All three books reviewed here focus on partisan divisions and political beliefs in the United States. Yet, beyond this important basic commonality, they differ greatly in methods and substance. In The Myth of Left and Right: How the Political Spectrum Misleads and Harms America, Hyrum Lewis and Verlan Lewis (henceforth “the Lewises”) have written a spirited polemic about ideology in American politics. These authors claim that ideology is a meaningless post hoc label for evolving party policies, rendering attachment to “liberalism” or ‘conservatism” simply destructive, mindless “tribalism.” They cite many policy reversals by parties and note that the terms “liberal,” “conservative,” “left,” “right,” and “ideology” were not always part of the American political lexicon. The Lewises are right that short-term partisan tactics can be reified as ideology and that overarching claims about parties’ directional shifts are often problematic. Debate may shift “leftward” on one policy, such as health care, while moving “rightward” on another one, like tax rates. New issues, like abortion, may arise, whereas others, such as Prohibition, disappear. Thus, global assessments of parties’ spatial movement are characteristically difficult.
The Lewises assert that the “myth” of ideology legitimates rigid partisanship and cue-taking on issues in a way that mere party loyalty does not; they contend that if readers would accept their debunking of ideology, the dysfunctions of American politics would diminish. They therefore give great weight to the shift from the older term “party principles” to ideology. The authors consider the decline of third-party movements as evidence that voters were more principled in the (very partisan!) nineteenth century, before the term “ideology” was widely used in the United States. Yet there has been much realignment in recent decades linked to parties’ changing issue stands.
The Lewises deny that they are attacking a straw man, yet they do not dispel concerns by making their foil an “essentialist” view of ideology as unchanging or inspired by just one underlying value or psychological trait. The insights that parties are opportunistic or that ideology is socially constructed and evolves are not new. Nor does acknowledging them render meaningless a discussion of polarization, as these authors seem to contend. Increased party cohesion—one aspect of polarization—is undeniable. Parties may also move in the ideological space without changing it: for example, many Democrats moved rightward on welfare in the 1990s while remaining the relatively pro-welfare state party. Similarly, by the 1940s most Republican politicians accepted much of the New Deal, seeking only to slow its expansion.
Parties evolve, but most shifts are gradual and continuities are notable. Abortion, gun control, and environmentalism were not partisan issues or even much discussed in the 1950s. Yet most of William F. Buckley’s “fusionist” 1950s conservatism—hostility toward labor unions and the welfare state, paired with endorsement of the traditional racial hierarchy and religion—is still with us, recent discussions of “Trumpism” notwithstanding.
Building blocks of ideology are combined differently over time and cross-nationally. Yet patterns suggest that some underlying logic or logics are evident. For example, throughout Western countries most parties that are more supportive of the redistribution of wealth also have been more favorable toward LGBTQ rights, even though the Lewises see no inherent connection between these two issues. The book would have been strengthened by an exploration of such seeming elective affinities. The Lewises downplay the possibility that policy shifts may reflect stable values promoted in new conditions. Yet, this view informed historians’ famous discussion of progressives and New Dealers seeking “Jeffersonian ends by Hamiltonian means.” Ideology necessarily includes not only normative goals but also causal understandings that may change along with social conditions. As E. E. Schattschneider wrote, “People have ideas about interests” (Party Government, 1942, p. 37). Shifts of this sort do not make ideology meaningless.
The Lewises’s evidence consists of many brief historical examples. They usefully recall numerous party policy reversals, but some of their claims are questionable. For example, the authors assert that “Goldwater wanted to roll back the New Deal, opposed civil rights legislation, favored abortion rights and opposed tax cuts while Reagan took the opposite position on all these issues” (p. 30). Barry Goldwater was publicly pro-choice late in life but did not address abortion in 1964 and opposed it in his last senate race. As president, Ronald Reagan, a Goldwater backer, did attack New Deal programs, proposing Social Security cuts in 1981 and undermining the NLRB. He also weakened civil rights enforcement (though he was greatly constrained by Congress in this area). Exaggerating the differences between the two men understates real continuity. Advocating avoidance of polarizing ideological talk in favor of issue-specific “granularity,” the authors note approvingly that Martin Luther King did not refer to “liberals” or “conservatives” in his Letter from a Birmingham Jail (1963, p. 89). However, King memorably denounced “the white moderate” in that document.
The Lewises’s assertion that conservatives/Republicans once simply advocated “limited government” before adding hawkishness and social issues to the mix is also dubious. Before the 1930s, Republicans supported more military spending than Democrats, and most became prohibitionists. They later supported tariffs, immigration restrictions, and laws constraining labor unions. This is not simply a question of historical accuracy. Ignoring all this supports a view of politics as driven by abstract ideas about government and downplays group attachments—in this case, the GOP’s connection to business lobbies and Protestants. While decrying tribalism, the authors say little about how real interest-based conflicts between groups underlie political disagreement.
In The Social Roots of American Politics: A Widening Gyre, Byron Shafer and Regina Wagner analyze decades of survey data regarding issue attitudes in the United States. The authors’ historical perspective is refreshing and unusual, given the current centrality of experiments in the study of American political behavior. They examine six major social divisions—class, party, race, region, religion, and sex—and four areas of public policy: social welfare, civil rights, national security, and “cultural values.” The evidence presented comes entirely from the American National Election Study (ANES) from 1952 to 2010; the authors build scales based on factor analysis for respondents in the demographics they aggregated in three eras, 1950–70, 1970–90, and 1990–10. Finally, Shafer and Wagner give special attention to the distinction between respondents coded as party activists and “rank-and-file” partisans, which is perhaps not a social cleavage but is still a useful set of categories to examine.
This book is primarily an extended descriptive exercise rather than an attempt to advance a new theory of political behavior. However, because decades of ANES surveys have been mined by many scholars, the findings that Wagner and Shafer report are mostly familiar. They show that the mid-twentieth-century partisan cleavage on social welfare policy was joined by conflicts over race and later “cultural” or religious issues. Men and women diverged a bit more on policy over time, while regional divisions waned and party divisions on national security grew. The authors’ finding that new issues added to, but did not displace, social welfare as a party divide is consistent with Geoffrey Layman and Thomas Carsey’s work on “conflict extension” (“Party Polarization and ‘Conflict Extension’ in the American Electorate,” American Journal of Political Science 46 [4], 2002). Moreover, their finding that party activists were usually more polarized on issues than voters is also in line with previous studies. Nevertheless, the authors note an important exception to this dynamic: upper-class Southern Democratic activists were to the right of most Southern Democrats on social welfare in the earliest period studied.
An important distinction that Shafer and Wagner could have made more clearly is that between voting along income lines—which has declined in recent years—and the association between social welfare attitudes and the vote, which actually increased. Admittedly, some of this trend postdates their data. They aggregate respondents into three eras stretching over nearly 70 years. This approach allows them to examine very specific subgroups; for instance, middle-class Southern Republican activists. But voters are asked different questions within a policy area from year to year, and context may vary greatly; for instance, some respondents are questioned during wars and others during peacetime, yet their responses are aggregated in one measure. Given the variation in questions asked and in political context, the authors’ claims about a demographic move “rightward” or “leftward” in a policy area are hard to evaluate. This is especially true because the appendix does not include a systematic listing of which survey items were used to build the issue scales from year to year, making replication difficult.
Shafer and Wagner attempt to parse out the share of change in partisan preferences on issue dimensions that derives from “compositional” (p. 51) shifts, such as the enfranchisement of Southern Blacks and their movement into the Democratic Party. They acknowledge, however, that their data do not permit them to “distinguish individual moves” (p. 56); they cannot assess how much of the polarization they show resulted from attitudinal shifts by partisans, issue-based changes of partisanship within regions, and generational turnover in the electorate.
Matthew Levendusky’s book, Our Common Bonds: Using What Americans Share to Help Bridge the Partisan Divide, focuses on partisan divisions, but his methods and data are different from those of the other authors. Rather than focusing on ideology or issue attitudes, his concern is affective polarization, or the growing hostility and mistrust Americans feel for the other party and its supporters—a trend that he, like many observers, sees as corrosive of democratic norms. Proposed remedies for polarization abound, but many focus on electoral rules such as open or nonpartisan primaries, ranked-choice voting, nonpartisan redistricting, or other institutional reforms. Reformers often take voters’ orientations as largely fixed. In contrast, Levendusky discusses strategies for mitigating and altering the attitudes underlying polarization.
He contends that these attitudes are based in part on misinformation about the other side and can be mitigated by reminders of shared identities: not only “Americanness” but also more mundane connections like shared allegiance to sports teams, friendships, or family ties cutting across party lines. Highlighting underestimated areas of policy agreement also proves effective. This study is based on a mix of national surveys, survey experiments, and a field experiment. Even the observational studies sometimes include something like a treatment, in that negative partisanship is a bit lower for respondents contacted around events that prime national identity such as the Fourth of July and the Olympics.
The experimental studies were conducted over several years but do not afford the welcome historical perspective the other books offer. They do, however, permit more rigorous causal claims to the extent we are interested in the external validity of these findings. This is no small assumption, especially for survey experiments, but even if the real-world impact may be smaller than what is reported, results may still be interesting given the substantive importance of the phenomenon under study. Levendusky acknowledges that he has not found the solution and that the problems he addresses are deeply rooted and, as such, can only be partially mitigated in the near term.
Several scholars have explored possible asymmetric aspects of polarization. Although this is not Levendusky’s focus, he still discusses Donald Trump as an atypically divisive head of state whose behavior might call into question the continued validity of findings based on experiments conducted before his time in office. Levendusky finds that appeals to American identity still work to reduce partisan animus in the Trump era, but—unsurprisingly—less than before for Democratic and nonwhite respondents. His discussion of how priming shared sports fandom may ameliorate affective polarization would have been enriched by a discussion about political conflict surrounding protests in the NFL in recent years. Mistrust for the opposite party may be based in some cases on misperceptions of their views and demographics, but it may also be a reasonable response to real-world behavior not only by irresponsible and polarizing political elites but also by the voters who stand by and enable them.
Levendusky finds that dialogue and shared experiences can bridge partisan divides. Yet the book does not include much discussion of public policies that might mitigate mistrust through shared experiences such as mandatory national service, be it civilian or military. Like almost all political scientists, Levendusky sees political parties as inevitable and necessary in a democracy. Instead of seeking to undermine them, he hopes for more constructive parties. But given his behavioral, individual-level focus, he offers no strategy for reforming them. Instead, many of his recommendations concern individual-level behavior; for instance, people should consume more local media and join local civic groups. Levendusky’s methodological approach is that taken by leading studies on the field, but the book is still written in an accessible manner so that concerned citizens outside academia may find it informative.
These books are all directed to important and related questions. The varied approaches of their authors demonstrate researchers’ breadth of concern about partisan polarization in the United States. Because that trend shows few signs of abating, we will continue to need studies by scholars using diverse methods to illuminate new developments.
