As the major political parties convene this summer, with all the usual noise, pomp, and expense, Americans can be counted on to let out a collective yawn, or maybe a grimace. But not so for political scientists. Academic experts see a lot to like--or at least a lot to study--in the American two-party system. In their considered view, a competitive party system ensures the legitimacy of opposition to government, promotes public debate about policy options, and gets citizens involved in the public sphere. The two-party system never does these things perfectly, but it does them well enough. Without it our system would collapse overnight, leaving gridlock and hyperpluralism--or so most political scientists think.
But if one looks closely at the views of those who are researching, thinking about, and writing about political parties, one finds an interesting division of opinion. One school of thought is that parties are in decline and, consequently, that we have a major problem. The public is right to be irritated. A second view holds that parties have changed dramatically but that they are just as strong as they used to be. The public ought to get used to the transformation and stop griping. A third school, best articulated by David Mayhew of Yale University, is that political scientists have attributed too much importance to party dynamics. They matter, but less so than the professional literature has suggested. In this light, the public's gripes are beside the point.
To make sense of the disagreement, we must first sift through the ruins of realignment theory. For about 20 years, American government students were instructed in this line of thinking. Its concepts still echo in political punditry. But the theory died a decade ago when it became clear it wasn't explaining with any precision the events of the actual political world.
Still, it was an elegant idea. Realignment theory held that not all elections were the same. In certain highly charged elections or in a string of two or three such elections, big and lasting shifts occurred in how voters behaved. A new voter coalition would assert control over our system, determining policy outcomes for a generation. Walter Dean Burnham, the theory's best-known proponent, suggested such elections might be a uniquely American surrogate for political revolution. Before realignment, there might be a third-party challenge, protests, and even civil disorder. Eventually, ambitious politicians would pick up the pressing issues and make them their own.
With the image of periodic political renewal, there was a soothing message in all this. Realignments allowed the political system to adjust to social and political stress, and to bring those citizens who might otherwise be absorbed in personal concerns into political action. The party system periodically restored its own vitality and that of the system as a whole.
Burnham explicitly warned, however, that the party system's capacity for "peaceful revolution" was not automatic. If and when the party system lost its ability to adapt, the branches of government would lock up. Governmental remedy as both an ideal and a practice would wither. Gradually, a propensity toward broad-based oligarchy would set in. After all, the wealthy are best protected by government that is deadlocked. Simultaneously, a huge class of unmobilized people would emerge as a "party of nonvoters." Their influence on the system would necessarily be weaker.
S cholars in the "party decline" school have inherited Burn-ham's worry. They agree that as a party system weakens it tends to pull the rest of the order down with it. Sidney Milkis of the University of Virginia, who is close in spirit to Burnham, makes such a case in Political Parties and Constitutional Government, a study of the rise and development of political parties since the founding. While Milkis does not share Burnham's open distaste for markets, capitalism, and social inequalities, he does adopt Burnham's democratic nationalism. For Milkis the weakening of parties has promoted broad discontent with American government and has generated an anemic civic culture.
In Milkis's account--and this is what makes his work so provocative--the cause of party decline, and thus of public cynicism, is not the depoliticizing force of market values, as Burnham has long argued. Instead, it is the particular development of the presidency. To put Milkis's claim bluntly, FDR killed the parties when he built a government competent enough to run a welfare state. In doing that, he changed the constitutional balance that had been supported by the parties since the time of Madison, Jackson, and Van Buren. Subsequent presidents failed to reverse Roosevelt's legacy.
Party competition first emerged, Milkis argues, when James Madison and Thomas Jefferson sought to develop a political opposition to Alexander Hamilton, John Adams, and the Federalist legatees of George Washington's two-term presidency. Madison, in particular, feared for the future of federalism and the separation of powers if Hamilton's economic nationalism were left politically unchecked. Milkis takes pains to point out that there was a second Madison, one less well-known but just as important as the more familiar Madison who framed the Constitution. The first disliked parties and factions; the second had no trouble embracing them in order to save his overall institutional design. Happily for Madison, the party system that he helped to launch evolved (thanks to the genius of Martin Van Buren) into a stable contest between two large confederations of state and local parties. And happily for the system as a whole, Milkis says, the parties won the political loyalties of voters scattered across a vast geographic expanse.
Americans came to appreciate the full range of national, state, and local institutions contemplated by the founders. Voters liked their town and county governments; they valued their state institutions; and they came to treasure not only the presidency but the Senate, the House, and the Supreme Court. America's elaborate mix of national, state, and local jurisdictions and offices might never have taken hold without the early development of decentralized but nationally competitive parties. This accomplishment helps to explain the persistence of the Constitution of 1787 despite the extraordinary events of the Civil War and the Reconstruction, and the huge expansion of the republic's size.
But our party system and institutions were never particularly well-suited for strong, positive government, Milkis argues. They were good for participation and office-seeking, but not for supple macroeconomic management or the competent bureaucratic delivery of social benefits, such as old-age income security or work relief. Here Milkis carries forward a long line of thinking about party politics and public administration that dates to the work of Herbert Croly and other progressives in the early decades of the 1900s.
FDR was the first president, in Milkis's view, who was forced to cope with the lack of fit between the institutional forms given to him and new executive tasks. He keenly understood the limits of the party system he inherited, and sought briefly to do something about them, through the ill-fated 1938 "Roosevelt purge" in which New Deal liberals were encouraged to run against reactionary and conservative Democratic incumbents in Congress. He hoped to transform his party into a programmatic, responsible organization. He failed miserably.
FDR did not try again, opting instead for the independent regulatory commissions, new bureaucracies, court-packing, and executive reorganization that he or congressional liberals had already launched or planned before the purge effort. Roosevelt grasped that he could, and probably should, soft-pedal his party as an instrument of executive governance. It was too loaded with southern conservatives and stand-pat careerists. Time was short, and there was much work to be done to save liberal capitalism from its enemies within and without.
But there was a hidden price for this understandable decision. The cost to the polity, one that was not immediately obvious, was reduced voter involvement. As Kennedy, Johnson, and Nixon perfected the New Deal state, they did so on the backs of social movements, professors, experts, and government executives and lawyers. Their mission was not to revitalize the remnants of the urban machines or to reform the conservative state parties and party factions that they scorned. They made the same choice Roosevelt did. So the decentralized system of confederated parties--imagined by Madison and perfected by Van Buren--collapsed, as one ward club or county committee after another (with the notable exception of Chicago) died on the liberal vine. These local institutions were the vital foundation of voter involvement; without them voter turnout began its long decline.
Not all political scientists are alarmed by such developments. In his important 1995 work, Why Parties?, John H. Aldrich responds to the passing of the ward heelers by saying, in effect, "so what?" He wants us to face up to a stark proposition: The forms of parties are going to change. As he notes, trenchantly, "The major political party is the creature of the politicians... . These politicians do not have partisan goals per se, and the party is only the instrument for achieving them." Politicians run the parties, and they will inevitably change the ways in which parties help them to be politicians.
Aldrich is no iconoclast, to be sure. His book is deeply thoughtful, gently argued, and quite rigorous. At the heart of Aldrich's case lies an extended comparison between two party systems: the system that emerged in the North during the 1820s and 1830s and that lasted until the Civil War, and the more familiar two-party system that has structured our politics since the 1960s. The first was intensely mobilizing and generated sharp increases in voter turnout until it reached extraordinary, indeed uniquely high levels. This was also a period of "team parties," in which politicians subordinated their individual identities to the corporate identity of their party since the path to power lay through making that trade-off.
Today's parties, in contrast, are service-providing organizations. They resemble a franchise for entrepreneurs. The individual candidates of the two parties meet certain programmatic requirements related to party ideology, but in terms of campaigning they act as free-lancers. They have no trouble behaving as highly competitive teams within government, particularly within the House of Representatives, but they do not cooperate with each other to rally voters. Stimulating turnout is up to an individual candidate if he or she chooses.
The point of Aldrich's contrast is not that there has been decline relative to some golden age. Instead, these are fundamentally different systems. Juxtaposed to this claim is a lucid demonstration of the central tendency of any competitive party system, regardless of differences in the campaign styles of politicians. Using simple modeling, Aldrich posits that a party system will solve pathologies that would otherwise plague politicians. Without a competitive party system, politicians could not cooperate around mutual policy gains, which can only come through repeated interaction and binding commitments that hold up across time. They would instead treat all their interactions with each other as one-shot games and thus fall prey to the noncooperative trap epitomized by the "prisoner's dilemma."
Second, without parties' resources and their capacity to stimulate, motivate, and inform voters, politicians could never solve a major dilemma facing voters, i.e., the propensity to avoid voting and to instead "free ride" on those who take the trouble to vote out of an irrationally strong sense of civic duty. If most of us were freeriders, there could be no genuinely popular electoral system.
Third, without the partisan organization of legislatures and government, politicians could never efficiently restrict the agenda of conflict and debate to a basic set of important issues. They would instead stumble in and out of fragile log-rolls that would incorporate many unrelated items. The result would be policy immobility, rendering deliberation and participation beside the point.
Professional politicians in a democracy obviously need parties to satisfy their policy and office-seeking ambitions. But the rest of us also need parties. No parties, no "positive externalities" (in the language of welfare economics)--no streams of consistent and related policies, no agenda for public debate, and little prospect of even a modicum of voter attachment to the polity and its concerns. Thus, our current party system provides essentially the same "positive externalities," Aldrich is saying, as the earlier party system.
B ut a somewhat different take on the same facts is offered by Steven Schier in By Invitation Only. During the golden age of party politics, roughly the period from 1830 to 1890, we had something approaching a genuinely participatory democracy in this country. Today we have, in its place, a vast congeries of professionally managed "activation," that is, the stimulation and enlistment of thousands of small subsets of the citizenry in service of the ambition of an interest group or a candidate. Several kinds of professional consultancies are available to the well-heeled or the well-organized to accomplish their preferred strategy of activation: pollsters, media consultants, fundraisers, gatherers of demographic data, opposition and issue researchers, speechwriters, schedulers, and so forth. Schier catalogues them all succinctly.
The basic idea here is that parties now compete in a broad marketplace of service providers for the politically ambitious. Their historic monopoly on access to office and influence disappeared with the rise of primaries, referenda, campaign finance regulation, and a privately operated system of broadcast communications.
The loser in the shift toward a competitive market in political techniques is the mass of ordinary citizens. Following politics and getting involved in it is up to them. If they do not have the education, confidence, partisan-ship, or time to do so, no one will ask them. Expending resources to activate the already motivated voters is cost-effective. It is less cost-effective to pursue those who are not listed in the databases of the consultants.
In this way, the political system is a bit like the medical system: technologically advanced, expensive, and replete with a variety of coverages and exclusions. As a nation, we spend a huge amount of money on electoral politics and employ all the latest campaign techniques, but we do not get much average-voter turnout in return.
Schier's final chapter offers an exceptionally thoughtful treatment of possible cures for this state of affairs. The bottom line for reform, he suggests, is making party affiliation more salient to political candidates than it currently is. In response, politicians might have stronger incentives to cooperate with one another in mobilizing voters, rather than worrying only about their own constituency.
We should reorganize campaign finance so that parties control more resources than they do now. And the states could provide ballots that are organized as party slates. More states could do what Maine and Nebraska do, which is to allocate votes in the electoral college to whomever carries a congressional district and give the "Senate votes" to the statewide winner. These are among the most plausible reforms of the many that Schier discusses.
C ould it be that such reforms overemphasize the importance of political parties to democracy? David Mayhew is the one leader of the political science profession who has consistently resisted such enthusiasms. In the course of his career, he has helped to show that political parties have little to do with whether Congress works well, that states with weak parties are not necessarily less generous with social policies (and are sometimes more generous), and that from 1947 to 1990 divided government at the national level simply had no effect on the production of important public policy, budgetary balance, or the frequency or disruptiveness of congressional investigations of the White House or the executive bureaucracies.
It could also be the case that the party system, as Aldrich says, is not in decline but simply has acquired new forms. One might retort that the earlier system made for more active citizens. But cross-national survey research does not show that countries with party systems more like our earlier system have citizens more satisfied with how their democracy works than ours.
Nonetheless, Walter Dean Burnham was right to think as long and as hard as he did about cycles of decline and renewal in American party politics. Perhaps critical realignments never really existed, but political decline and renewal are hardly fanciful inventions of Burnham's towering intellect. They are the oldest and most important issues of political thought, going back to Aristotle.
For all their faults, political parties have been the essential foundation of both citizen involvement and citizen awareness of the issues facing a democratic polity. Perhaps nothing will come of letting our two-party system continue to become just one among many channels for citizen involvement, rather than the premier channel. It is more likely, though, that good things would come from trying, as Schier suggests, to make our party system more salient for voters and politicians than it currently is. ¤
Richard M. Valelly is a professor of political science at Swarthmore College.
