This article first appeared in The National Voter, vol. 53, No. 1, September/October 2003, pp. 8-12. Except for brief quotations embodied in articles or reviews, no part of this article may be reproduced or transmitted in any form or by any electronic, mechanical or other means without permission in writing from the publisher, League of Women Voters of the United States.
Race for the Nomination—In Search of Reform
By William G. Mayer
Over the last several decades, few institutions in American national government have been as consistently controversial as the presidential nomination process. The basic rules of this process were almost completely rewritten in the early 1970s, and from the moment the reforms first took affect, they faced a firestorm of criticism. The purpose of this article is to present a brief overview of how the current system works, some of the major criticisms that have been lodged against it, and what major alternatives have been proposed for reshaping the process.
The Current System
The American presidential nomination contest is a highly complex system, including a multitude of distinct rules and procedures, nearly all of which have generated some measure of controversy and a lengthy academic literature describing their alleged effects. What follows, then, is clearly not an exhaustive description of the current system, but rather an attempt to highlight its most important features.
First, the final decision about which candidate becomes a party's presidential nominee is still made by the national conventions, which are usually held in mid- to late summer of the election year. Though the convention decision has become increasingly pro forma, much like the voting in the Electoral College, the ultimate goal of the candidates—the centerpiece of all their planning—is still to win a majority of the delegates to their party's national convention.
Second, the national parties each promulgate general guidelines that specify how states are to select their national convention delegates and what sorts of things they must and must not do. As a general matter, national Democratic rules regulate the delegate selection process in considerable detail; the Republicans tend to give more discretion to states.
Third, the national rules in both major parties allow states to select their delegates in one of two ways: by primary or by a caucus-convention system. Primaries are elections, generally held under the auspices of state government, that are used to select or bind the national convention delegates. Caucuses are a multi-stage process, usually beginning with mass meetings held in each precinct or voting district, which select delegates to district and state conventions, which in turn select the national convention delegates. In recent years, the delegate selection process in both parties has clearly been dominated by primaries. In 2000, for example, 37 states held Democratic primaries, 42 states held Republican primaries.
Fourth, presidential primaries and caucuses both tend to be open to essentially any Democrat or Republican who wants to participate. Some states require participants to be a registered member of that party, but no further demonstration of past or future support for the party is required.
Fifth, besides choosing between caucuses and primaries, each state has considerable latitude in deciding when to select its delegates. In 2004, Republican rules require states to hold their primary or caucus between the first Monday in February and the third Tuesday in June; Democratic rules permit primaries and caucuses to be held between the first Tuesday in February and the second Tuesday in June, but provide specific exemptions for Iowa and New Hampshire. Within those "windows," the states themselves decide whether to select their delegates on, say, the first Tuesday in February or the third Tuesday in May.
As noted earlier, the contemporary presidential nomination process has been criticized from a variety of perspectives. To many observers, perhaps the greatest single shortcoming of the "reformed" system is that it has taken the nomination decision away from organized, institutional parties and entrusted it to ordinary voters who participate in primaries and caucuses. When compared to the system that preceded it, the current system gives substantially less weight to the views of party leaders and elected officials. Beginning in 1984, the Democrats made some attempt to rectify this problem by granting automatic delegate status to senators, governors, members of the House of Representatives and national committee members, but the number of "superdelegates" is probably not large enough to alter the system's fundamental dynamics.
Another major criticism of the new process is simply that it is too long and too expensive. In the nomination races of the 1950s and 1960s, even the most ambitious presidential aspirants did not announce their candidacies until the beginning of the election year or the last few months of the year preceding the election. Today, major presidential candidates routinely launch their campaigns in the winter or early spring of the year before the election—more than a year before the national conventions, and at least a year and a half before the General Election. For sitting senators and House members, such a long period of active campaigning almost inevitably requires presidential candidates to neglect their governmental responsibilities. It also, of course, increases the cost of running for president.
A third major criticism of the presidential nomination process concerns the outsized role that it gives to two small and not very representative states, Iowa and New Hampshire. Because these states hold the first caucus and first primary, respectively, they see far more of the candidates than any other states, receive substantially more press coverage, and have far more influence on the final outcome. By contrast, California, which traditionally held its primary in the first week in June, generally found that the nomination races were effectively over by the time that state selected its delegates.
Given all the advantages that accrue to early voting states, over the last two decades the delegate selection calendar has become increasingly front-loaded. Where primaries and caucuses were once spread out rather evenly over the delegate selection season, most primaries and caucuses now take place within a few weeks after the delegate selection season formally begins, with the result that the nomination races are effectively settled by early or mid-March.
Front-loading has a number of undesirable effects on the way we select our presidential candidates. First and most important, it greatly compresses the time that voters have to learn about the major candidates. Most voters do not start to pay attention to the nomination races until the delegate selection process begins. When the system was less front-loaded, this meant that the voters had three or four months to watch the candidates and learn more about their policies and personal abilities before reaching a final decision. Today, as the decision gets made more rapidly, the system has become less flexible, less deliberative, and less rational.
Front-loading also undermines both the extent and the quality of voter participation. As the primary and caucus calendar has become more front-loaded, it has become routine for nomination races to be settled in the early spring. Even with front-loading, this means that lots of states select their delegates at a time when everything of significance has already been decided. In 2000, for example, 25 states held their primaries after both Bill Bradley and John McCain had officially withdrawn. The result, not surprisingly, is that voter turnout in the presidential primaries declines quite substantially in the later stages of the race.
If the contemporary nomination process is easy to criticize, it is a good deal more difficult to say what to do about it. Although a number of remedies have been proposed, it is not clear that any would solve all of the problems just discussed; many would probably carry significant negative consequences of their own. While it is impossible to do justice to all of the many proposals that have been made for changing the nomination rules, to get a sense of the difficulties and complexities of the problem, it is worth taking a look at two of the most talked-about proposals: a national primary and a system of regional primaries.
Ever since the Progressive Era, many of those who were dissatisfied with the presidential selection system have wanted to scrap the current system, with its complicated pattern of fifty distinct primaries and caucuses, and replace it with a single national primary. A national primary has two principal virtues: it is simple and straightforward; and it would treat all states equally. In particular, it would guarantee that no state had the kind of outsized, disproportionate role now played by Iowa and New Hampshire, and that no state would hold its primary after the effective nomination decision had already been made. In return for those benefits, however, a national primary has a number of severe problems that make it a highly questionable option.
Above all, a national primary would give an enormous advantage to early front-runners and candidates who were already well known and well financed. One advantage of having the current process start in small states like Iowa and New Hampshire is that it gives long-shots, outsiders and insurgents a good venue in which to make the case for their candidacies. A full-scale campaign in both states can be mounted for a fraction of what it would cost to compete in a national primary. And, precisely because the total electorate in both states is so small, face-to-face, "retail" politics counts for a lot more than it would in a country with 200 million eligible voters. If the fierce scramble for campaign money is one of the less attractive features of the current system, it would only grow worse under a national primary.
Depending on how it is structured, a national primary might also lead to the nomination of a candidate who is supported by a small, highly committed minority, but is considered unacceptable by a large segment of that candidate’s own party. If there are six or ten declared presidential candidates seeking a given party's nomination—which almost always occurs today except when an incumbent president is running for a second term—a candidate could win a national primary with as little as 25 or 30 percent of the vote. This might lead one or both parties to nominate someone totally inappropriate, who appeals to a very small minority.
To avoid such a problem, most recent national primary proposals have included a provision requiring the winner to receive some minimum percentage of the total vote (usually, 40 or 50 percent). If no candidate exceeds that threshold, there would be a runoff election between the top two finishers. Yet a runoff provision does not guarantee that at least one of the finalists will be minimally acceptable to most party voters. Nor is it clear that Americans would welcome the possibility of holding three national elections (two national primaries and then a General Election) within the space of three months.
Though they differ in their details, regional primary proposals generally call for some central agency—usually either the national parties or the federal government—to divide the states up into a number of regions, each of which would hold a set of primary elections on a single date. In 1999, for example, the National Association of Secretaries of State (NASS) put forward a plan that created four such regions—East, South, Midwest, and West. On the first Tuesday of each month between March and June, one of these regions would go to the polls, with the order rotating every four years.
Perhaps the principal advantage of a regional primary system, according to its supporters, is that it would make campaigning easier and more efficient, by cutting down on travel time and allowing candidates to advertise in media markets that cut across state lines. Regional primaries also hold out the hope of forcing candidates to confront the particular, often unique problems that confront each region.
There are also, however, a number of important disadvantages to regional primaries. Depending on how many regions are established, regional primaries might also give a huge advantage to early front-runners and those candidates with ample war chests. If the states are divided into just four regions, for example, every candidate would be forced to compete in twelve different states at the same time. Even with all the purported efficiencies of regional campaigning, long-shots and insurgents would almost certainly find it difficult to run a competitive race under such circumstances.
In addition, regional primaries would confer a significant advantage on any candidate who happened to be particularly strong in whatever region went first. Region is a very important variable in explaining primary outcomes: in almost every recent presidential nomination contest, at least one candidate has run significantly better in one region than in the others. Given the importance of momentum, this might mean that major-party nomination decisions would hinge on the essentially random factor of which region went first. In 1992, for example, Bill Clinton's candidacy would likely have been doomed if the southern states had voted last: for the first five weeks of that year's delegate selection season, Clinton didn't win a single primary or caucus outside the South.
In recent discussions of how to reform the presidential nomination process, a great deal of attention has generally been given to comprehensive reform proposals such as national or regional primaries, both of which would make significant, fundamental changes in the system we currently use for selecting presidential candidates.
For those who are concerned about front-loading or the length and expense of the process, a better alternative might be to accept the basic framework that has now been in place for more than three decades, and look for incremental reforms that might alleviate some of its more pressing shortcomings.
For example, both parties already have rules that place broad limits on the timeframe during which primaries and caucus can be held. Separately, or together, the parties might want to consider limiting the number of states that can vote prior to some specific date, or pushing back the entire delegate selection calendar so that it begins in March rather than January or February.
Then, too, many of the most criticized features of the current system are attributable, in whole or in part, to the difficulties that candidates frequently have raising the huge amount of money necessary to run a nationwide nomination campaign. Increasing the contribution limits, or the federal matching fund ratio, might make the results of Iowa and New Hampshire less consequential by allowing losing candidates to stay in the race longer, and might give more successful candidates a better opportunity to expand their campaign to other states.
Finally, while the number of automatic delegate seats given to party and elected officials currently account for only one-sixth of all the delegates at the Democratic National Convention (and none of the Republican delegates), they do seem to have had the positive effect of giving the party leadership a somewhat greater voice in the presidential nomination process.
Incremental reforms such as those posited above, of course, will not resolve all of the shortcomings of the current nominating process. Still, while it is easy to appreciate some of the attractions of comprehensive reform proposals—in particular, they seem so much more "rational" and orderly than the current process—it is unlikely that any of these comprehensive reforms would actually make the system better. In substantial respects, they would probably make the system worse.
William G. Mayer is Associate Professor of Government at Northeastern University.