How Divided Are We?
The 2004 election left our country deeply divided over whether our country is deeply divided. For some, America is indeed a polarized nation, perhaps more so today than at any time in living memory. In this view, yesterday's split over Bill Clinton has given way to today's even more acrimonious split between Americans who detest George Bush and Americans who detest John Kerry, and similar divisions will persist as long as angry liberals and angry conservatives continue to confront each other across the political abyss. Others, however, believe that most Americans are moderate centrists, who, although disagreeing over partisan issues in 2004, harbor no deep ideological hostility. I take the former view.
By polarization I do not have in mind partisan disagreements alone. These have always been with us. Since popular voting began in the 19th century, scarcely any winning candidate has received more than 60 percent of the vote, and very few losers have received less than 40 percent. Inevitably, Americans will differ over who should be in the White House. But this does not necessarily mean they are polarized.
By polarization I mean something else: an intense commitment to a candidate, a culture, or an ideology that sets people in one group definitively apart from people in another, rival group. Such a condition is revealed when a candidate for public office is regarded by a competitor and his supporters not simply as wrong but as corrupt or wicked; when one way of thinking about the world is assumed to be morally superior to any other way; when one set of political beliefs is considered to be entirely correct and a rival set wholly wrong. In extreme form, as defined by Richard Hofstadter in The Paranoid Style in American Politics (1965), polarization can entail the belief that the other side is in thrall to a secret conspiracy that is using devious means to obtain control over society. Today's versions might go like this: “Liberals employ their dominance of the media, the universities, and Hollywood to enforce a radically secular agenda”; or, “conservatives, working through the religious Right and the big corporations, conspired with their hired neocon advisers to invade Iraq for the sake of oil.”
Polarization is not new to this country. It is hard to imagine a society more divided than ours was in 1800, when pro-British, pro-commerce New Englanders supported John Adams for the presidency while pro-French, pro-agriculture Southerners backed Thomas Jefferson. One sign of this hostility was the passage of the Alien and Sedition Acts in 1798; another was that in 1800, just as in 2000, an extremely close election was settled by a struggle in one state (New York in 1800, Florida in 2000).
The fierce contest between Abraham Lincoln and George McClellan in 1864 signaled another national division, this one over the conduct of the Civil War. But thereafter, until recently, the nation ceased to be polarized in that sense. Even in the half-century from 1948 to (roughly) 1996, marked as it was by sometimes strong expressions of feeling over whether the presidency should go to Harry Truman or Thomas Dewey, to Dwight Eisenhower or Adlai Stevenson, to John F. Kennedy or Richard Nixon, to Nixon or Hubert Humphrey, and so forth, opinion surveys do not indicate widespread detestation of one candidate or the other, or of the people who supported him.
Now they do. Today, many Americans and much of the press regularly speak of the President as a dimwit, a charlatan, or a knave. A former Democratic presidential candidate has asserted that Bush “betrayed” America by launching a war designed to benefit his friends and corporate backers. A senior Democratic Senator has characterized administration policy as a series of “lies, lies, and more lies” and has accused Bush of plotting a “mindless, needless, senseless, and reckless” war. From the other direction, similar expressions of popular disdain have been directed at Senator John Kerry (and before him at President Bill Clinton); if you have not heard them, that may be because (unlike many of my relatives) you do not live in Arkansas or Texas or other locales where the New York Times is not read. In these places, Kerry is widely spoken of as a scoundrel.
In the 2004 presidential election, over two-thirds of Kerry voters said they were motivated explicitly by the desire to defeat Bush. By early 2005, President Bush's approval rating, which stood at 94 percent among Republicans, was only 18 percent among Democrats—the largest such gap in the history of the Gallup poll. These data, moreover, were said to reflect a mutual revulsion between whole geographical sections of the country, the so-called Red (Republican) states versus the so-called Blue (Democratic) states. As summed up by the distinguished social scientist who writes humor columns under the name of Dave Barry, residents of Red states are “ignorant racist fascist knuckle-dragging NASCAR-obsessed cousin-marrying road-kill-eating tobacco-juice-dribbling gun-fondling religious fanatic rednecks,” while Blue-state residents are “godless unpatriotic pierced-nose Volvo-driving France-loving leftwing Communist latte-sucking tofu-chomping holistic-wacko neurotic vegan weenie perverts.”
To be sure, other scholars differ with Dr. Barry. To them, polarization, although a real enough phenomenon, is almost entirely confined to a small number of political elites and members of Congress. In Culture War? (2004), which bears the subtitle “The Myth of a Polarized America,” Morris Fiorina of Stanford argues that policy differences between voters in Red and Blue states are really quite small, and that most are in general agreement even on issues like abortion and homosexuality.
But the extent of polarization cannot properly be measured by the voting results in Red and Blue states. Many of these states are in fact deeply divided internally between liberal and conservative areas, and gave the nod to one candidate or the other by only a narrow margin. Inferring the views of individual citizens from the gross results of presidential balloting is a questionable procedure.
Nor does Fiorina's analysis capture the very real and very deep division over an issue like abortion. Between 1973, when Roe v. Wade was decided, and now, he writes, there has been no change in the degree to which people will or will not accept any one of six reasons to justify an abortion: (1) the woman's health is endangered; (2) she became pregnant because of a rape; (3) there is a strong chance of a fetal defect; (4) the family has a low income; (5) the woman is not married; (6) and the woman simply wants no more children. Fiorina may be right about that. Nevertheless, only about 40 percent of all Americans will support abortion for any of the last three reasons in his series, while over 80 percent will support it for one or another of the first three.
In other words, almost all Americans are for abortion in the case of maternal emergency, but fewer than half if it is simply a matter of the mother's preference. That split—a profoundly important one—has remained in place for over three decades, and it affects how people vote. In 2000 and again in 2004, 70 percent of those who thought abortion should always be legal voted for Al Gore or John Kerry, while over 70 percent of those who thought it should always be illegal voted for George Bush.
Division is just as great over other high-profile issues. Polarization over the war in Iraq, for example, is more pronounced than any war-related controversy in at least a half-century. In the fall of 2005, according to Gallup, 81 percent of Democrats but only 20 percent of Republicans thought the war in Iraq was a mistake. During the Vietnam war, by contrast, itself a famously contentious cause, there was more unanimity across party lines, whether for or against: in late 1968 and early 1969, about equal numbers of Democrats and Republicans thought the intervention there was a mistake. Pretty much the same was true of Korea: in early 1951, 44 percent of Democrats and 61 percent of Republicans thought the war was a mistake—a partisan split, but nowhere near as large as the one over our present campaign in Iraq.
Polarization, then, is real. But what explains its growth? And has it spread beyond the political elites to influence the opinions and attitudes of ordinary Americans?
The answer to the first question, I suspect, can be found in the changing politics of Congress, the new competitiveness of the mass media, and the rise of new interest groups.
That Congress is polarized seems beyond question. When, in 1998, the House deliberated whether to impeach President Clinton, all but four Republican members voted for at least one of the impeachment articles, while only five Democrats voted for even one. In the Senate, 91 percent of Republicans voted to convict on at least one article; every single Democrat voted for acquittal.
The impeachment issue was not an isolated case. In 1993, President Clinton's budget passed both the House and the Senate without a single Republican vote in favor. The same deep partisan split occurred over taxes and supplemental appropriations. Nor was this a blip: since 1950, there has been a steady increase in the percentage of votes in Congress pitting most Democrats against most Republicans.
In the midst of the struggle to pacify Iraq, Howard Dean, the chairman of the Democratic National Committee, said the war could not be won and Nancy Pelosi, the leader of the House Democrats, endorsed the view that American forces should be brought home as soon as possible. By contrast, although there was congressional grumbling (mostly by Republicans) about Korea and complaints (mostly by Democrats) about Vietnam, and although Senator George Aiken of Vermont famously proposed that we declare victory and withdraw, I cannot remember party leaders calling for unconditional surrender.
The reasons for the widening fissures in Congress are not far to seek. Each of the political parties was once a coalition of dissimilar forces: liberal Northern Democrats and conservative Southern Democrats, liberal coastal Republicans and conservative Midwestern Republicans. No longer; the realignments of the South (now overwhelmingly Republican) and of New England (now strongly Democratic) have all but eliminated legislators who deviate from the party's leadership. Conservative Democrats and liberal Republicans are endangered species now approaching extinction. At the same time, the ideological gap between the parties is growing: if there was once a large overlap between Democrats and Republicans—remember “Tweedledum and Tweedledee”?—today that congruence has almost disappeared. By the late 1990's, virtually every Democrat was more liberal than virtually every Republican.
The result has been not only intense partisanship but a sharp rise in congressional incivility. In 1995, a Republican-controlled Senate passed a budget that President Clinton proceeded to veto; in the loggerhead that followed, many federal agencies shut down (in a move that backfired on the Republicans). Congressional debates have seen an increase not only in heated exchanges but in the number of times a representative's words are either ruled out of order or “taken down” (that is, written by the clerk and then read aloud, with the offending member being asked if he or she wishes to withdraw them).
It has been suggested that congressional polarization is exacerbated by new districting arrangements that make each House seat safe for either a Democratic or a Republican incumbent. If only these seats were truly competitive, it is said, more centrist legislators would be elected. That seems plausible, but David C. King of Harvard has shown that it is wrong: in the House, the more competitive the district, the more extreme the views of the winner. This odd finding is apparently the consequence of a nomination process dominated by party activists. In primary races, where turnout is low (and seems to be getting lower), the ideologically motivated tend to exercise a preponderance of influence.
All this suggests a situation very unlike the half-century before the 1990's, if perhaps closer to certain periods in the 18th and 19th centuries. Then, too, incivility was common in Congress, with members not only passing the most scandalous remarks about each other but on occasion striking their rivals with canes or fists. Such partisan feeling ran highest when Congress was deeply divided over slavery before the Civil War and over Reconstruction after it. Today the issues are different, but the emotions are not dissimilar.
Next, the mass media. Not only are they themselves increasingly polarized, but consumers are well aware of it and act on that awareness. Fewer people now subscribe to newspapers or watch the network evening news. Although some of this decline may be explained by a preference for entertainment over news, some undoubtedly reflects the growing conviction that the mainstream press generally does not tell the truth, or at least not the whole truth.
In part, media bias feeds into, and off, an increase in business competition. In the 1950's, television news amounted to a brief 30-minute interlude in the day's programming, and not a very profitable one at that; for the rest of the time, the three networks supplied us with westerns and situation comedies. Today, television news is a vast, growing, and very profitable venture by the many broadcast and cable outlets that supply news twenty-four hours a day, seven days a week.
The news we get is not only more omnipresent, it is also more competitive and hence often more adversarial. When there were only three television networks, and radio stations were forbidden by the fairness doctrine from broadcasting controversial views, the media gravitated toward the middle of the ideological spectrum, where the large markets could be found. But now that technology has created cable news and the Internet, and now that the fairness doctrine has by and large been repealed, many media outlets find their markets at the ideological extremes.
Here is where the sharper antagonism among political leaders and their advisers and associates comes in. As one journalist has remarked about the change in his profession, “We don't deal in facts [any longer], but in attributed opinions.” Or, these days, in unattributed opinions. And those opinions are more intensely rivalrous than was once the case.
The result is that, through commercial as well as ideological self-interest, the media contribute heavily to polarization. Broadcasters are eager for stories to fill their round-the-clock schedules, and at the same time reluctant to trust the government as a source for those stories. Many media outlets are clearly liberal in their orientation; with the arrival of Fox News and the growth of talk radio, many are now just as clearly conservative.
The evidence of liberal bias in the mainstream media is very strong. The Center for Media and Public Affairs (CMPA) has been systematically studying television broadcasts for a quarter-century. In the 2004 presidential campaign, John Kerry received more favorable mentions than any presidential candidate in CMPA's history, especially during the month before election day. This is not new: since 1980 (and setting aside the recent advent of Fox News), the Democratic candidate has received more favorable mentions than the Republican candidate in every race except the 1988 contest between Michael Dukakis and George H. W. Bush. A similarly clear orientation characterizes weekly newsmagazines like Time and Newsweek.
For its part, talk radio is listened to by about one-sixth of the adult public, and that one-sixth is made up mostly of conservatives.
1 National Public Radio has an audience of about the same size; it is disproportionately liberal. The same breakdown affects cable-television news, where the rivalry is between CNN (and MSNBC) and Fox News. Those who watch CNN are more likely to be Democrats than Republicans; the reverse is emphatically true of Fox. As for news and opinion on the Internet, which has become an important source for college graduates in particular, it, too, is largely polarized along political and ideological lines, emphasized even more by the culture that has grown up around news blogs.
At one time, our culture was only weakly affected by the media because news organizations had only a few points of access to us and were largely moderate and audience-maximizing enterprises. Today the media have many lines of access, and reflect both the maximization of controversy and the cultivation of niche markets. Once the media talked to us; now they shout at us.
And then there are the interest groups. In the past, the major ones—the National Association of Manufacturers, the Chamber of Commerce, and labor organizations like the AFL-CIO—were concerned with their own material interests. They are still active, but the loudest messages today come from very different sources and have a very different cast to them. They are issued by groups concerned with social and cultural matters like civil rights, managing the environment, alternatives to the public schools, the role of women, access to firearms, and so forth, and they directly influence the way people view politics.
Interest groups preoccupied with material concerns can readily find ways to arrive at compromise solutions to their differences; interest groups divided by issues of rights or morality find compromise very difficult. The positions taken by many of these groups and their supporters, often operating within the two political parties, profoundly affect the selection of candidates for office. In brief, it is hard to imagine someone opposed to abortion receiving the Democratic nomination for President, or someone in favor of it receiving the Republican nomination.
Outside the realm of party politics, interest groups also file briefs in important court cases and can benefit from decisions that in turn help shape the political debate. Abortion became a hot controversy in the 1970's not because the American people were already polarized on the matter but because their (mainly centrist) views were not consulted; instead, national policy was determined by the Supreme Court in a decision, Roe v. Wade, that itself reflected a definition of “rights” vigorously promoted by certain well-defined interest groups.
Polarization not only is real and has increased, but it has also spread to rank-and-file voters through elite influence.
In The Nature and Origins of Mass Opinion (1992), John R. Zaller of UCLA listed a number of contemporary issues—homosexuality, a nuclear freeze, the war in Vietnam, busing for school integration, the 1990-91 war to expel Iraq from Kuwait—and measured the views held about them by politically aware citizens. (By “politically aware,” Zaller meant people who did well answering neutral factual questions about politics.) His findings were illuminating.
Take the Persian Gulf war. Iraq had invaded Kuwait in August 1990. From that point through the congressional elections in November 1990, scarcely any elite voices were raised to warn against anything the United States might contemplate doing in response. Two days after the mid-term elections, however, President George H. W. Bush announced that he was sending many more troops to the Persian Gulf. This provoked strong criticism from some members of Congress, especially Democrats.
As it happens, a major public-opinion survey was under way just as these events were unfolding. Before criticism began to be voiced in Congress, both registered Democrats and registered Republicans had supported Bush's vaguely announced intention of coming to the aid of Kuwait; the more politically aware they were, the greater their support. After the onset of elite criticism, the support of Republican voters went up, but Democratic support flattened out. As Bush became more vigorous in enunciating his aims, politically aware voters began to differ sharply, with Democratic support declining and Republican support increasing further.
Much the same pattern can be seen in popular attitudes toward the other issues studied by Zaller. As political awareness increases, attitudes split apart, with, for example, highly aware liberals favoring busing and job guarantees and opposing the war in Vietnam, and highly aware conservatives opposing busing and job guarantees and supporting the war in Vietnam.
But why should this be surprising? To imagine that extremist politics has been confined to the chattering classes is to believe that Congress, the media, and American interest groups operate in an ideological vacuum. I find that assumption implausible.
As for the extent to which these extremist views have spread, that is probably best assessed by looking not at specific issues but at enduring political values and party preferences. In 2004, only 12 percent of Democrats approved of George Bush; at earlier periods, by contrast, three to four times as many Democrats approved of Ronald Reagan, Gerald Ford, Richard Nixon, and Dwight D. Eisenhower. Over the course of about two decades, in other words, party affiliation had come to exercise a critical influence over what people thought about a sitting President.
The same change can be seen in the public's view of military power. Since the late 1980's, Republicans have been more willing than Democrats to say that “the best way to ensure peace is through military strength.” By the late 1990's and on into 2003, well over two-thirds of all Republicans agreed with this view, but far fewer than half of all Democrats did. In 2005, three-fourths of all Democrats but fewer than a third of all Republicans told pollsters that good diplomacy was the best way to ensure peace. In the same survey, two-thirds of all Republicans but only one fourth of all Democrats said they would fight for this country “whether it is right or wrong.”
Unlike in earlier years, the parties are no longer seen as Tweedledum and Tweedledee. To the contrary, as they sharpen their ideological differences, attentive voters have sharpened their ideological differences. They now like either the Democrats or the Republicans more than they once did, and are less apt to feel neutral toward either one.
How deep does this polarization reach? As measured by opinion polls, the gap between Democrats and Republicans was twice as great in 2004 as in 1972. In fact, rank-and-file Americans disagree more strongly today than did politically active Americans in 1972.
To be sure, this mass polarization involves only a minority of all voters, but the minority is sizable, and a significant part of it is made up of the college-educated. As Marc Hetherington of Vanderbilt puts it: “people with the greatest ability to assimilate new information, those with more formal education, are most affected by elite polarization.” And that cohort has undeniably grown.
In 1900, only 10 percent of all young Americans went to high school. My father, in common with many men his age in the early 20th century, dropped out of school after the eighth grade. Even when I graduated from college, the first in my family to do so, fewer than one-tenth of all Americans over the age of twenty-five had gone that far. Today, 84 percent of adult Americans have graduated from high school and nearly 27 percent have graduated from college. This extraordinary growth in schooling has produced an ever larger audience for political agitation.
Ideologically, an even greater dividing line than undergraduate education is postgraduate education. People who have proceeded beyond college seem to be very different from those who stop with a high-school or college diploma. Thus, about a sixth of all voters describe themselves as liberals, but the figure for those with a postgraduate degree is well over a quarter. In mid-2004, about half of all voters trusted George Bush; less than a third of those with a postgraduate education did. In November of the same year, when over half of all college graduates voted for Bush, well over half of the smaller cohort who had done postgraduate work voted for Kerry. According to the Pew Center for Research on the People and the Press, more than half of all Democrats with a postgraduate education supported the antiwar candidacy of Howard Dean.
The effect of postgraduate education is reinforced by being in a profession. Between 1900 and 1960, write John B. Judis and Ruy Teixeira in The Emerging Democratic Majority (2002), professionals voted pretty much the same way as business managers; by 1988, the former began supporting Democrats while the latter supported Republicans. On the other hand, the effect of postgraduate education seems to outweigh the effect of affluence. For most voters, including college graduates, having higher incomes means becoming more conservative; not so for those with a postgraduate education, whose liberal predilections are immune to the wealth effect.
The results of this linkage between ideology, on the one hand, and congressional polarization, media influence, interest-group demands, and education on the other are easily read in the commentary surrounding the 2004 election. In their zeal to denigrate the President, liberals, pronounced one conservative pundit, had “gone quite around the twist.” According to liberal spokesmen, conservatives with their “religious intolerance” and their determination to rewrite the Constitution had so befuddled their fellow Americans that a “great nation was felled by a poisonous nut.”
If such wholesale slurs are not signs of polarization, then the word has no meaning. To a degree that we cannot precisely measure, and over issues that we cannot exactly list, polarization has seeped down into the public, where it has assumed the form of a culture war. The sociologist James Davison Hunter, who has written about this phenomenon in a mainly religious context, defines culture war as “political and social hostility rooted in different systems of moral understanding.” Such conflicts, he writes, which can involve “fundamental ideas about who we are as Americans,” are waged both across the religious/secular divide and within religions themselves, where those with an “orthodox” view of moral authority square off against those with a “progressive” view.
To some degree, this terminology is appropriate to today's political situation as well. We are indeed in a culture war in Hunter's sense, though I believe this war is itself but another component, or another symptom, of the larger ideological polarization that has us in its grip. Conservative thinking on political issues has religious roots, but it also has roots that are fully as secular as anything on the Left. By the same token, the liberal attack on conservatives derives in part from an explicitly “progressive” religious orientation—liberal Protestantism or Catholicism, or Reform Judaism—but in part from the same secular sources shared by many conservatives.
But what, one might ask, is wrong with having well-defined parties arguing vigorously about the issues that matter? Is it possible that polarized politics is a good thing, encouraging sharp debate and clear positions? Perhaps that is true on those issues where reasonable compromises can be devised. But there are two limits to such an arrangement.
First, many Americans believe that unbridgeable political differences have prevented leaders from addressing the problems they were elected to address. As a result, distrust of government mounts, leading to an alienation from politics altogether. The steep decline in popular approval of our national officials has many causes, but surely one of them is that ordinary voters agree among themselves more than political elites agree with each other—and the elites are far more numerous than they once were.
In the 1950's, a committee of the American Political Science Association (APSA) argued the case for a “responsible” two-party system. The model the APSA had in mind was the more ideological and therefore more “coherent” party system of Great Britain. At the time, scarcely anyone thought our parties could be transformed in such a supposedly salutary direction. Instead, as Governor George Wallace of Alabama put it in his failed third-party bid for the presidency, there was not a “dime's worth of difference” between Democrats and Republicans.
What Wallace forgot was that, however alike the parties were, the public liked them that way. A half-century ago, Tweedledum and Tweedledee enjoyed the support of the American people; the more different they have become, the greater has been the drop in popular confidence in both them and the federal government.
A final drawback of polarization is more profound. Sharpened debate is arguably helpful with respect to domestic issues, but not for the management of important foreign and military matters. The United States, an unrivaled superpower with unparalleled responsibilities for protecting the peace and defeating terrorists, is now forced to discharge those duties with its own political house in disarray.
We fought World War II as a united nation, even against two enemies (Germany and Italy) that had not attacked us. We began the wars in Korea and Vietnam with some degree of unity, too, although it was eventually whittled away. By the early 1990's, when we expelled Iraq from Kuwait, we had to do so over the objections of congressional critics; the first President Bush avoided putting the issue to Congress altogether. In 2003 we toppled Saddam Hussein in the face of catcalls from many domestic leaders and opinion-makers. Now, in stabilizing Iraq and helping that country create a new free government, we have proceeded despite intense and mounting criticism, much of it voiced by politicians who before the war agreed that Saddam Hussein was an evil menace in possession of weapons of mass destruction and that we had to remove him.
Denmark or Luxembourg can afford to exhibit domestic anguish and uncertainty over military policy; the United States cannot. A divided America encourages our enemies, disheartens our allies, and saps our resolve—potentially to fatal effect. What General Giap of North Vietnam once said of us is even truer today: America cannot be defeated on the battlefield, but it can be defeated at home. Polarization is a force that can defeat us.
1 The political disposition of most radio talk-show hosts is explained by William G. Mayer in “Why Talk Radio Is Conservative,” Public Interest, Summer 2004.
2 True, the “elite effect” may not be felt across the board. With most of the issues Zaller investigated, even well-informed citizens would have had little first-hand experience, and so their minds were of necessity open to the influence of their “betters.” Results might have been different had he measured their views on matters about which most Americans believe themselves to be personally well-informed: crime, inflation, drug abuse, or their local schools.