The Joan Shorenstein Center on the Press, Politics and Public Policy. John F. Kennedy School of Government, Harvard University
The Vanishing Voter
About the Project
Weekly Updates
Voter Involvement Index
Results Archive
Shorenstein Center Home

Public Involvement and the 2000 Nominating Campaign: Implications for Electoral Reform

Policy Implications: Alternative Nominating Systems

Does the current system of nominating presidential candidates serve to broaden and deepen public engagement and information? Our findings lead us to conclude that it is not well designed for this purpose. It may serve the needs of the parties, particular candidates, and the voters in a few states, such as Iowa and New Hampshire. It does not, however, serve the general public's interests.

"Front loading" of the primaries (the tendency of states to schedule their contests on the earliest available dates) is a major reason. Front loading forces the candidates to campaign heavily in the months before the first contests, contributing to the public's perception that the campaign "starts too early" and "lasts too long." Yet because the public's attention is relatively limited until the Iowa and New Hampshire contests, citizens do not learn very much about the candidates in the months leading up to these contests. Then, soon after the public starts to pay closer attention and acquire information about the candidates, the heavily front-loaded Super Tuesday brings the active campaign to a sudden halt. Yet the conventions are still months away, heightening the public perception that the campaign is needlessly long. And because the nominating campaign is basically "over" in people's minds, they disengage from the campaign and begin to lose some of their knowledge of the candidates. Meanwhile, citizens in states that have yet to hold their primaries and caucuses feel disenfranchised.

For some of these reasons and others, the Democratic and Republican parties are reviewing alternatives to the present system. They are examining a number of options, including rotating regional primaries, population-based primaries, and restrictions on front loading. The Secretaries of State from a majority of the states have proposed a specific change for the 2004 election: rotating regional primaries.

Would any, or all, of the proposed alternatives result in increased levels of public engagement and information? It is this question to which we now turn.

Criteria for Evaluating the Alternatives

Assuming that the public's level of engagement and information are overriding considerations and that the public's complaints about the current system should be taken into account, an alternative nominating system would be preferable to the current one if it created:

  1. A shorter campaign;
  2. A nominating process that remained competitive for a longer period of time in order to give the public a greater opportunity to engage the campaign and to become informed about the candidates;
  3. A briefer interval between the decisive contests and the conventions in order to help people sustain the levels of public engagement and information they had attained when the nominating campaign peaked; and
  4. A system that increases the likelihood voters in all states will have an effective voice in the selection of the nominees.

Evaluating the Alternatives

These criteria are not necessarily consistent; the first and second ones, for example, can conflict in some situations. Nevertheless, they are a basis for a public-centered evaluation of alternative nominating systems. The following are brief evaluations of some of the proposed alternatives:

  1. The Current System Without the Front Loading: If front loading could be reversed and the system restored to its original 1970s form, the public would likely be more responsive to the process. The campaign would still be a lengthy one-which would be a source of public dissatisfaction-but there would be more Tuesdays during the initial phase in which a single state's primary was scheduled.

    This would increase the likelihood that the race would remain competitive for a longer period, which would result in higher levels of public involvement and public learning. It would also increase the public's sense that the process was "fair": a larger number (but not all) of the states would have an effective voice in the selection of the nominees. And if the race were competitive for a longer period, the public's sense that the process itself was overly long would probably diminish.

  2. Rotating Regional Primaries with Iowa and New Hampshire's Opening Positions Preserved: This system would probably not reduce the public's dissatisfaction with the nominating process and could conceivably increase it. This system would not shorten the campaign appreciably. It would reduce the number of dates on which state contests were held but the process would still stretch from February to June.

    Public involvement and learning would increase with the holding of the Iowa and New Hampshire contests but would still be far short of an optimal level when the first regional primary was held. This primary date might well mark the end of the race; a candidate with money, name recognition, and party support could sweep or nearly sweep the region's contests. A candidate with these advantages would almost certainly prevail by the second round. Voters in other regions would then feel they had been denied a voice in the outcome. This perception would be heightened by the regional identities that the system would generate. The perception that the process is unfair would be especially pronounced if the victorious candidate had his or her political base in an opening region.

  3. Population-Based (Delaware) System. This system, as it has been proposed, would start with contests in the ten least populous states, followed a few weeks later with the contests in the next ten least populous states, and so on. It would end with contests in the ten most populous states. This population-based system would not shorten the campaign appreciably. It would reduce the number of dates on which the state contests were held but the process would still stretch from mid-winter through early June.

    This system is more likely than the regional system to produce a sustained race although it is unclear whether the typical race would extend to the last set of primaries. In all likelihood, it would not last through all five rounds. Repeated instances of this outcome would generate intense dissatisfaction with the system in the most populous states their voters would feel completely disenfranchised by the system since the nominations would always be decided before their primaries were held. However, the system would probably create more overall public involvement and learning than the regional system because it could be expected to unfold more slowly.

    (NOTE: If this system receives serious consideration, the standard proposal might warrant revision. Iowa and New Hampshire's opening positions should probably be preserved. These two contests would serve to trigger public involvement and learning in advance of the first multi-state wave of contests. If Iowa and New Hampshire led off, it might make sense to cluster the remaining states in groups of twelve in order to confine the campaign to four subsequent dates. Moreover, it might make sense to include a regional factor in the selection of the sets of twelve states-that is, the first twelve states would not be the twelve least populous states (excluding New Hampshire) but instead the three least populous states from each of the four regions. This configuration would increase the public's sense that the process is a fair one. If the Delaware plan as originally proposed was adopted, no southern state would be included in the first cluster of primaries.)

  4. National Primary System. (NOTE: There is little sentiment for a national primary within party circles; however, the purpose of this paper is to present a public-centered view of the nominating process.) A national primary is far and away the public's preferred alternative. There are, as previously indicated, two reasons for this sentiment: a national primary would dramatically shorten the campaign and would be "fair" in that all states would vote on the same date.

    The critical question about a national primary is whether the voters would get involved deeply enough far enough in advance of primary day to cast an informed vote. To have any chance of working effectively, the system would require activation mechanisms in advance of the national primary, such as a beauty contest in New Hampshire or a series of very high-profile nationally televised debates.

  5. "Old-Style" National Party Convention System. This option would shorten the presidential campaign (and reduce the role of money in presidential nominations). A national convention system would also reduce the involvement and information burdens that the current system places on ordinary voters; these burdens would be transferred to party activists. A hard-to-answer question about this system is whether people would regard the system as "fair." Our polling indicates that Americans are willing to countenance a return to the convention system. But a full test probably wouldn't come until the system's first application.

    If the convention and the run-up period (which presumably, as in the old days, would include a few primary elections) were dominated by vigorous and open competition for delegate support, the general public would likely accept the system's legitimacy. (Objections from opinion leaders probably eliminate this system from serious consideration by either party. It is commonly accepted in American politics that the selection of presidential nominees is best accomplished through a primary-based system.)

  6. Hybrid System: There is a hybrid alternative that would satisfy all four requirements. This system would start on the first Tuesday of April with the first of five single-state contests each scheduled a week apart. (For illustrative purposes, let's assume the campaign starts in Iowa; then goes to New Hampshire in the second week; Georgia or South Carolina in the third week; Washington or Oregon in the fourth week; and Ohio or Michigan in the fifth week.) After the last of these contests, there would be a four-week interval that would allow the candidates to campaign around the country and participate in televised debates. Then, all remaining states would hold their contests on the same day (roughly the first Tuesday in June).

    In this system, the campaign would be dramatically shorter; there would be a nine-week period of intense activity during which the public could engage the campaign; and all states would be active participants, since the nominating races would not be decided until the final day. This hybrid system combines features of the current system, before front loading altered it, with features of a national primary (45 states would hold their contests on the final day).

What Lies Ahead >>

BackShorenstein Center HomeKennedy School Home

Please email comments and suggestions regarding this web site to our .
Copyright ©2001 by the President and Fellows of Harvard College.
Reporting copyright infringements