Continental Divide

Among elected officials, journalists, and average citizens, intensifying partisan polarization is thought to be one of the dominant political trends of our times. Yet it has proved remarkably controversial among political scientists.

Some dimensions of polarization are not in question. For the first time in modern history (and maybe ever), there is no ideological overlap between the two major parties. According to the standard political-science index, the most conservative Democratic senator is a bit more liberal than is the most liberal Republican senator. Even more remarkably, the same is true in the House of Representatives. If the core of what we mean by the “center” of American politics is the overlap between the parties, then in a very real sense the center has disappeared.

At the same time, party unity, discipline, and centralization are stronger than they were a generation ago. Back then, legislative bipartisanship meant starting from the center and building out. Today, it usually means that one party tries to pick off a handful of the other party’s members. President Barack Obama came into office pledging to change this. While he has managed to enact some of his major policy initiatives, he has failed to bridge the partisan divide in Congress. Historians will debate whether he could have succeeded with a different strategy. But it’s hard to deny that he began with the deck stacked against him.

The core of the disagreement about polarization concerns not the state of affairs in Washington, but rather the relationship between party elites and the electorate. In 2006, Morris P. Fiorina and two colleagues published what soon became a well-known book, Culture War? The Myth of a Polarized America. The title accurately foreshadows their argument: Increases in polarization are principally confined to political elites. The American people are mostly moderate in their orientation and are not much more divided than they were a generation ago. Yes, voters have sorted themselves out between the parties along ideological lines. But that doesn’t mean they’ve moved left and right at the expense of the center. Rather, they are responding to the signals that elites are sending from the top. If the major parties sent more moderate signals, voters would respond to those as well, many with relief.

The volume under review, The Disappearing Center, represents a frontal challenge to the Fiorina thesis. With a wealth of data, most of it drawn from the gold standard American National Election Studies, Alan Abramowitz offers a powerful argument that polarization has taken deep root in the electorate. Compared with two decades ago, many fewer Americans are located at or near the center of the ideological spectrum, many more are at or near the extremes. Voters are far more polarized than nonvoters; voters who are deeply engaged in politics—who follow the news, possess relatively high levels of political information, and care a lot about electoral outcomes—are more polarized than are less engaged voters. And contrary to what one might intuitively expect, more educated voters are more polarized than are the less educated. Indeed, the rising level of formal education over the past generation seems to be among the principal causes of increased polarization.

As a consequence of these trends, Abramowitz argues, we now have a two-party system in which ideology, partisan preference, and voting are much more closely linked than they were thirty or even twenty years ago. Today, most liberals are Democrats, and most conservatives are Republicans; Democrats overwhelmingly vote for Democratic candidates, Republicans for Republicans. Not surprisingly, there is far more alignment between presidential and congressional voting than there once was. The number of House districts where the presidential and congressional majorities are divided between the parties has declined, as has the number of Senate seats in which the senatorial winner is from one party and the presidential winner from the other.

Another important consequence of polarization is increasing divisions along geographical lines. While many observers believe that the rising number of safe House seats reflects more skillful gerrymandering, political scientists (of whom Abramowitz is the latest) have found that artful line-drawing accounts for no more than one-third of the increase. Much more stems from the demographic trend Bill Bishop has called the “big sort”: simply put, more and more people now seek to live near others who think the way they do. In jurisdictions such as counties (whose boundaries don’t change much) and states (where they don’t change at all), the number of supermajority victories has risen sharply. This is the main reason close national elections today feature many fewer contested states than did comparable elections in the 1960s and ’70s.

This is not to say that the trends of recent decades have affected the political parties in exactly the same way. Instead, we now have what I call asymmetrical polarization. While nearly three-quarters of Republican voters are conservative, and the rest moderate, Democrats are far more diverse—about 40 percent liberal, 40 percent moderate, and the remaining 20 percent conservative. Democrats remain an ideological coalition, both in the electorate and, to a lesser but still significant extent, in the Congress. As a result, it is relatively easy for Republican leaders to hold their troops together; Democratic leaders must negotiate. (The tortured path to a health-care-reform majority is Exhibit A.)

I am not the right reviewer to offer a fine-grained methodological assessment of the way Abramowitz uses the data. I don’t think I will be alone, however, in concluding that he makes a powerful prima-facie case for his conclusion. In the wake of his book, the burden of proof has clearly shifted to those who would continue to affirm that polarization is confined to political elites.

A more imponderable question is what to make of all this. The political scientists who sixty years ago called for a more “responsible” party system offering clearer choices and a tighter link between electoral outcomes and policy results would have no choice but to applaud the developments of recent decades. Abramowitz offers a mixed verdict: while polarization has helped increase the number of engaged citizens, the movement toward a quasi-parliamentary party system comports imperfectly with our decidedly nonparliamentary constitutional system. In the absence of hard-to-achieve compromise, or hard-to-sustain supermajorities, the inevitable outcome of polarization is gridlock. The more challenging the issue, the more likely is that result.

My own verdict is more negative. While it is true that today’s party system offers the voters clearer choices (which they understand better) and a closer connection between those choices and governing agendas, polarization has degraded our politics in many ways. It is connected to the collapse of shared assumptions or even a common body of agreed-upon facts, without which deliberation is impossible. It has contributed to a zero-sum mentality that affects not only the legislative process but governance itself: if nominations are routinely caught in the partisan crossfire, staffing the government becomes more difficult than it should be, and the judiciary becomes dangerously politicized. By forcing majorities to rely on their own ranks to pass legislation, polarization gives excessive power to individuals and small factions while reducing the majority’s incentive to accept the minority’s good ideas. It spurs the decline of civility and exacerbates the tendency to regard the other party as beyond the pale and its victories as somehow illegitimate. Polarization makes it impossible to solve problems—such as our ever more dangerous fiscal crisis—that can only be addressed if the parties agree to share responsibility for necessary but unpopular measures.

Perhaps worst of all, the unending partisan contestation of recent decades has fed Americans’ mounting distrust and contempt for their governing institutions. To be sure, a measure of mistrust is a healthy check on governmental threats to liberty. But as James Madison warned, an excess of mistrust makes effective self-government all but impossible. That is the point we have now reached.