- What Are Parties For?
Parties are the core institution of democratic politics. As E.E. Schattschneider puts it, “democracy is unthinkable save in terms of parties.” Many analysts of the democratic process have emphasized the relationship, but perhaps no one more systematically than Joseph Schumpeter. He defined democracy as a system in which the people select the governors from among competing elites, meaning organized groups led by politicians.
Those who seek to gain or keep office must present the voters with alternative policies, personalities, or critiques of extant institutions— political, economic, or otherwise. As John Aldrich reiterates, office seekers are rational actors who will do all they can to appeal to the voters, who in turn have interests and values stemming from various identities such as economic position, gender, ethnicity, race, caste, religion, region, and so forth.
Politicians have always sought to maximize their appeal. Before the 1930s, they either guessed what voters wanted by judging crowd reactions, or had confederates report their impressions of public opinion. But such techniques pale beside modern opinion surveys.
Negative campaigning—criticism, fair or dirty, of opponents—has always marked polities with free elections. The longest continuous such system, that of the United States, has witnessed vicious campaign tactics almost from the start. Thomas Jefferson and his Democratic-Republicans [End Page 169] were denounced as tools of the immoral French Revolutionists. Their Hamiltonian rivals, the Federalists, were called Britain’s hirelings. Allegations of corruption and immorality were common. Charges of sexual misconduct and even of miscegenation were known long before the Civil War, as were tales of bribery and vote-buying. Nothing, it seems, is new under the electoral sun.
Aldrich repeats that “democracy is unworkable save in terms of parties.” They harness a “nation’s leadership . . . to public desires and aspirations.” But if this is true, Aldrich asks, why do so many analysts of contemporary politics conclude that U.S. parties are in decline? In a work intended to demonstrate the utility of rational-choice theory, Aldrich ascribes recent changes to politicians’ personal (rather than partisan) ambitions. The fulfillment of these, he notes, following Richard Fenno, is “sought in government, not parties.” He sees the so-called decline of parties as “the change from what I call the ‘party in control’ of its ambitious office seekers and holders to the ‘party in service’ to them,” rather than as a real decline.
Aldrich does not deny that institutions and history shape the nature and behavior of parties. Thus he notes that in a pluralistic country like the United States (and all others, I would add), parties are needed to bring together “many and diverse groups” into governing majorities. Those who determine how such coalitions operate are the office seekers and holders and others who seek benefits from the parties.
In discussing the U.S. two-party system, Aldrich relies heavily on Maurice Duverger (and also, it should be noted, on Kenneth Arrow, Anthony Downs, William Riker, and other formal theorists). Without going into the details of Aldrich’s formal rational-choice analysis, I would argue that the effort does not contribute much to our knowledge of the changes in the behavior of parties beyond what we can understand by long-established methods like systems analysis and individual-actor analysis, both of which also assume rational behavior.
Aldrich tells us, not surprisingly, that “the reason for affiliating with a major rather than a minor party is relatively simple to explain”: the chances of getting elected are higher in a major party. The needed “theoretical underpinnings” that he seeks to provide for Duverger’s law—“that elections decided by plurality or majority rule yield a two-party system”—are the “wasted-vote thesis” and the recognition that it is “in the party’s interests . . . to be one of the two major parties.” Hence parties seek to recruit the strongest candidates, and these in turn seek to reduce the effective choices to two.
While these assumptions are valid most of the time, they do not explain the emergence of third...