Party autonomy and two-party electoral competition.

• Jurisdiction: United States
• Date: 01 January 2001
• Author: Cain, Bruce E.

There is widespread sympathy these days for giving U.S. voters much more freedom and choice in making their selections. Some see choosing between only two parties as the political equivalent of not having cable television--excusable in the fifties, but not in the modern era. Even the increasingly prevalent use of the term "duopoly" for a two-party system has a pejorative implication of unnatural constraint.(1) The term "competition" in U.S. politics has traditionally meant only that the voter has more than just one viable candidate, party, or option to choose from.(2) Now, however, there are new claims to be assessed: that two parties are not sufficient given the complexities of modern society,(3) and, more to the point of this Article, that voters should be free to participate in several party primaries simultaneously.(4) The desire to have more choices has been amplified, and perhaps distorted, in recent years by the popularity of economic modes of analysis that suggest many parallels between markets and elections.(5) As highly competitive markets with numerous small firms tend to spur product innovation and lower prices, then perhaps giving voters more choices will result in better representation. But is it always true that offering voters more choices is necessarily better?

For those who would answer yes to the previous question, the Supreme Court's recent decisions in California Democratic Party v. Jones,(6) striking down California's blanket primary system, and in Timmons v. Twin Cities Area New Party,(7) defending a ban on fusion candidates in support of a two-party system, erect new judicial barriers to a better democracy. Instead of promoting more choice, these decisions seem to some to restrict voter choices and weaken the prospect of minor party challenges. I argue against the prevailing assumption that more choices necessarily result in better democracy. Democratic theory recognizes numerous forms of legitimate party systems and electoral rules, and gives us no basis to conclude that multiparty systems are more democratic than two-party systems, or that closed party nomination rules are less democratic than open or blanket systems that offer voters more choices. All conform to the basic requirements of a democracy as laid out by Joseph Shumpeter, Robert Dahl, Anthony Downs, and other modern theorists.(8) Looking at the effects of implementing a blanket system, there is no compelling state reason for forcing parties to adopt such rules when they would quite understandably prefer not to. Moreover, I argue that letting the parties decide how to bestow their party labels with only minimal constitutional constraint serves important functional purposes in a two-party democracy. It also saves the Court from being a party to yet another, unnecessary lock-in of one particular representational theory (i.e., hypercentrism).

There are two basic questions raised in California Democratic Party v. Jones. The first, largely a political science question, concerns the relative merits of different nomination systems and whether it can be said indisputably that a blanket primary is better than a closed primary. I will argue that it cannot. The second question is whether the state should be allowed to impose a nomination system upon the parties. Based on a principle of functional party autonomy, I will argue that the answer is no.(9) While V.O. Key's work usefully reminds us that party ties can also be found between elected representatives and the general electorate,(10) the central question in Jones concerns the party organization's right to establish the process and criteria by which it officially confers its party label upon candidates on the November ballot. Whether a significant number of party registrants and some elected officials might have favored the state's action is not relevant since the question at issue concerns the official party organization.

In the sections that follow, I will first set out evidence concerning the relative merits of different primary nomination systems. Much of this is derived from recent political science work in this area, especially a recent study of California's only election under the blanket primary system.(11) Secondly, I will consider the case for the Court's intervention based on a functional theory of party autonomy.(12) Unlike the position taken by the Brennan Center for Justice in the Jones litigation and by Rick Hasen in this volume, this theory does not distinguish between major and minor parties and does not hold that minority parties deserve different or greater protections than major parties themselves. I argue that the function and value of protecting minority voices (such as activists and party supporters with noncentrist positions) within the major parties is as valuable as protecting the minor parties. Finally, I will consider the question of how to reconcile the Court's intervention with my general aversion to judicial interference with the political sector. I argue that in Jones, the Court acted to prevent the state from the unnecessary lock-in by initiative of a particular theory of competition (i.e., hypercentrism). In other areas, such as redistricting, I have opposed the Court's attempt to define criteria of fairness that would lock in rigid and highly disputable theories of representation.

1. THE BLANKET PRIMARY CONSIDERED

   The basic political science question at the core of the Jones case is whether the blanket primary system enacted by Proposition 198 imparts any important, let alone compelling, state interest on California's electoral system. This divides into an empirical assessment--what are the actual effects of implementing a blanket system?--and a normative/legal assessment--to what degree are the measurable effects important for a democratic state and how do they fit into a sensible scheme of legal regulation? At the Jones trial and in subsequent studies of the blanket primary's impact in the one California election before the Court overturned Proposition 198, political scientists examined many different possible effects, such as the blanket primary's impact on campaign spending levels and incumbency reelection rates. Some of these causal relationships are valuable in assessing the pure policy problem of predicting what might happen when blanket rules are put in place, but do not address core state interests. The primary goal of imposing the blanket system was not to lower campaign costs or incumbency reelection rates. Indeed, there is evidence that some of these effects went the other direction; for example, incumbents were rewarded with even more support under the blanket system.(13) In general, the ancillary empirical findings serve to remind us that any set of rules has both good and bad unintended effects.

   The two potential state interests I want to focus on are higher participation and ideological moderation.(14) With respect to the former, it is clearly important for a democracy to have extensive participation by its citizens for several reasons. First, it prevents skewed or biased outcomes that may occur when key segments of the electorate are not active participants. Unless the portion of those who actually vote is representative of the entire population as a whole, election outcomes can get skewed.(15) In that case, a process that yields biased decisions can lose legitimacy in the eyes of citizens who feel shut out. Finally, since participation is partly educative--i.e., people inform themselves in order to make their choices--getting people to participate raises awareness and knowledge of the pertinent election issues. So if there is a strong and narrowly tailored link between the blanket system and wider participation, the state might plausibly have an important reason for imposing the blanket system upon the parties.

   But is it true? A blanket primary system is defined as one that allows voters the freedom to vote in any party election he or she chooses for each race. Thus, a voter can vote in the Republican primary for the president, the Democratic primary for state senator, and so forth. Since there is no restriction by party affiliation, independents and party registrants can freely enter any party's race. In a closed primary, only those who are registered with the party (typically some specified number of days in advance of the election) can vote in that party's primary. Independents and other party registrants are not allowed to participate unless they affiliate with that party prior to the election. In an open primary system, there is no prior party registration requirement, but the voter is allowed to take only one party ballot and may not cross party lines for different races. Finally, in the semi-open systems, independents are allowed to cross party lines, but no other party registrants are allowed to do so.

   The expectation that the blanket primary system might increase participation is based on the understanding that, at a minimum, independents would be able to vote in contests from which they were previously excluded, as well as the belief that some party registrants might become more energized by the opportunity to participate in another party's primary. The 1998 California primary provided a natural experiment for these predictions. With the blanket rules in effect, the turnout proved to be 29.8% of the eligible vote. Comparing this with the three previous non-presidential California primaries beginning with 1986, the 1998 figure represents an increase of 2.4% over the closed primary average of 27.4%. As Wendy K. Tam Cho and Brian Gaines demonstrate, assuming a constant turnout rate among partisans and nonpartisans, the surge can be completely explained by the addition of the nonpartisan registrants.(16) They conclude that there is no obvious sign in the aggregate data that the blanket primary energized party registrants or that the surge carried over into the general election in November.

   In sum, the effect...