The Myth of a Non-Polarized America Carl Desportes Bowman

The Myth of a Non-Polarized America

Carl Desportes Bowman

Reprinted from The Hedgehog Review 12.3 (Fall 2010). This essay may not be resold, reprinted, or redistributed for compensation of any kind without prior written permission. Please contact The Hedgehog Review for further details.

By 2005, a long, hard-fought scholarly debate, which had lasted over twenty years, finally reached a consensus: there was no “culture war” in America. Extinguishing the debate originally ignited by James Davison Hunter’s 1991 book on the topic,1 a legion of reputable scholars produced evidence to support this conclusion. Paul DiMaggio and co-authors were among the first, publishing an analysis of General Social Survey data in 1996 concluding that Americans, if anything, were becoming more similar.2 Shortly thereafter, Alan Wolfe’s qualitative study, One Nation After All: What Americans Really Think.... added to the cases. His interviews with American suburbanites from four urban areas led Wolfe to conclude, “there is little truth to the charge that middle-class Americans, divided by a culture war, have split into two hostile camps.”3 Among scholars, Morris Fiorina’s 2005 book, Culture War? The Myth of a Polarized America dowsed whatever flame remained in the culture war embers. The evidence was in. Polarization was passé.

The conclusion, however, was oddly incompatible with continuing public commentary and the human experiences that lent it credibility. Scholars declared the culture war a fiction, but political activists, media commentators, and gas-station philosophers continued to speak, think, and behave as if the war were ongoing. Could they be so wrong? Could it be, as a blurb on the back of Wolfe’s book claimed, that “our politicians, media polls, and social critics have no idea what Americans think”? Was polarization, as Fiorina claimed, nothing more than a myth?

As it turns out, even the scholars weren’t entirely persuaded.

New Evidence

The last three years have produced a fresh harvest of books on polarization, several of which present new empirical evidence of American cultural and political polarization.4 This essay briefly presents the argument of two of these books, followed by additional evidence from the Institute for Advanced Studies in Culture’s 2003Survey on Difference and Democracy.

First, let us consider Alan Abramowitz’s The Disappearing Center. Abramowitz contends that the deeply partisan divide in Congress mirrors divisions in the broader public. Even though Abramowitz acknowledges that such differences are most pronounced among the politically engaged, he rejects Fiorina’s postulation of a basic disconnect between political elites and the public.5 Instead, Abramowitz contends that on a variety of issues “the overall distribution of ideological preferences among the public, and especially among the politically engaged, has shifted from a unimodal distribution toward a bimodal distribution.”6 Graphically, the argument is that the contour of public opinion has shifted from A, where most cluster in the center, to B, where groups have moved away from the center in opposite directions. (See Figure 1.)

Additionally, Abramowitz observes that:

• Social cleavages based upon economic class differences are diminishing, as are those associated with religious affiliation.
• Yet differences anchored in religious commitments have widened. A “religious divide” in public culture, most notably among white voters, pushes cultural issues to the center of public debate.
• This religious and ideological divide is increasingly aligned with party identification
• Reduction in the number of competitive congressional districts, among other things, is related to the residential migration of progressives and conservatives to areas that fit them culturally. Increasingly Americans surround themselves with people who see the world as they do.7
• The polarized rhetoric and policies of the major parties, rather than causing a generalized estrangement from politics, is associated overall with increasing political engagement.

These observations are grounded in a longitudinal analysis of the federally funded National Election Studies and other national data. Scales of political liberalism-conservatism are employed to measure movement toward the ideological extremes. Contrary to findings by DiMaggio, Ambramowitz’s longitudinal comparisons suggest that Americans have indeed moved further away from one another ideologically.

Abramowitz challenges what he calls the “conventional wisdom” upheld by Fiorina that the public is politically uninterested, uninformed, and unsophisticated.

According to Fiorina, the ideological disputes that engage political elites and party activists have little resonance among the American mass public…this portrait of the American public as generally apathetic with regard to politics is fundamentally flawed. In the first place, sweeping generalizations about the political beliefs and behavior of ordinary Americans ignore the fact that there are vast differences in political interest, knowledge, and activity among the public…evidence from recent surveys of the electorate show that today the engaged public is quite large.8

And it is this “quite large engaged public” that, Abramowitz concludes, has polarized, aligning their party preferences with their ideologies.

Evidence that polarization extends beyond the political elite is also presented in Marc J. Hetherington and Jonathan D. Weiler’s Authoritarianism and Polarization in American Politics. The reasons scholars have difficulty finding it, they say, are both conceptual and methodological. Conceptually, polarization may exist only on certain clusters of issues, defying detection if issues are inappropriately aggregated or if the wrong issues are examined. Polarization can also shift from issue to issue over time, so tracking the level of disagreement on a particular issue might miss it. And the lack of consensus as to what constitutes “polarization” exacerbates the situation; what one scholar calls polarization another might characterize as nothing more than “partisan sorting.”

Empirically, Hetherington and Weiler make an intriguingly counter-intuitive argument: polarization may increase even as disagreement declines. That is, less extreme views on a topic (such as gay marriage) can increase public discussion, which may then be manipulated by partisan leaders who fan the flames to their own advantage. Visibility of the issue might thereby increase and the emotional stakes rise to the point where the public is more polarized even though their “opinions” remain much the same. In other words, emotional salience is as basic to polarization as difference of opinion.

Hetherington and Weiler level a bevy of methodological criticisms against studies that have found little polarization. One objection is that the “movement of opinion toward opposing poles” is an unrealistic criterion. Survey respondents, they point out, commonly shy away from extreme responses and gravitate toward the middle. Moreover, there is no scholarly consensus as to how much divergence is required to describe opinions as polarized. Differences as large as 15 percentage points are dismissed by Fiorina, note Hetherington and Weiler, especially if two groups are on the same side of the midpoint. Yet historically, such differences are nearly as large as the mid-1960s opinion gap on segregation between northerners
and southerners, when the regions were widely viewed as polarized.

All things considered, Hetherington and Weiler conclude,

the definition most often used by scholars to test for polarization [movement toward the poles] is apparently impossible to meet. Given the place that politics occupies in the life of most people, their preferences are unlikely ever to cluster at ideological poles. To the extent that we use the term polarization, we must carve out a different understanding of it. To us, the essence of polarization is when hot-button issues become salient concerns for a large percentage of people.… It might be the case that people’s preferences are not that different from those of their opponents, but they do not see it that way.9

In Search of Explanations

What then makes something a “hot-button issue”? The most volatile issues of our period, Hetherington and Weiler contend, are those for which responses are structured by authoritarianism.

After a brief tour of authoritarianism’s checkered scholarly history, the authors supply their own conceptual understanding: an authoritarian disposition includes the need for order, a tendency to simplify the world into black and white categories, a tendency to feel threatened by ambiguity and outsiders, and a reliance upon established authorities to preserve order. The fundamental divide in America, they say, “is not between two groups with the same psychological disposition who merely disagree.” But more fundamentally, disagreements are “animated by fundamentally different dispositions” which inform “dramatically different worldviews.”10

In short, they argue that distinct personality styles are the key to understanding which issues polarize. Economic issues, for example, being equally threatening to authoritarians and nonauthoritarians, generate little polarization. Cultural issues, by contrast (pertaining to race, sexuality, gender, family, religion, patriotism, terrorism, criminal justice, civil liberties, and immigration), polarize because they uniquely threaten authoritarians. Since authoritarians and nonauthoritarians were similarly threatened by the attacks of 9/11, their responses were similar, fostering national solidarity. But many things that threaten authoritarians—gay marriage, non-Christian religions, non-English speakers, to name a few—are not so threatening to those lacking their predilection for order, authority, and tradition. Hetherington and Weiler conclude that conflicts between the two groups “go far beyond disagreements over policy choices and even ideology, to conflict about core self-understandings of what it means to be a good person and to the basis of a good society.”11

The authors seek to identify the deeper sources of cultural conflict, and yet, rather than explaining polarization culturally, in terms of contrasting moral orders with disparate conceptions of moral authority, they invoke George Lakoff’s psycho-linguistic distinction between the “strict father” and the “nurturant parent.” In fact, the use of psychological categories to explain polarization is typical of recent studies that venture into the terrain of explanatory frameworks. This excerpt from Naomi Cahn and June Carbone’s Red Families, Blue Families is a good example:

Conservatives celebrate the “strict father,” who enforces relatively fixed and hierarchical values, an identification that is similar to our notion of “traditionalists.” Liberals prefer the “nurturing mother” who makes context-based decisions designed to promote individual well-being, a concept similar to our “modernists.” Those who share the strict father mentality see the world as dangerous; children need to be protected, and it is the responsibility of the strict parent to impose discipline on the children. Children are born bad and learn through punishment. By contrast, the nurturing mother mentality views the world as basically safe, with parents responsible for nurturing their children with empathy and responsibility. Different ways of framing issues—including the calls for a reaffirmation of traditional values versus insistence on the need for greater acceptance of diverse family forms—appeal to different world views.12

Whatever the theoretical merits of Lakoff’s argument, its application by others readily leads to an association of one moral orientation with pathology, that of the strict father.13 Hetherington and Weiler, for instance, write that those scoring high on authoritarianism “tend to rely more on emotion and instinct than those scoring low because they (1) have, on average, fewer cognitive tools and (2) feel more threat from the often ambiguous nature of the complicated world around them.”14 In the presence of threat, their “cognition breaks down” and emotion “picks up the slack.” How fortunate for nurturant, low-authoritarian progressives that they can rationally contemplate such matters! The ability of low-authoritarians to deal in abstraction indicates that they “likely make up a disproportionate share of those who progress to…[Kohlberg’s] more advanced stages of moral reasoning.”15 In such analyses, the lines between “conservative moral orientation” and “lower form of moral reasoning” are thin, if they exist at all.

While Hetherinton and Weiler use language and arguments that in many cases echo Hunter’s Culture Wars, they mention his work only in passing.16 However, a look at Hunter’s interpretive framework is helpful. First and foremost, he sees the conflict as cultural in the broadest sense. As a cultural conflict, it entails a polarization of elites, and consequently, of public discourse on a convergence of issues. Second, Hunter focuses attention upon radically different understandings of the world; partisan identification and the ballot box are not the analytic bottom line. Like Lakoff, he digs beneath transient political issues to ask what the enduring sources of our differences are, but a different lens is brought to bear. Rather than highlighting family metaphors, his analysis highlights these questions: What criteria do people rely upon to deem something “good” or some action “appropriate”? How does this inform their understanding of whether the “good and appropriate” applies to others? How do answers to these questions map onto understandings of: the past and future, family roles, goals for children, personal freedom, social obligation and responsibility, and one’s place in the scheme of things?

Basic to the cultural conflict that Hunter describes is a critical difference in how Americans approach the “criteria for something good” question above. Although they may seek counsel from others, some people take the modernist view that they are their own moral arbiters, that in the end, their own conscience, goal, or desire is the final authority. Others see moral authority residing outside themselves—in “God,” a sacred text or tradition, a faith community, or even the ways of their ancestors. The key, in a nutshell, is whether “the good” is internally or externally given (and in the case of the former, the extent to which it is radically subjectivized or understood to require input from, and obligations toward, something else). This basic difference shapes the way Americans respond culturally. They experience cultural tensions because their lives are constructed differently from the ground up. Whether conceived as cultural conflict, polarization, or in some other way, the clutter of fleeting (and not so fleeting) political differences are grounded in radically different cultures. 

Evidence of Culture

“Culture” is difficult to grasp, let alone measure. The 2003 Survey on Difference and Democracy, tried to get at these deep cultural differences in the kind of questions survey participants were asked.17 For example, the survey asked: “We all have hopes and dreams for our country, but we don’t all hope for the same things. For each of the following, which phrase comes closest to your own personal wish for the future of life in America?” Respondents were then asked to pick the phrase from sets of paired alternatives that best represented their hope for the future. Among the pairs were:

In each of these cases, the first statement represents a more fluid moral vision, while the second represents a more uniform or traditional one.

Whether Americans settle on one side or the other varies depending on the statement’s content. A clear majority, for example, hopes for increasing religious and family diversity and rejects the notion of the “stay-at-home mom.” Yet an equally clear majority hopes for clearer moral guidelines and increasing unity of moral commitments. Even so, responses to the items correlate highly, enabling us to construct a “Future Hopes Index” with scores ranging from “progressive” to “traditional.”18 The overall distribution of responses on the scale approximates a normal distribution, with few Americans at either extreme. (See Figure 2.)

Fiorina’s Culture War? The Myth of a Polarized America documents similar response patterns on many issues. His conclusion was noted above: no “culture war” in the general population. But when Future Hopes are considered in light of social forces that may shape them, such as religiosity (see left graph in Figure 3), or when the Future Hopes of different voting blocks are compared (see right graph in Figure 3), the picture is different. In both cases, the comparison reveals groups whose typical responses differ markedly.19

The most religious Americans hope for a more traditional and morally unified nation than those whose religious beliefs are “not too important.”20 Regarding voting, nearly 90 percent of those who score 0.5 or more toward the “traditional” side on Future Hopes said they would vote for George W. Bush, compared to only 43 percent of those whose scores were similarly progressive (-0.5 or less).21 This degree of cultural constraint—the fact that voting preference is so closely linked to something as diffuse as the moral order hoped for by Americans—may not document a deep rift dividing Americans into two opposing camps, but it effectively underscores the relationship of moral orders to religious commitment, on the one hand, and political consequences, on the other.

Alan Wolfe’s case against American polarization conceded that if a culture war exists anywhere, it is on the issue of homosexuality.22 The division on this issue detected in his qualitative interviews is borne out by the Survey on Difference and Democracy. We asked respondents whether they “completely agree,” “mostly agree,” “mostly disagree,” or “completely disagree” with several statements on the issue, among them: 1) Homosexual couples—that is couples of the same sex—should have the right to marry. 2) Homosexual couples should be allowed to adopt children. 3) Homosexuality should be considered an acceptable alternative lifestyle. Responses to these statements were then combined in a simple additive “Approval of Homosexuality Index,” with scores ranging from 3 (strong opposition) to 12 (strong approbation). The distribution of scores on the index can be seen in Figure 4. The departure from the bell-shaped distribution seen earlier is striking. Fully 27 percent of Americans give the “completely disagree” response to all three statements. What is more, the next most common combined response comes from Americans who “completely agree” with all three statements. The polarization is even greater among Americans who express a strong interest in politics.

A crude sense of the combined power of cultural factors for predicting political behavior can be seen by bisecting the response distributions of both the Future Hopes and Approval of Homosexuality scales, leaving us with only two categories for each. If scores on the two scales are unrelated, we would find Americans distributed evenly across the resulting four categories, with about 25 percent in each. But the actual distribution reveals that those who hope for a traditional, morally unified America tend to disapprove of homosexuality, while those who envision a fluid morality are more approving. (See Figure 5.) This is hardly surprising. Even though the survey statements about homosexuality make no explicit reference to America’s future, divergent concerns about the future often spark discussions of homosexuality. The two are interwoven not only empirically, but conceptually.

It is not unusual for scholars to discount such cultural measures, arguing that they have little impact beyond the purely cultural realm. Yet if such were the case, one would not expect to see the relationship between “culture” and voting expressed in Figure 6. The vote differential between those with a consistently traditional moral orientation (rejecting homosexuality and holding a traditional vision of the future) and those with a consistently progressive moral orientation (accepting homosexuality and holding a fluid vision of the future) is dramatic. The former are more than three times likelier to have voted for Bush in 2000. Keep in mind the distribution of Americans among the four categories in the chart: the consistently traditional and consistently progressive bars together account for 70 percent of those surveyed.

Thus far, cultural ideals and political behavior appear closely related, but perhaps the relationship is merely an artifact of something else, like political ideology or education. To explore the question, I estimated a logistic regression model the statistical details of which need not concern us.23 Suffice it to say that the technique enables me to simulate a situation in which certain phenomena are constrained, in order to isolate the impact of others. If differences in political ideology (self-identification as a liberal, moderate, or conservative), for example, could somehow be eliminated, would Future Hopes still predict voting patterns? Differently stated, are differences in moral visions of America (Future Hopes) still connected to voting behavior if we consider only those who call themselves “liberal” or “conservative,” so that differences between liberals and conservatives do not enter into the picture? If purely cultural differences were inconsequential, we would expect their predictive value to vanish in the presence of such controls. In this case, I was trying to predict whether people said they would vote for “George W. Bush” in a hypothetical election.

In the presence of controls, not only did Future Hopes and moral assessments of homosexuality retain importance, but standard demographic factors associated with voting—gender, education, urbanism, and region of the country—lost theirs. Not to say that our data suggest education and urbanism are unrelated to voting. But they do suggest that the impact of demographic factors is mediated through cultural differences, differences in moral outlook. Political ideology, of course—whether people consider themselves to be “liberal” or “conservative”—remains an important predictor of voting even when other cultural factors are held constant. But the same is true for the cultural indices, they remain important even when political ideology and all of the standard demographic factors are controlled.

An additional chart will help to visualize this. (See Figure 7.) The chart depicts the impact of both political ideology and moral vision (Future Hopes) upon the likelihood of selecting Bush when income and approval of homosexuality are held constant. The fact that each curve rises to the right tells us that visions of America’s moral future predict hypothetical voting outcomes even when differences related to income and views of homosexuality are statistically erased. The top curve tells us that self-identified “conservatives,” even the rare ones who cherish a progressive hope for moral diversity, are likely to choose Bush, but they are virtually certain to choose Bush if they espouse a traditional vision of the future. The slope of the other two curves suggests that the voting preferences of moderates and liberals may be more directly tied to their vision for America’s future.24

The empirical story is a complex one and difficult to unravel, but it suggests several things. First, the fact that educational, gender, residential, and regional voting differences largely disappear after controlling for cultural factors and political ideology suggests that their influence is mediated by culture—that social context is important precisely because it is the context in which cultural narratives are born, live, and are transformed through the dynamic discourse of people who share similar circumstances. Second, it suggests that cultural measures should be taken seriously by those attempting to sort through the contemporary cultural and political polarization. Third, relying exclusively upon measures of political ideology or partisanship to convey the story of polarization leaves much unsaid. Moral orientation, as reflected by particulars such as opposition to homosexuality, are closely entwined with political partisanship and ideology, to be sure, but not reducible to them. The independent sources and consequences of moral orientation, broadly speaking, merit investigation. As this occurs, measures of faith orientation and commitment that go well beyond denominational affiliation or even church attendance should be attended to. How often one prays, for example, rarely enters into such investigations, but may be at least as strong an indicator of religious commitment as attendance at religious services.

A Final Suggestion

A final suggestion is this: before making pronouncements about the extent of American polarization, analysts should expend more effort digging into the shifting sands of cultural consensus and dissensus.

The Survey on Difference and Democracy, for instance, reveals that most Americans share this much in common: They see terrorism, nuclear weapons, chemical and biological weapons, and infectious diseases as serious threats to the world’s future. Religious and cultural differences between nations, in contrast, are not considered seriously threatening; just as religious, cultural, and family diversity within America are broadly accepted, and evidence to the contrary notwithstanding, tolerance has been enshrined as a cardinal American virtue. Even so, Americans share a longing that in spite of diversity, Americans will stand behind their own convictions and live them out without government intrusion. Generally speaking, moral indifference—not caring about being good—is viewed more negatively than moral difference. Americans also generally share a belief that the U.S. should remain the world’s strongest military power, that its world leadership role is pivotal, and that it is a force for good in the world. They believe America should see to its own interests before helping citizens of other nations, including those who might seek to immigrate. Americans are overwhelmingly patriotic, identifying themselves much more with country than as global citizens. Flag burning and disrespect for the military are broadly frowned upon. Everyone isn’t expected to know English, but to show no interest in learning it strikes most as “un-American.”

Americans, however, also evidence systematic differences: Some consider deterioration of the natural environment, excessive material consumption, American world domination (and its attending resentments) as serious threats to the world’s future; while others, a very different group culturally, see atheism, lack of spirituality, Islam, and communism as serious threats. The former tend to worry about unchecked economic development and consumption, and hope that one day wealth will be more equally shared, while the latter hope that religious faith and spirituality will flourish, nurtured by public schools that emphasize faith, courtesy, and character-building. The former question America’s role in the world, believing we too quickly tell other nations what to do and too aggressively, even violently, pursue our own interests, while the latter see our nation as a city on a hill, a superior cultural and religious model for others to emulate. The former are offended by the very notion of moral uniformity, rejecting calls for government control over the media’s moral content; the latter are offended by anything potentially associated with Godlessness or socialism. The moral reasoning of the former tends toward, “as long as we don’t hurt others we should all just live however we want,” and “we should be more tolerant of people who adopt alternate lifestyles,” while the latter’s moral reasoning tends in this direction: “those who violate God’s rules will be punished,” and “we would all be better off if we could live by the same basic moral guidelines.”25

But the “some see”/“others see” structure of the previous paragraph raises the logical question, in each case, of “Who?” This question is not easily answered. We all breathe cultural crosswinds, finding ourselves speaking multiple cultural languages and endorsing contradictory cultural arguments. At the end of the day, the conflict is truly a cultural one within each of us and between the like-minded groups we inhabit. In fact, the narrative in the previous paragraphs about national similarities and differences is grounded in the thinking of ordinary Americans; it was derived from their answers to the 2003 Survey on Difference and Democracy. But it was also constructed by analyzing patterns of similarity and difference in how questions within entire sets of survey questions were answered, revealing systematic similarities and differences in our operative cultural frameworks.

These American differences are neither random nor ad hoc. Many of them map onto religious/secular divisions in our society. They also relate to differential views of American exceptionalism, differential understandings of what constitutes “us” versus “the other,” and differential cultural narratives regarding the past, the future, and our place in the scheme of things. Beyond that, they also map onto political preferences and voting patterns, and increasingly, as Bill Bishop and others have pointed out, onto the news one receives, the residential community one lives in, and which public events spark celebration on the one hand, and anger or resentment, on the other.

Whether the abundant empirical evidence for an enduring cultural and political polarization constitutes a “culture war,” I will leave to others to decide. The language of war may be too much of a distraction, turning our eyes from systematic and creative analysis of our national similarities and differences. Certainly, the premature conclusion that a pervasive, culturally anchored polarization does not plague us undercuts the very types of investigation that are needed at a time such as this.

Endnotes

1. James Davison Hunter, Culture Wars: The Struggle to Define America (New York: Basic, 1991).
2. Paul DiMaggio, John Evans, and Bethany Bryson, “Have Americans’ Social Attitudes Become More Polarized?” American Journal of Sociology 102.3 (1996): 690–755.
3. Alan Wolfe, One Nation After All: What Middle-Class Americans Really Think About... (New York: Penguin, 1998) 320–1.
4. See Alan I. Abramowitz, The Disappearing Center: Engaged Citizens, Polarization, & American Democracy (New Haven: Yale University Press, 2010); the expanded edition of Andrew Gelman, Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do (Princeton: Princeton University Press, 2010); Naomi Cahn and June Carbone, Red Families versus Blue Families: Legal Polarization and the Creation of Culture (New York: Oxford University Press, 2010); Irene Thomson, Culture Wars and Enduring American Dilemmas(Ann Arbor: University of Michigan Press, 2010); Bill Bishop, The Big Sort: Why the Clustering of Like-Minded America Is Tearing Us Apart (New York: Houghton Mifflin Harcourt, 2008); Sean Theriault, Party Polarization in Congress(Cambridge: Cambridge University Press, 2008); and Marc Hetherington and Jonathan Weiler, Authoritarianism and Polarization in American Politics(Cambridge: Cambridge University Press, 2009).
5. See Morris P. Fiorina, with Samuel J. Abrams and Jeremy C. Pope, Culture War? The Myth of a Polarized America, 2d ed. (New York: Pearson Longman, 2006).
6. Abramowitz 7.
7. This phenomenon is analyzed in detail in Bill Bishop’s The Big Sort.
8. Abramowitz 15–6.
9. Hetherington and Weiler 24–5.
10. Hetherington and Weiler 42.
11. Hetherington and Weiler 11.
12. Cahn and Carbone 64–5.
13. The best exposition of Lakoff’s cognitive-science based family metaphor is found in George Lakoff, Moral Politics: How Liberals and Conservatives Think (Chicago: University of Chicago Press, 1996).
14. Hetherington and Weiler 34.
15. Hetherington and Weiler 198.
16. See Hunter.
17. The 2003 Survey on Difference and Democracy was a nationally representative mail survey conducted by the University of Virginia’s Institute for Advanced Studies in Culture and fielded by the Gallup Organization.
18. Index scores were generated using Urbano Lorenzo-Seva and Pere J. Ferrando’s (2006) FACTOR software. Item weights for the nine items were assigned via a Principal Components Analysis of the polychoric correlation matrix. The eigen value for the single requested factor was 4.352. The ten Berge & Hofstee coefficient alpha estimate was .866. Polychoric correlations were computed on the original, uncollapsed items, each of which was measured with this four category, ordinal response set: Exactly (Statement) A, Mostly A, Mostly B, Exactly B.
19. All graphs were initially generated with SYSTAT version 13. Some post-editing of SYSTAT metafiles was required.
20. The survey question was, “How important to you are your religious beliefs?” The “not too important” distribution in the chart includes respondents who said their beliefs are “not at all important.”
21. Respondents were asked who they would vote for (Bush, Gore, or someone else) if the 2000 election were held again today (in 2003). An estimated 28 percent of Americans score above 0.5 on the Future Hopes Index, while an estimated 30 percent score lower than -0.5.
22. Alan Wolfe, “The Culture War that Never Came,” in James Davison Hunter and Alan Wolfe, Is There a Culture War? A Dialogue on Values and American Public Life (Washington: Brookings, 2006) 47.
23. If you are one of those quantitatively oriented types (like myself) whom this dismissal might offend, please see footnote number 24.
24. The dependent variable for all logistic regression models was the question, “If the election were being held today and the choice was again between Al Gore for the Democrats and George W. Bush for the Republicans, who would you be most likely to vote for? The three response categories—Gore, Bush, “Someone else”—were collapsed to two, so that the logistic regression would predict simply whether respondents chose Bush or not. Predictors in the final model (presented in the chart) were: the Future Hopes Index, the Approval of Homosexuality Index, an ordinal measure of household income, and political ideology (measured as response to the question, “Which word best describes your overall political beliefs: conservative, moderate, or liberal?”). Political ideology was entered as two dummy variables: conservative and moderate. Models were estimated using Stata. Scott Long’s SPOST routines were used to generate predicted probabilities. Other variables such as education, region of the country, urbanism of residence, and gender were included in other models, but generally lost statistical significance once the predictors from the final model were included.
25. The narrative in this paragraph and the one before is based upon Multidimensional Scalings of Euclidean Distance measures representing differences in reponse to questions for several subset items included in theSurvey on Difference and Democracy.

Carl Desportes Bowman is Director of Survey Research at the Institute for Advanced Studies in Culture. His work in cultural analysis alternates between contemporary examination of American public culture and historical analysis of the Brethren, most notably in his book Brethren Society: The Cultural Transformation of a Peculiar People.