American society is at war with itself. Arrayed on one side of the battlefield are culturally orthodox conservatives, who view the world through a morally traditionalist and religious lens. On the other side are the culturally progressive, liberal in values and secular in spirit. On the day’s most pressing cultural and social issues—abortion, contraception, stem-cell research, same-sex partnerships and marriage, the role and importance of family, drugs, immigration, even healthcare, the economy and the constitution—the two sides are engaged in a battle of ideas, unwilling, indeed unable, to compromise, each certain of the supreme rightness and morality of its own position. The fight is for the heart and soul of America. Or so the argument goes. The goal of this chapter is to strip away the hyperbolic, and frequently hysterical, language, to examine whether the United States has indeed polarized ideologically or is in a process of polarization.
The analysis is divided into two main parts. The first explores whether there could be said to be a culture war among America’s elites. Are its politicians, office seekers, opinion formers, commentators and self-appointed guardians of the public interest at war with each other? The answer is a resounding yes. On nearly every indicator, the divide between liberals and conservatives is wide and deep and getting more so. On this, nearly all academic observers are in agreement. However, there is much disagreement in the academy over the extent to which ordinary Americans could be said to be engaged in a culture war. The second part of this chapter explores the arguments of scholars who contend that the American public is polarized on cultural issues and of those who dismiss the idea that such a deep and fundamental divide exists among ordinary citizens. The chapter attempts to resolve the differences of the two sides by bringing new evidence to bear regarding the extent to which Americans feel strongly about cultural issues. In short, while there is little evidence to suggest that people’s positions on the key culture war issues are changing and becoming more polarized, those same issues are becoming more important to Americans. They are increasingly at the forefront of people’s minds, to be discussed, argued and fought over. This finding lends support to the argument that there exists a culture war at the mass level, even while other data demonstrate that issue polarization has not occurred. The chapter ends with a brief examination of the causes of polarization and what can be done about it.
There is no absolute agreement about where one should look for evidence of a culture war, nor what type or level of evidence is necessary to establish its veracity or not, but nearly all scholars agree that the facts stack heavily in favour of the proposition that America’s elites are very deeply divided and getting more so.
Polarization in Congress
The most compelling evidence of elite polarization comes from the US Congress, the highest lawmaking body in the United States. In the middle of the twentieth century, bipartisanship—where members of both parties worked together to achieve common goals and voted together on the same side of an issue—was much more common than it is today. Ideologically, there was considerable overlap between the two main parties. The Republican caucus included many members of a liberal persuasion, especially on social issues, and the Democratic caucus many conservatives. Today, however, conservative Democrats and, especially, liberal Republicans are an endangered species. The 2010 midterm elections saw Tea Party candidates oust sitting liberal Republicans in several primary contests. The few remaining liberal Republicans faced hostile criticism from within their own party and were challenged by more conservative Republicans in the run-up to the 2012 elections. For example, Indiana’s Richard Lugar, the joint-longest serving Republican in the Senate, was defeated in the GOP primary by Richard Mourdock, and Utah’s Orrin Hatch, the other joint-longest serving Senate Republican, had to fight off a tough primary challenge from former state senator Dan Liljenquist. Hatch had in 2010 seen Bob Bennett, his fellow Republican Utah senator, defeated by Tea Party favourite Mike Lee. Moreover, the upper chamber’s most liberal Republican member—three-term senator, Olympia Snowe—said in February 2012 that she would not seek reelection in November, specifically blaming the decline in bipartisanship for her decision: “I do find it frustrating that an atmosphere of polarization and ‘my way or the highway’ ideologies have become pervasive in campaigns and in our governing institutions… I see a vital need for the political center in order for our democracy to flourish and to find solutions that unite rather than divide us. It is time for change in the way we govern.”1 With Snowe’s retirement, the number of Senate Republicans supporting abortion rights can be counted on one hand. On the other side, the Democratic Party’s most conservative member, Senator Ben Nelson of Nebraska, also announced his retirement in 2012. Other prominent moderate senators—James Jeffords, Lincoln Chaffee, Arlen Specter, Blanche Lincoln and Evan Bayh—had already made their exits from the chamber, replaced by more radical members. The result is a “dysfunctional and paralysed” Congress, divided by ideology and party.2
Figures 1 and 2 place the contemporary congressional polarization in its historical perspective. Constructed by Keith Poole and based on Poole and Rosenthal’s influential DW-Nominate scores, the figures show the positions of the most moderate and conservative Republicans (at the 10th and 90th percentiles, respectively) and moderate and liberal Democrats (again at the 10th and 90th percentiles, respectively) over time.3 While polarized parties were the norm in the pre-second world war period, from the end of the war until the late 1960s there was a considerable ideological overlap between the parties in the House of Representatives (where ten percent of Democrats were actually more moderate than ten percent of Republicans). From the 1970s onwards, however, the parties began to diverge, a trend that accelerated in the 1980s and beyond. While the most liberal wing of the Democratic Party (measured at the 90th percentile) has not become especially more liberal, the most conservative wing of the Republican Party (90th percentile) has veered sharply to the right. Moreover, the moderate wing (measured at the 10th percentile) of the Democratic Party has grown noticeably more liberal while the GOP’s moderate wing (10th percentile) has become considerably more conservative. Today, there is no ideological overlap between the parties in the House. The gap between the parties’ moderate wings is much larger than during the bipartisan post-war years and as large as it was at the start of the twentieth century. The gap between their more radical wings has never been as large. The trends in the Senate are not quite so stark, but they are in the same direction, especially in the centre of the ideological spectrum.
[Figures 1 and 2 about here]
Extra-Legislative Elite Polarization
It is more difficult to uncover robust data on elite polarization outside the US Congress, but many academics argue the phenomenon is nonetheless real. Buchanan’s speech at the 1992 Republican Party convention is perhaps the best quoted and most notorious—in which he claimed that “there is a religious war going on in our country for the soul of America. It is a cultural war...[in which the Clintons are trying to impose] abortion on demand..., homosexual rights, discrimination against religious schools, women in combat.... [which ]is not the kind of change we can tolerate in a nation that we still call God's country”—but there are many other examples. Newt Gingrich’s career, from young congressman, to Speaker and his recent attempt to win the GOP presidential nomination, was built on incendiary, polarized language. His claim on 10th February 2012 that the Obama “administration is waging war on religion” is a notable example, which none of the other 2012 Republican presidential hopefuls disowned.
Polarization is also reflected in the many books of political commentators on both sides of the divide. The titles capture the contempt, anger and suspicion directed towards the other side. Liberal-leaning, pro-Democratic ones include Al Franken’s Lies and the Lying Liars Who Tell Them: A Fair and Balanced Look at the Right; James Carville’s We’re Right, They’re Wrong: A Handbook for Spirited Progressives; Daniel Kurtzman’s How to Win a Fight with a Conservative; Thomas Franks’ The Wrecking Crew: How Conservatives Rule; Charles Pierce’s Idiot America: How Stupidity Became a Virtue in the Land of the Free; Chris Hedges’ American Fascists: The Christian Right and the War on America; Peter Beinart’s The Good Fight: Why Liberals—and Only Liberals—Can Win the War on Terror and Make America Great Again, and so on ad infinitum.
On the conservative side, the list includes Glenn Beck’s Arguing with Idiots: How to Stop Small Minds and Big Government; Bill O’Reilly’s Pinheads and Patriots: Where you Stand in the Age of Obama; Ann Coulter’s Demonic: How the Liberal Mob is Endangering America and her How to Talk to a Liberal (If You Must); David Limbaugh’s Crimes Against Liberty: An Indictment of President Barack Obama; Jerome Corsi’s Where’s the Birth Certificate? The Case that Barack Obama is not Eligible to be President; Aaron Klein’s Red Army: The Radical Network That Must Be Defeated to Save America; and, perhaps most unpleasantly, David Freddoso’s Gangster Government: Barack Obama and the New Washington Thugocracy.
What is remarkable about these tomes, apart from their hyperbolic titles, is the absolute belief in the rightness of their own side and the unwavering certainty in the wrongness of the other. Indeed, there can be no room for compromise because the other side is not just wrong, but actually evil and very likely involved in a grand conspiracy against the United States of America. The “paranoid style”, first noted by Richard Hofstadter in the 1960s, is particularly prominent among, but not restricted to, conservative commentators. Obama’s presidency has generated a publishing boom on the right, with no argument as to his motives, objectives and lineage too outlandish to print.
The hysterical, compromise-free tone of the printed word is mirrored in the electronic sphere, where websites publish stories in which balance is an alien concept and conservative and liberal bloggers write in apocalyptic terms. The battle also rages on the airwaves after the Federal Communication Commission repealed its Fairness Doctrine in 1987, which had required balance in programming. Talk radio hosts such as Rush Limbaugh, Sean Hannity and Glenn Beck attract millions of listeners each week with their particular brands of anti-liberal rhetoric and baiting. Limbaugh, for example, felt obliged in March 2012 to call Sandra Fluke, a Georgetown University student and reproductive rights campaigner, a “slut” and a “prostitute” in response to her attempt to require her Jesuit university to provide contraception on its healthcare plan. Conservative viewers are also well served by cable television, particularly Fox News. While talk radio is dominated by conservative hosts and listeners, MSNBC views the world via a liberal lens and is an antidote to Fox News. Because conservatives and liberals tend to watch and listen to conservative and liberal shows, respectively, such programming has the effect of reinforcing Americans’ ideological positions.
The evidence presented above, and the view of most observers and academics, is quite clear. There is a culture war raging among America’s elites. There is, however, no similar agreement on whether the elite-level conflict is reflected at the level of ordinary Americans. In the following section, we present and critique the arguments of both sides, and bring new data to bear.
THE MASS PUBLIC
As noted in the introduction, academic opinion is divided on whether ordinary Americans are engaged in a culture war. To illuminate this debate, this section presents and critiques the work of two scholars, Morris Fiorina and Alan Abramowitz, who with their various co-authors have taken opposite positions. On the one hand, Abramowitz concludes that his “evidence indicates that since the 1970s, ideological polarization has increased dramatically among the mass public.”4 On the other, Fiorina argues “the simple truth is that no battle for the soul of America rages, at least none that most Americans are aware of.... The myth of the culture war rests on misinterpretation of election returns, lack of hard examination of polling data, systematic and self-serving misinterpretation by issue activists, and selective coverage by an uncritical media more concerned with news value than getting the story right.”5
Abramowitz’s and Fiorina’s first dispute regards the issue of geographic polarization. The basic idea is that red (Republican) states are getting redder (more Republican) as native residents become increasingly conservative and/or liberals leave and are replaced by conservative newcomers. The process in blue (Democratic) states is the said to be the same. Contrary to many media reports and popular stereotypes, Fiorina holds that America is not clearly and deeply divided into red and blue states. Inter-state differences—in party identification, ideology, religion, beliefs and attitudes—are frequently small and often statistically insignificant. For example, large and equal numbers of red state residents and blue state residents identify themselves as ideological moderates (about 30%), while small and equal numbers identify themselves as extremely liberal (3-4%) and extremely conservative (4-5%). Moreover, the two parties were closely matched in half the states in the 2000 presidential election (defined by less than 55% of the electorate voting for either one of the two parties). The evidence is clear, claims Fiorini: American political parties may occupy the ideological poles, but the voters and states do not.6
Abramowitz slices the data differently. While Fiorina’s analysis is temporally static, offering only a snapshot at one point in time, Abramowitz examines geographic polarization over time, comparing the nationally competitive presidential elections of 1960 and 1976 with those in 2000 and 2004. He shows that the average margin of victory across states has risen from around 8 to 14 percentage points and that the number of states considered to be competitive has declined by around ten, confirming that “red states have been getting redder while blue states have been getting bluer.”7 As polarization is as much a trend as a state, Abramowitz’s dynamic overtime analysis is most convincing than Fiorina’s, even though Fiorina is right to point out that a plurality of states is not solidly Republican or Democratic but marginal.
Both scholars address the issue of religious polarization. The idea is that the Republican Party is increasingly the party of God-fearing Americans, particularly evangelicals, while the Democratic Party’s constituency is increasingly secular. Fiorina points out that the correlation between partisanship and religious denomination has actually declined in recent years (mainline Protestants are less Republican and Catholics and less Democratic, for example), but agrees with Abramowitz that the correlation between partisanship and religiosity (one’s religious commitment, usually measured by frequency of church attendance) has grown stronger. While more religious Americans are increasingly supportive of the Republican Party, this shift is counterbalanced by the movement of secular and less religious Americans towards the Democratic Party. Thus, while neither party has particularly benefitted electorally from the changes, religiosity is nonetheless an increasingly powerful predictor of voting behaviour.
Where Abramowitz and Fiorina disagree is over the relative importance of religiosity. Fiorina argues, based on his analysis of presidential elections up to 2000, that religiosity is more important than it was but that it has not replaced the old New Deal cleavage of class. Indeed, income, as a proxy for class, seems to be growing in importance with poor Americans increasingly voting for Democratic candidates and richer ones for Republicans. The exception to this trend, acknowledges Fiorina, was the 2000 presidential election when income played a lesser role in determining vote choices, both absolutely and relative to religiosity. While Fiorina suggests that it would be premature to make any robust conclusions about trends based on the 2000 election alone, evidence from the 2004 presidential election suggests that 2000 was not aberration, but the beginning of a trend away from class and towards religion. Abramowitz’s data are persuasive. Controlling for other factors, religiosity is a considerably more powerful predictor of vote choice in 2004 among white voters than any other social characteristic, including income. Abramowitz predicted that the trend would deepen as secular voters increased as a proportion of the electorate and religiously committed voters aligned increasingly with the GOP.8 While 2008 did not see a deepening of the trend, it certainly saw its consolidation.9
One of the most popular arguments levied in favour of the culture war thesis is a reworking of the argument presented above that political parties in Congress are more clearly divided along ideological lines. Similarly, at the mass level, social conservatives are increasingly likely to support and vote for Republican candidates and social liberals for Democratic ones. The evidence seems indisputable and the phenomenon has been called party or partisan polarization. Fiorina does not challenge the conventional wisdom that this has occurred. Instead, he challenges its extent and what it means.10 First, regarding its extent, Fiorina acknowledges the increasing importance of cultural and social concerns, but argues that ordinary Americans are less differentiated by culture than are party elites and points to a large body of research that emphasises the increasing importance of class in structuring partisan affiliations. Second, and regardless of any evidence presented on the first point, Fiorina rejects its importance. The reason is that party affiliations and vote choices may have changed, but issue and ideological positions have not. Americans have sorted themselves more neatly by ideology and party, mirroring the elite level sorting in Congress, with Republican liberals switching to the Democratic Party and Democratic conservatives switching to the GOP. For this reason, Fiorina prefers the term party sorting to party polarization.
While Fiorina engages Abramowitz and other culture war theorists on questions of party, religious and geographic polarization, he does not actually believe that these are relevant indicators of polarization. According to Fiorina, polarization is best defined and measured by the distribution of politically-related attitudes: “The most direct way to measure polarization of political positions is to measure political positions.”11 Conservatives may be moving to red states and religious Americans may be increasingly lining up in support of the Republican Party, but these trends to not demonstrate that Americans are becoming more polarized. If Americans are polarizing, their opinions should be changing. Specifically, Americans’ opinions should be moving towards the poles of the distribution and away from the centre. For example, fewer people should over time have centrist opinions on abortion, and more should believe it should never be legal or be legal in all circumstances; and fewer Americans should define themselves as ideological moderates and more should self-identify as strong liberals or strong conservatives. Figures 3a, 3b and 3c present these theoretical distributions graphically. Most analysts would agree that opinion on any particular issue is not polarized in 3a, is somewhat polarized in 3b and is most polarized in 3c. To recap, Fiorina’s contention is that on most issues, even the hot-button ones that allegedly define the culture war, the distribution of opinion in the American population is best represented by the normal bell-curve in Figure 1a.
[Figures 3a, b and c about here]
The evidence on political positions appears to stack up in Fiorina’s favour. His own work and that of DiMaggio et al and Evans, based on exhaustive analysis of the available data, offer seemingly compelling evidence that Americans are not polarizing.12 In the aggregate, the issue positions of Americans—on abortion, the role and reach of government and so on—over the past four decades are best described as stable and centrist. Ideologically, a plurality of Americans still describe themselves as moderates, and only tiny percentages as extremely liberal or conservative. Adding moderates to “don’t knows” makes this group the majority.
Abramowitz’s response,13 based on the same data as Fiorina uses,14 is that Americans hold increasingly coherent ideological positions across a range of issues—with fewer conservatives taking any liberal positions and fewer liberals taking conservative ones. DiMaggio et al reached a similar conclusion regarding issue cohesion, although they called it issue constraint and consolidation. Fiorina argues in turn that issue coherence or consistency is not akin to polarization, because the overall distribution of positions on the issues has not changed—3a or 3b still best describes the distribution. But his focus on the aggregate distribution ignores the important individual-level changes occurring in the American population. That one is less likely to find a self-described ideological liberal expressing restrictionist views on reproductive rights or immigration and a self-identified conservative expressing liberal positions means that ideological sorting has occurred. Liberals and conservatives now differ more clearly and deeply on the issues because individuals have changed their issue positions. Moreover, ideological sorting is not restricted to the most politically knowledgeable and sophisticated citizens. Hetherington estimates that about three quarters of Americans could be said to be sorted by ideology and party.15 Fiorina’s aggregate data mask this important individual-level change. Whether such sorting constitutes evidence of political polarization is, however, disputed. Abramowitz says it does and Fiorina it does not. In part, the dispute rests on the causes of any shifts. We return to this point in the conclusion. The dispute also rests on how one defines polarization and cannot be resolved empirically, but if a more clear ideological divide between liberals and conservatives based on an individual-level shift in issue positions does not constitute polarization, then perhaps Fiorina may be accused of defining it out of existence.
Regardless of whether ideological sorting constitutes evidence of popular polarization, Fiorina has been criticised for underplaying the importance of issue salience. As Hetherington notes, “polarization suggests an intensity that draws on attitudes that people hold deeply,”16 but Fiorina does not weigh issues according to their relative salience in public discourse or people’s consciousness. Even if Americans are not more divided on cultural issues—with the issue distributions remaining constant, and perhaps even bell-shaped—“salience can make issues seem more polarizing even if the distance between the groups remains relatively small... If anything, the issue environment has become increasingly conducive to a culture war,” argues Hetherington.17 Hetherington offers the contrasting examples of civil rights in the 1960s and gay rights in the early twenty-first century. The preference distributions on both issues are very similar, but many more people identified civil rights as one of the most important problems facing America. It is perhaps this difference in issue salience, not the difference is preference distributions, which explains why civil rights was much more polarizing in the 1960s than is gay rights today.
While Hetherington does not conduct a rigorous analysis, his issue salience theory could potentially help explain part of the polarization puzzle. The puzzle is that Fiorina’s issue distribution data show little evidence of polarization, yet a widespread perception remains that ordinary Americans are polarized and/or polarizing. If cultural issues have become more salient at the same time that the distribution of opinions has remained constant, then one could perhaps conclude that Americans are more polarized and America is experiencing a culture war. The remainder of this chapter tests Hetherington’s salience theory.
First, an appropriate over time measure of issue salience needs to be constructed. This is not straightforward and there is no generally accepted instrument. However, Hetherington18 and others19 suggest that the Most Important Problem (MIP) question may offer analytical leverage. The American National Election Studies (ANES) surveys have asked Americans to name the most important problem facing America roughly every two years since 1966. Because the question is “open”, respondents do not pick from a pre-existing list of problems but instead must consider unprompted the relative importance that they attach to various issues and pick the one they consider the most pressing. The question taps into Hetherington’s understanding of issue saliency at the level of the individual citizen.
To construct our measure of issue saliency, the ANES MIP responses were scoured for any mention of cultural issues. Because issues evolve over time, the categories are not constant. Issues drop in and out of the response set. While this would normally be regarded as a potential problem by analysts, its flexibility is the key reason why the MIP question is an attractive measure. Table 1 below sets out which categories are included in each year.
[Table 1 about here]
If Hetherington’s issue salience theory is correct, the percentage of respondents mentioning cultural issues should increase over time. Figure 4 demonstrates that this is indeed the case. Apart from an uptick in 1970 and 1972, which was largely driven by anxiety over drugs, concern about cultural issues remained generally low in the 1960s and 1970s, despite a febrile political atmosphere generated by increasing opposition to the Vietnam war and growing concern over the rise of the anti-authoritarian and anti-traditionalist counter-culture. Even the Supreme Court’s 1973 Roe vs. Wade decision, which protected the rights of women seeking to terminate a pregnancy and is regarded as a staple of the culture war today, had no effect on the salience of cultural issues (it was not identified by any respondents as the most important problem). Concern began to increase again in the early 1980s, reaching its all-time high of 17.2 percent when George H. W. Bush faced Michael Dukakis in the 1988 presidential election. It dropped off over the next four years, perhaps as the first Gulf war and the recession diverted attention away from cultural concerns, but rose sharply from 1992 through the presidency of Bill Clinton, reaching 16.8 percent in 2000.
The salience of cultural issues appears to have declined dramatically between 2000 and 2004 (the MIP question was not asked in 2002). This fall is particularly strange because some commentators identified the 2004 presidential election between George W. Bush and John Kerry as the time when cultural and moral issues came to fore, a claim supported by data from the 2004 National Election Pool (NEP) exit poll, in which 22 percent of respondents said “moral values” was the most important issue determining their presidential vote. While the poll’s methodology and the interpretation that moral values “decided” the presidential election have been subject to scholarly criticism,20 other surveys (by Pew and the Los Angeles Times, for example) confirm that moral values were a significant source of concern. Why did the ANES survey not pick up on this concern, or at least measure it at a much lower level than the NEP and other 2004 polls? The answer lies in a combination of factors to do with the respective survey designs and the important issues of the day. Most notably, the ANES changed the wording of the MIP question. In previous years, it had asked respondents to think about the present but in 2004 it directed them to consider the past four years. Given that 9/11 had occurred during the specified four-year time period, America had engaged in wars in Afghanistan and Iraq, and respondents could select only their single most important concern, it is easy to understand why concern about cultural issues would be trumped by such events. We cannot know what the result would have been had the question not changed and events been different, but it is likely that they significantly diminished the number selecting cultural issues. We should thus be wary about the 2004 data point, and not conclude that cultural issues didn’t matter. Unfortunately, the ANES has still not released the results of its 2008 MIP question.
[Figure 4 about here]
In sum, and recognising the 2004 data point is problematic, it seems appropriate to conclude that Hetherington is right. Thinking about issues in terms of their salience, rather than just their distribution as per Fiorina, offers empirically interesting and theoretically important insights into the culture war debate. Even if Fiorina is correct, and Americans are not significantly more divided in terms of where they stand on the hot-button issues such as abortion, gay rights, family values and the role of religion in public and private life, it is very likely that they are more polarized because these issues have grown in importance over the past three decades. This finding poses a challenge to scholars who continue to reject the argument that ordinary Americans are engaged in battle over moral and cultural issues. It also allows observers to place recent events, such as the rise of the Tea Party and the vehement opposition of ordinary conservatives to the Obama administration, in the context of a polarizing polity. The apparently feverish climate of political discourse can in part be explained by the increased salience of cultural issues among ordinary Americans, not just political elites.
CAUSES OF POLARIZATION
Has elite polarization caused ordinary Americans to become more polarized, or are elites responding, as perhaps they should in a democratic polity, to polarization at the mass level? Alternatively, it is possible that elite and mass polarization may be causally unrelated, and instead be the product of one of more other factors.
Despite the potentially messy causal relationships, Hetherington is clear that elite polarization is responsible for the increased salience of cultural issues at the mass level.21 But how? One potential mechanism is that political operatives like Karl Rove, George W. Bush’s close advisor, have adopted electoral strategies that seek to turn out the party’s base rather than moderate voters.22 To do so, parties, especially the Republicans, have focused their campaigns on divisive issues such as abortion, gay rights and school prayer that appeal to core supporters and motivate them to vote in large numbers. Alternatively, suggests Thomas Frank, Republicans may have used cultural issues to divide the Democratic coalition, splintering off conservative Democrats and realigning them with the GOP despite the fact that the Democratic Party best represents their economic interests.23 As noted above, however, other scholars suggest that economics actually matters more today, not less as Frank suggests. An increasingly partisan and ideological media may also be part of the story, widening, or at least entrenching, ideological differences between partisans, as conservatives tune into Fox while liberals watch MSNBC, for example. The politically polarized nature of the blogosphere and talk radio may exacerbate this effect, suggests Hetherington.24 A subsequent question, however, is why has the media polarized? Is it caused by the same things that are causing other aspects of elite polarization, or by the polarization going in other parts of the elites’ universe, such as Congress, or is it a response to technological developments and/or changing tastes at the mass level? The causal story is difficult to disentangle.
The causes of elite-level polarization are equally difficult to establish.25They are opaque, contested and many. Of the probable causes identified by Hetherington in his encyclopaedic review of the literature, perhaps the most notable is the Democratic Party’s adoption of the civil rights agenda in the 1940s and ‘50s and the subsequent loss of its southern hegemony.26 The “grand bargain” between northern liberal and southern conservative Democrats unravelled (where southerners had supported liberal economic programmes while northerners turned a blind eye to segregationist practices), producing more ideologically cohesive parties. The Democratic Party’s congressional caucus became more universally liberal as its pro-segregation southern members lost their seats to conservative Republicans or switched parties. It is also likely that procedural changes enacted in the 1970s, particularly primary contests,27 interacted with the growing radicalisation of grassroots party activists to produce congressional and presidential candidates less representative of the median voter and more representative of each party’s ideological poles.28 While primaries were designed to weaken the grip of party elites and unsurprisingly have hitherto been interpreted as evidence of party decline, it appears they have in fact engendered party resurgence in both Congress and country. Of course, the subsequent question of why Republican Party activists have moved rightwards and Democratic ones leftwards is a thorny one. Any explanation should probably acknowledge the role of elite discourse and particularly the role elites play in framing issues,29 but it is also likely that activists in turn played a role in elites’ own polarization.30 Such endogeneity makes it very difficult to pin down the relative contribution of the respective causes. That there are likely many other varied and interconnected causes—institutional, non-institutional and cultural31—further undermines efforts to settle on a robust explanatory model. More research in this area is required.
CONSEQUENCES AND SOLUTIONS
The key consequence of polarization, say its critics, is a malfunctioning polity.32 In the absence of very large majorities in both chambers of Congress, and particularly the Senate, the parties’ failure to work together militates against the passage of important solutions to today’s problems. Indeed, the parties are more likely to work against than with each other. For example, “the single most important thing we want to achieve,” declared Senate Minority Leader Mitch McConnell, is “for President Obama to be a one-term president.”33 Given the Republicans’ priority is defeating the incumbent president rather than addressing the financial crisis or growing the economy or rethinking America’s role in the world, it is perhaps not surprising that only 12 percent of the American people have a “great deal” or “quite a lot of confidence” in the US Congress (compared to 35% in the presidency and 37% in the Supreme Court),34 and that only 10 percent “approve of the way Congress is handling its job”35—the lowest level since Gallup began asking this question in 1974. Yet, the opprobrium of the American people has, hitherto, failed to refocus its members’ attentions on substantive issues of policy and governance. At the time of writing, the institution is convulsed by partisan conflict and there appears to be little prospect that the parties will emerge from their trenches in an entente cordiale.
Prospective solutions to elite polarization are varied. Those that require constitutional change—such as aligning the electoral terms of representatives, senators and presidents—will struggle to clear the hurdle of the constitution’s own amendment procedure. Quirk’s suggestion is to attack the problem at one of its alleged sources: the primary process.36 This could be done at the state level and on a state-by-state basis, and would not require the congressional turkeys to vote for Christmas. California’s new nonpartisan blanket primary for state and congressional elections, approved by voters as Proposition 14 in 2010, arrays all prospective candidates on the same list irrespective of party affiliation. Its effects will be watched eagerly, but could include forcing politicians to seek out the median voter rather than those at the ideological poles.
The discussion above highlights a broad consensus among academics and observers of the political scene that American elites are deeply divided on cultural and social issues. No such consensus exists on the issue of polarization among ordinary Americans. Fiorina’s and Abramowitz’s work represents two sides of this debate. Much of the dispute between them is ultimately definitional. What Abramowitz sees as polarizing, Fiorina sees as sorting. There is no clear, scientific way to settle their disagreement and the debate in the wider literature. Nonetheless, this chapter is sympathetic to Fiorina’s position that geographic, religious and party sorting does not constitute evidence of mass polarization.
It is less sympathetic to the idea that ideological sorting does not constitute evidence of mass polarization. We know that self-identified liberals increasingly hold fewer conservative positions on the issues and self-identified conservatives fewer liberal ones. The aggregate distribution of opinion may not have changed, with roughly stable numbers of liberal, moderate and conservative positions on any given cultural or social issue, but this masks an important change at the individual level. Liberals have moved towards the liberal end of the distribution and conservatives towards the conservative end. They are more divided ideologically because they have fewer overlapping value positions. Is that polarization or sorting? Probably both, but a firm answer depends in part on whether the shift is a consequence of real value change that pits more ideologically coherent liberals and conservatives against each other, or whether liberals and conservatives take more coherent positions across issues not because their opinions have fundamentally changed but because elite partisans are giving more clear signals about where they should say they stand. In other words, is the value change real or simply rhetorical? The answer is that we do not know. More research is required. Most likely, both processes are at work and interacting with each other. Importantly, however, public opinion is a combination of the two. When political scientists say that elites shape public opinion, they mean that elites influence what people think, feel and say without being able to distinguish clearly between them. Given that public opinion is what the public says its opinion is, we should probably, until we know better, interpret the data to mean that the public is more coherently divided ideologically because the public has told us it is more coherently divided ideologically. If this interpretation is correct, it undermines Fiorina’s proposition that ordinary Americans are not polarized or engaged in a culture war.
Fiorina has also been criticised by Hetherington for fixating on the distribution of opinions and ignoring the relative importance of the issues under investigation. Even if there has been no change in terms of the distribution of opinion on hot-button cultural issues, Americans are likely to be more divided if those issues have grown in importance. Polarization, for Hetherington at least, is a product of opinion distribution and issue salience. To explore this idea, this chapter presented some new data on the salience of cultural issues in the US over the past four decades. The trend is clear. The number of Americans identifying cultural issues as the most important problem facing the country increased markedly during the 1980s and 1990s. Put differently, that these have become increasingly important to ordinary Americans suggests the culture war is both an elite- and mass-level phenomenon, whose reach should alarm all observers concerned about the health of American democracy.
Alan I. Abramowitz, (2011), The Disappearing Center: Engaged Citizens, Polarization, and American Democracy (New Haven: Yale University Press).
Alan I. Abramowitz and Kyle L. Saunders (2008), “Is Polarization a Myth?” Journal of Politics, Vol. 70, pp. 542-555.
Edward G. Carmines and Geoffrey C. Layman, (1997), “Issue Evolution in Post-war American Politics: Old Certainties and Fresh Tensions” in Byron E. Shafer (ed.), Present Discontents: American Politics in the Very Late Twentieth Century (Washington DC: CQ Press).
Paul DiMaggio, John Evans, and Bethany Bryson, (1996), “Have Americans’ Social Attitudes Become More Polarized?” The American Journal of Sociology, Vol. 102, No. 3, pp. 690-755
John H. Evans, (2003), “Have Americans’ Attitudes Become More Polarized?—An Update,” Social Science Quarterly, Vol. 84, No. 1, pp. 71-90.
Morris P. Fiorina, (2005), Culture War? The Myth of a Polarized America (New York, Pearson Longman).
Fiorina, Morris P., and Samuel J. Abrams, (2008), “Political Polarization in the American Public,” Annual Review of Political Science, Vol. 11, pp. 563-588.
Morris P. Fiorina and Samuel J. Abrams, (2009), Disconnect: the Breakdown of Representation in American Politics (Norman, Oklahoma: University of Oklahoma Press).
Thomas Frank, (2004), What’s the Matter with Kansas? How Conservatives Won the Heart of America (New York: Metropolitan Books).
Jonathan Freedland, “Barack Obama’s Presidency, Three Years On—Is it Time to Give Up Hope?” The Guardian, 19 January 2012.
Kevin Fullam and Alan R. Gitelson, (2009) “A Lasting Republican Majority? George W. Bush’s Electoral Strategy” in Andrew Wroe and Jon Herbert (eds.), Assessing the George W. Bush Presidency: A Tale of Two Terms (Edinburgh: Edinburgh University Press).
John Green, Pew Research Center Forum on Religion and Public Life's Faith Angle Conference, December 2008. Marc J. Hetherington, (2009), “Putting Polarization in Perspective,” British Journal of Political Science, Vol. 39, p. 413-448.
Marc J. Hetherington, (2001), “Resurgent Mass Partisanship: The Role of Elite Polarization,” American Political Science Review, Vol. 95: 619-631
Geoffrey C. Layman and Edward G. Carmines, (1997), “Cultural Conflict in American Politics: Religious Traditionalism, Postmaterialism, and US Political Behavior,” Journal of Politics, Vol. 59, No. 3, pp. 751-777.
Kara Lindaman and Donald P. Haider-Markel, (2002), “Issue Evolution, Political Parties, and the Culture Wars,” Political Research Quarterly, Vol. 55, No. 1, pp. 91-110.
Young Min, Salma I. Ghanem, and Dixie Evatt, (2007), “Using a Split-Ballot Survey to Explore the Robustness of the ‘MIP’ Question in Agenda-Setting Research: A Methodological Study,” International Journal of Public Opinion Research, Vol. 19, pp. 221-236
Keith Poole, “Congressional Policy Shifts, 1879-2011,” Voteview Blog, 15 Feb 2012, at http://voteview.com/blog/.
Paul J. Quirk, (2011), “A House Dividing: Understanding Polarization,” The Forum, Vol. 9, Issue 2, Article 12.
Howard Schuman, ‘The Validity of the 2004 “Moral Values” Question’, The Forum, 2006, Vol. 4, No. 2, Article 5.
Michael Shear, “Is Bipartisanship Back? Don’t Count on It,” New York Times, 1 March 2012.
Tom W. Smith, (1985), “The Polls: America’s Most Important Problems; Part I: National and International,” Public Opinion Quarterly, Vol. 49, pp. 264-274.
Figure 1. Poole-Rosenthal DW-Nominate House Polarization Score
Narcotics control; drug addiction; pep pills, LSD, marijuana etc
Problems of young people; e.g. drinking, discipline, mixed up thinking
Licensing, control of guns; other weapons
Moral, religious decay of nation (general)
Problems of young people; e.g. drinking, sexual freedom, discipline, mixed up thinking, hippies; communication with young; generation gap
Codes Included by Year
363, 365, 380, 382, 383
363, 365, 380, 382, 383
363, 365, 380, 382, 383
363, 364, 365, 380, 382
Notes: 2008 MIP Q not yet coded by ANES; MIP Q not asked in 2002; MIP numerical codes are associated wording are the ANES’s own not the authors; Codes included by year selected by author
1 Snowe quoted in Michael Shear, “Is Bipartisanship Back? Don’t Count on It,” New York Times, 1 March 2012.
2 Snowe quoted in Shear, “Is Bipartisanship Back? Don’t Count on It.”
3 Keith Poole, “Congressional Policy Shifts, 1879-2011,” Voteview Blog, 15 Feb 2012, at http://voteview.com/blog/.
4 Alan I. Abramowitz and Kyle L. Saunders (2008), “Is Polarization a Myth?” Journal of Politics, Vol. 70, p. 542.
5 Morris P. Fiorina, (2005), Culture War? The Myth of a Polarized America (New York, Pearson Longman), p. 8.
6 Fiorina, Culture War?pp. 11-32.
7 Abramowitz and Saunders, “Is Polarization a Myth?” p. 548.
8 Abramowitz and Saunders, “Is Polarization a Myth?”
9 John Green, Pew Research Center Forum on Religion and Public Life's Faith Angle Conference, December 2008.
10 Morris P. Fiorina and Samuel J. Abrams, (2008), “Political Polarization in the American Public,” Annual Review of Political Science, Vol. 11, pp. 578-580.
11 Fiorina and Abrams, “Political Polarization in the American Public,” p. 569.
12 Paul DiMaggio, John Evans, and Bethany Bryson, (1996), “Have Americans’ Social Attitudes Become More Polarized?” The American Journal of Sociology, Vol. 102, No. 3, pp. 690-755; John H. Evans, (2003), “Have Americans’ Attitudes Become More Polarized?—An Update,” Social Science Quarterly, Vol. 84, No. 1, pp. 71-90.
13 Abramowitz and Saunders, “Is Polarization a Myth?” p. 544-545.
14 Fiorina and Abrams, “Political Polarization in the American Public,” p. 573.
15 Marc J. Hetherington, (2009), “Putting Polarization in Perspective,” British Journal of Political Science, Vol. 39, p. 438.
16 Hetherington, “Putting Polarization in Perspective,” p. 429.
17 Hetherington, “Putting Polarization in Perspective,” pp. 430 and 435.
18 Hetherington, “Putting Polarization in Perspective,” pp. 434-436.
19 Young Min, Salma I. Ghanem, and Dixie Evatt, (2007), “Using a Split-Ballot Survey to Explore the Robustness of the ‘MIP’ Question in Agenda-Setting Research: A Methodological Study,” International Journal of Public Opinion Research, Vol. 19, pp. 221-236; and Tom W. Smith, (1985), “The Polls: America’s Most Important Problems; Part I: National and International,” Public Opinion Quarterly, Vol. 49, pp. 264-274.
20 Howard Schuman, ‘The Validity of the 2004 “Moral Values” Question’, The Forum, 2006, Vol. 4, No. 2, Article 5.
21 Hetherington, “Putting Polarization in Perspective,” p. 441.
22 Kevin Fullam and Alan R. Gitelson, (2009) “A Lasting Republican Majority? George W. Bush’s Electoral Strategy” in Andrew Wroe and Jon Herbert (eds.), Assessing the George W. Bush Presidency: A Tale of Two Terms (Edinburgh: Edinburgh University Press).
23 Thomas Frank, (2004), What’s the Matter with Kansas? How Conservatives Won the Heart of America (New York: Metropolitan Books).
24 Hetherington, “Putting Polarization in Perspective,” p. 443.
25 See, for example, Paul J. Quirk, (2011), “A House Dividing: Understanding Polarization,” The Forum, Vol. 9, Issue 2, Article 12.
26 Hetherington, “Putting Polarization in Perspective,” p. 421.
27 Morris P. Fiorina and Samuel J. Abrams, (2009), Disconnect: the Breakdown of Representation in American Politics (Norman, Oklahoma: University of Oklahoma Press).
28 Alan I. Abramowitz, (2011), The Disappearing Center: Engaged Citizens, Polarization, and American Democracy (New Haven: Yale University Press).
29 Marc J. Hetherington, (2001), “Resurgent Mass Partisanship: The Role of Elite Polarization,” American Political Science Review, Vol. 95: 619-631; Geoffrey C. Layman and Edward G. Carmines, (1997), “Cultural Conflict in American Politics: Religious Traditionalism, Postmaterialism, and US Political Behavior,” Journal of Politics, Vol. 59, No. 3, pp. 751-777; but see Kara Lindaman and Donald P. Haider-Markel, (2002), “Issue Evolution, Political Parties, and the Culture Wars,” Political Research Quarterly, Vol. 55, No. 1, pp. 91-110.
30 Edward G. Carmines and Geoffrey C. Layman, (1997), “Issue Evolution in Post-war American Politics: Old Certainties and Fresh Tensions” in Byron E. Shafer (ed.), Present Discontents: American Politics in the Very Late Twentieth Century (Washington DC: CQ Press).
31 This is Hetherington’s typology in “Putting Polarization in Perspective”.
32 Hetherington, “Putting Polarization in Perspective,” pp. 427-429; Quirk, “A House Dividing,” pp. 8-11.
33 Quoted in Jonathan Freedland, “Barack Obama’s Presidency, Three Years On—Is it Time to Give Up Hope?” The Guardian, 19 January 2012.