Party nominee Henry Wallace wanted. Which bundle parties adopt depends on which coalitions gain the most influence over nominations.
I test the claims with a deep historical investigation of party transformation on two different issues—civil rights and cultural conservatism. History offers the chance to reveal complex processes in which multiple variables—groups, politicians, interests, laws, demographics, and ideologies—are changing at the same time. By conducting a thorough investigation into the interactions between different actors, we can discern who is responding to whom and why. As Truman writes, both parties and groups value unity and the appearance of unity, so "most groups are careful to reveal as little as possible to the outsider concerning such internecine struggle." They are less careful about discussing such struggles in private letters, memos, conference notes, and meeting minutes, which are archived and made available to the public decades later. Removed from the heat of the moment, this evidence reveals different motives than politicians or group leaders admit publicly at the time parties are changing.
Political behavior scholars usually evaluate claims about parties using statistical tests. In testing Claim 2, for example, one might treat party change as a dependent variable and particular cases of candidate-centered change and group-centered change as independent variables. However, each data point of party transformation requires an in-depth investigation to determine who or what is determining the outcome. Party transformation is typically a long process in which politicians, groups, and others are attempting to influence each other. There are simply not enough clear-cut data points—where the outcome was for sure decided by one force or another—for a statistical test. As much has been written about civil rights and cultural conservatism, this book is the first to publish some of the data needed to evaluate the role of groups and parties in party transformations. Given the importance of understanding why parties change, the inability to test the impact of variables through statistical relationships should not prevent us from making inferences from other forms of evidence.
The transformations examined are not simply two case studies but two long processes, each with many data points speaking to the book's claims. For example, Claim 2 implies that the Democratic Party would support a civil rights plank in 1948, because important component groups of the Democratic Party had placed such importance on it. But they also imply that a) Adlai Stevenson would reassure racial liberals that he was a civil rights supporter in 1952, b) the 1956 convention would select Estes Kefauver as the vice presidential nominee, and c) that groups would force John F. Kennedy to distance himself from the South. The nominating contests and separate conflicts within each contest offer plentiful evidence from which to draw inferences.
I chose to focus on civil rights and cultural conservatism because they are the most significant party transformations since the New Deal; parties have maintained the same relative positions on economic and social welfare policies since then. Civil rights and cultural issues, on the other hand, comprise distinct dimensions of issues that changed after the New Deal economic alignment. They are not single issues, but an entire cluster of related issues. While a test of party transformation might conceivably overlook the transformation of the two parties on single issues (tariffs, for example), no theory of party transformation can ignore civil rights and cultural issues. They constitute the major axis upon which parties have turned. For entire decades, one could not predict the position of members of Congress (MCs) on civil rights or cultural issues from their positions on economic issues. Over time, these sets of issues became predictably related; in Poole and Rosenthal's widely used metric of party ideology, second dimension NOMINATE scores were eventually predicted by first dimension NOMINATE scores. Rarely have such orthogonal clusters of issues become straightforward partisan issues.
The civil rights transformation not only bears on the question of how parties transform, but also the current makeup and electoral fortune of America's two major parties. Some historians and political scientists argue that the Democratic Party's New Deal coalition unraveled because of racial issues. Others claim that the civil rights realignment presaged the Republican Party realignment on cultural issues by driving culturally conservative white Southerners away from the Democratic Party. Either way, it was a pivotal change in party history.
However, this transformation took place when party conventions, instead of primary voters, formally selected nominees. Numerous other changes, including the growth of the mass media and direct mailing, also confound attempts to generalize from the civil rights transformation to the modern era. As such, the group-centered transformation might be ascribed to a bygone era of smoke-filled rooms and backroom promises. In order to show that the findings are still relevant in an age of primaries, a more contemporary party transformation is needed. Additionally, Hopkins and Grossman argue that the parties are not symmetrical. While the Democrats are a coalition of groups, the Republican Party is a party of ideology. If true, the group-centered civil rights change of the Democratic Party does not apply to Republicans. An account of change in both parties is needed to examine whether the claims apply to both parties.
The cultural conservative transformation of the Republican Party shows that groups are still the most consistent and powerful agents of change, although nominations are legally in the hands of voters. Various religious sects overcame their prejudices toward other sects and also formed partnerships with political conservatives. Resourceful conservatives confronted Republican politicians with issues they preferred to avoid or market selectively. Two separate waves of cultural conservative groups were able to complete party transformation with comparatively little help from candidates or the general public by being active in primaries, particularly in important states holding contests early in the season. The cultural conservative transformation differed from the civil rights transformation in some ways that will be highlighted in the relevant chapters and the conclusion. However, these differences do not undermine the fundamental support they provide for the book's theoretical claims.
To demonstrate how my analysis might be applied beyond the focal pair of issues, I investigate nineteenth-century populism and gay rights. Space does not permit me to provide the same rigorous examination that I use for civil rights and cultural conservatism, but the available data suggests considerable symmetry. With the Populists, transformative groups failed to transform a party because they did not meet the conditions I specify for successful coalition building. Instead, they allowed rifts between farmers and laborers to fester and did little to repair racial divisions. The gay rights transformation, representing one of the most important issues to contemporary liberals, reaffirms that groups are still "first to the party" for Democrats.
I have examined dozens of archives to uncover the actions, attitudes, and motives of the key players in the realignment on civil rights in the 1940s and 1950s. Altogether, I sifted through tens of thousands of documents from ten different states, including records from unions, interest groups, MCs, presidents, and presidential aides. For the realignment on cultural issues, I have looked through the Ronald Reagan and George H. W. Bush presidential libraries and conducted interviews with more than fifty individuals from thirty different organizations. The result is a deeper and more credible account of the dynamics of party change than could be obtained by other means. Some of the findings corroborate existing sources or suspicions, but others offer new insight into the changes taking place. Both major transformations are worthy stories in their own right, only partly told in other accounts.
In searching through the archives, I focused most of my efforts on presidential nominations, relying on secondary sources for congressional and state nominations where possible. Influence over presidential candidates signals a group's influence in a party better than most other measures. Groups need influence over presidential nominees who can fight for lasting policy changes when unified government and public opinion created a favorable context. As we will see, coalitions with transformative ambitions treated national tickets as the cornerstone of their strategy. They thought that a party's national ticket and platforms were more indicative of where a party stands than more localized state and congressional contests. Civil rights groups and cultural conservatives in particular desired policy changes through the judicial system, where presidential appointments played an inordinate role. With congress deadlocked on controversial proposals like discrimination in the military and abortion funding, presidents also signed important executive orders. Frances Lee shows that congressional parties often change their issue positions to match those of a president from their own party and vote more uniformly along party lines..
Party transformation is not so much an explicit coalitional bargain accepted by all groups in a party as much as a shift in which coalition gains the most influence in nominations, often over the continued objections of other groups. If a coalition with a new set of views continuously nominates loyal candidates over the objections of older groups, the party has changed. Parties are large, multifaceted organizations with affiliated politicians in the White House and Congress, and at the state level. Groups may have more influence in one arena than in others, often influencing state parties in sympathetic parts of the country before they influence national leaders. All of these divisions are important to parties in the process of transformation, but presidential nominations are the best single vantage point from which to observe these divisions and who comes out on top.
Historical evidence has its limits, of course. People only know what they witnessed or what they learned secondhand, and seldom know all of the relevant details. They may innocently misremember events. Some political actors prefer not to keep written records (or to dispose of them). Both in written accounts and oral histories, people are apt to exaggerate their own role or distort a course of events to create a narrative that serves their own purposes. When multiple accounts diverge, one can only draw tentative conclusions. For most of the key points in this book, though, multiple pieces of historical evidence confirm the same sequence of events.
The Plan of the Book
The next four chapters delve into the civil rights transformation. African Americans, long neglected by both parties, became part of a coalition that forced the Democratic Party to favor civil rights. Chapter 2 documents the transformation of the prestigious NAACP into a supporter of labor unions and liberal causes more broadly. Internal changes caused the NAACP to accommodate allies who could bring about the party change it wanted. It rejected an alliance with the CIO in the 1930s and changed with its organizational needs in the 1940s. Chapter 3 explains the CIO's interest in working with the NAACP to promote civil rights. The politically ambitious CIO thought black voters could help defeat an opposing party faction—conservative southern Democrats. Both organizations developed constructive ways of improving relations with each other to effect a formidable alliance. Chapter 4 shows that this alliance bore fruit at the 1948 Democratic Convention, which marked the transformation of racial equality from party taboo to a litmus test. The CIO and its allies passed a civil rights plank at the 1948 Democratic Convention against the wishes of most party leaders. As I demonstrate in Chapter 5, Democratic Party nominees did not retreat from the positions taken in 1948. Reflecting the new party equilibrium, serious presidential contenders distanced themselves from whatever ties they had to the southern wing of the party. Future candidates and platforms improved on the civil rights positions of 1948.
A new set of cultural issues confronted parties by the 1970s, including abortion, school prayer, and gay rights. They were slow to grab the attention of the religious conservatives now thought to be a natural constituency for these issues. As Chapter 6 describes, conservative religious sects were neither united nor uniformly culturally conservative. Religious leaders and denominations, like civil rights and labor, were groups that operated according to their own institutional incentives that militated against cooperation. Chapter 7 explains how these sects underwent a transformation themselves before they could bring a cultural conservative transformation to the Republican Party. Conservative critics of the Republican Party formed institutional outlets for their new grievances and helped socialize them into a set of coalition norms. Chapter 8 describes the Christian Right's first attempts to infiltrate the new system of binding primary elections. Republican politicians left to their own devices were about as willing to prioritize cultural conservatism in the 1970s as Democrats were willing to prioritize civil rights in the 1940s. Chapter 9 explains the origins and tactics of a second wave of conservative Christians, which focused on capturing state parties. Since momentum in early states became important for party nominations, state parties in Iowa and South Carolina provided a powerful asset to culturally conservative candidates. Chapter 10 shows that following Reagan's presidency, all viable candidates for the Republican presidential nomination presented themselves as allies of the Christian Right and needed to take them seriously in office.
Chapter 11 looks beyond the two major transformations of the book to investigate party change in other situations. First, it explores the unsuccessful attempts of nineteenth-century Populists to change both the two-party system and then the Democratic Party. The Populist revolt was perhaps the most notable failed transformation in the country's history. Moreover, it casts doubt on the inevitability of a CIO-NAACP alliance by showing that racial solidarity can trump class solidarity in politics. Since the absence of effective coalition formation led to its failure, it strengthens the case for Claim 1. The penultimate chapter also examines the addition of gay rights to the Democratic agenda as an example of contemporary party change in the Democratic Party.
The concluding chapter summarizes the aforementioned transformations and the case for treating them as parallel. As such, we can draw the same conclusion in spite of some interesting differences. I discuss the current state of the parties, including Trump's improbable success in the 2016 Republican primaries and election, in light of the book's evidence. While most of the book considers how parties work, the conclusion evaluates what parties mean for democracy. Parties are often treated as bulwarks of democracy, but the process of party transformation should temper our judgments. Well-organized groups can change parties independently of ideology or politicians, who presumably would cater to voters if unimpeded by groups. Groups continued to do so even after the passage of reforms designed to broaden voter participation in party nominations.
Politicians go hunting where the ducks are, as Barry Goldwater said, but a lot happens before politicians go hunting. Caretakers make some hunting grounds much more appealing to hunters just like some groups can offer more to politicians than other groups—volunteers, money, networks, and expertise. While hunters might focus on the current season just like politicians focus on the current election, caretakers attract prey by minding future hunting seasons. Both natural and manmade climate change alter the desirability of some hunting grounds over time just as new laws and demographic changes affect the appeal of some groups as allies or constituents. And hunting grounds can exclude hunters who drag dirt from other hunting grounds, just like groups in a party can snub politicians who associate too strongly with opposing groups in a party. Deciding where to hunt is only the end of a long process in which some places are, by design and historical accident, more desirable than others.