Stability and Instability

Are presidential nomination campaigns relatively stable and predictable or are they unstable and unpredictable? Early news media expectations of upcoming nomination campaigns tend to see the frontrunner as winning all primaries and sweeping to victory. This is usually correct in terms of the outcome but often incorrect in terms of the process. The early frontrunner has won the nomination every time from 1980 on. However, in every contested campaign in the post-reform era, the frontrunner has suffered setbacks in the early contests. Often another candidate develops sufficient early momentum to challenge the frontrunner. Whether intentional or not, the expectation of stability allows the news media to describe the actual events of the campaign as unexpected and exciting.

Despite the consistent success of early frontrunners, presidential nomination campaigns have often been depicted as unstable. This view holds that the dynamics of the campaign (e.g., momentum) are such that several outcomes are possible. Aldrich assumes that “nomination campaigns can not be understood in anything but dynamic terms” and shows that the more active candidates participating, the more unstable the campaign (Aldrich 1980a, 1980b). The nomination campaigns of the 1970s and early 1980s seem to support this view. In 1972, frontrunner Muskie was defeated by a relatively unknown long-shot, McGovern. The 1976 Democratic campaign was like a roller coaster with Carter's popularity increasing dramatically in the early contests then declining later on; established candidates were eliminated, only to be replaced by fresh candidates. In the 1980 and 1984 campaigns, the early frontrunners (Reagan, Mondale) won the nomination after weathering a substantial and extended challenge. Nonetheless, the winner of the “invisible primary,” the candidate who leads in pre-Iowa polls and fundraising, usually wins the nomination (Adkins and Dowdle 2001; Mayer 1996). Because of frontloading and other rules changes, frontrunners now seem more likely to win than they were in the 1970s and 1980s.

Types of Campaigns

Presidential nomination campaigns can be categorized into four basic types: one-candidate campaigns, two-candidate campaigns, multi-candidate campaigns with an early dominant frontrunner, and multi-candidate campaigns without such a frontrunner. If the incumbent president is running, few if any others will run.

• One-candidate campaigns (Democratic 1996; Republican 1972, 1984) When there is a popular incumbent, he is not seriously challenged for his party's nomination. If George W. Bush's popularity remains high, he is unlikely to be challenged in the Republican primaries.
• Two-candidate campaigns (Democratic 1980, 2000; Republican 1976, 1992). Several recent presidents (Gerald Ford, Jimmy Carter, George H. W. Bush) with low approval ratings were challenged by a major candidate representing the other ideological wing of the party. Such campaigns are characterized by their zero-sum nature: one candidate's success is matched by his opponent's failure.

When the incumbent president is not running, it is common for five to nine candidates to enter the race.

• Multi-candidate campaigns with a clear frontrunner (Democratic 1972, 1984; Republican 1980, 1988, 1996, 2000). Here, one candidate dominates the early polls. The advantages enjoyed by the frontrunner are generally sufficient to gain the nomination. Nonetheless, it is typical for the frontrunner to stumble early and to be seriously challenged by one of the other candidates. Note that such challenges were more effective before the advent of a massive Super Tuesday (1972–1984) than after (1988–2000).
• Multi-candidate campaigns without a clear frontrunner (Democratic 1976, 1988, 1992). In some multi-candidate campaigns, there is no clear frontrunner. (There is always someone who leads the polls but that may represent little more than name recognition.) The 2004 Democratic campaign appears to be following this pattern. In such cases, the campaign appears to be especially sensitive to the results of the early contests.

Components of Presidential Nomination Campaigns

The dynamics of the presidential nomination system can be described in terms of interactions between the candidates' strategic behavior, the voters' behavior, and that of other actors (especially the news media) in response to the outcomes of state contests (see Norrander 1996). These interactions are conditioned by the rules and procedures governing nomination campaigns. The rules have both normative and strategic implications (Geer 1986; Gurian 1990; Haskell 1992; Crespin 2001). Institutional factors influence the strategic calculations candidates make in allocating resources. Resource allocations, in turn, affect the outcomes of the primaries (Norrander 2000; Haynes, Gurian, and Nichols 1997; Dunn 1994; Parent, Jilson, and Weber 1987). The results of the primaries alter each candidate's probability of nomination (viability) directly via the allocation of delegates and indirectly via media coverage, fundraising, and changes in the field of candidates (Aldrich 1980a, 1980b; Marshall 1981; Damore 1997). Institutional factors directly affect campaign spending, media coverage, and the allocation of delegates, and indirectly affect vote outcomes and fundraising; thus they influence the choice of the eventual nominee (Lengle and Shafer 1976).

During the early contests, relatively minor institutional factors can have substantial effects on the remainder of the campaign. It is easy to speculate that the outcome of some previous campaign would have been different under different circumstances. However, the following example suggests that those circumstances need not be very different. In the 1992 Democratic campaign, Bill Clinton's first victory was in the Georgia primary on March 3. The timing of this contest was helpful to the Arkansas governor. Since he had lost the New Hampshire primary and all the other early contests, his viability was in doubt; Georgia provided him with a much-needed boost going into Super Tuesday one week later. Originally, the Georgia primary was scheduled as part of Super Tuesday, but the state legislature moved the primary up one week. If Clinton had not had a victory in a southern state, he likely would have gone into Super Tuesday without the benefit of momentum. His prospects on Super Tuesday, and thus in subsequent primaries, would have been diminished. Thus, a one week change in the timing of one primary altered the dynamics in a way that may have changed the final outcome.

“Viability” is central to the unfolding of the nomination process. The dynamics of presidential nominations winnow the field of candidates until only one remains. Low viability and negative media coverage diminish a candidate's ability to raise money and thus to continue campaigning (Crespin et al. 2000). Viability is the subjective evaluation of a candidate's chances of nomination. (Electability refers to a candidate's chances of election.) The basis of such evaluation changes over the course of the campaign. In the invisible primary, a candidate's poll standings and fundraising success are the best measures of viability. During the early contests, viability is measured largely by news media verdicts. After that, the delegate count becomes the best measure of viability.

Voting in presidential primaries is substantially different from voting in a presidential election. Studies have shown that the major factors influencing individual votes in the general election are party identification, issues, and candidate evaluations. However, party cues are absent in the primaries since voters must choose among candidates in the same party. Information about issues tends to be considerably lower in the primaries than in the general election; by the time most voters become familiar with the candidates' positions, the race is effectively over (Brady and Johnston 1987). Candidate evaluations do play a part in voters' decisions but information about the candidates' experience, ability, and character is often limited. In the primaries, candidates' chances of success are powerful influences on vote choice. Most voters are unwilling to vote for a candidate with little chance of nomination; thus a candidate's viability is an important consideration (Abramson et al. 1992). Partisans generally prefer a candidate who has a good chance at defeating the other party's nominee in the general election; thus a candidate's electability may also be considered (Abramowitz 1989). The role of viability and electability as voting cues is critical because they are essential components of momentum.

By momentum, we refer to the process by which success in one primary leads to increases in the candidate's chances of success in subsequent primaries. Each good or poor showing by a candidate leads to good or poor (or non-existant) media coverage which in turn leads to favorable or unfavorable changes in the evaluations of voters in subsequent primaries, increases or decreases in fundraising, etc. (Marshall 1981; Mutz 1995). Since favorable media coverage, higher popularity, and greater fundraising generally contribute to higher vote outcomes, a strong showing in one primary tends to lead to strong showings in subsequent primaries. In presidential nomination campaigns, nothing succeeds like success.

Phases of the Campaign

During the invisible primary (see Buell 1996), potential candidates “test the waters” and some decide to enter the race. The relative stability of the invisible primary phase is often upset during the early contests. The competition in this phase is intense. Long-shot candidates know that unless they can attract the media spotlight with an unexpected victory, their quest for the presidency is likely over. Frontrunners realize that expectations for them are high and that other candidates are doing their best to derail their campaign. Furthermore, the early contests are typically in small states with low spending limits where frontrunners have the least advantage.

The strategic positions of the candidates may change dramatically during the early contests. In most multi-candidate campaigns, a long-shot has emerged from the early contests to become a major contender for the nomination. When there is a clear frontrunner, that candidate typically suffers a setback in the early contests. When there is an obvious runner-up, that candidate may be eliminated during this phase (e.g., John Connally in 1980, John Glenn in 1984, Phil Gramm in 1996). The relative viability of the candidates after Iowa and New Hampshire is often very different than it was before.

The sequence of contests can have a profound effect on the unfolding of the campaign and potentially the identity of the nominee. Typically, Iowa and New Hampshire each receive 10–20% of the total national news media coverage devoted to the nomination campaign; the other states typically receive less than 2% each. In every modern campaign, the eventual nominee finished in the top three in Iowa and in the top two in New Hampshire. Candidates like George H. W. Bush, Gary Hart, and John McCain became major contenders based largely on strong performances in Iowa or New Hampshire. A different sequence of contests might have lead to a different field of candidates, different opportunities or obstacles for certain candidates, different patterns of media coverage, different levels of information among voters in some states, etc. Plausible arguments have been made that Carter in 1980, Hart in 1984, and Clinton in 1992 would not have fared as well if the sequence of contests been only slightly different.

After New Hampshire there are a few scattered primaries followed by one or more multi-state events such as Super Tuesday. In the 1980s, several southern states agreed to hold an early regional primary (see Norrander 1992; Stanley and Hadley 1989). Since then, other southern and non-southern states have followed their lead. By 2000, the southern Super Tuesday had been eclipsed by a larger quasi-national primary involving California, New York, and a dozen other states. In the 1970s and 1980s, a candidate who did unexpectedly well in Iowa or New Hampshire could become a major contender in a matter of weeks as voters and contributors flocked to support the rising alternative to the frontrunner (Bartels 1988). The advent of an increasingly important and expensive Super Tuesday soon after New Hampshire seems to have diminished the time and opportunity for such candidates. For example, John McCain had a resounding victory over frontrunner George W. Bush in New Hampshire in 2000, but despite positive media coverage, fundraising success, and victories in other early contests, he was unable to compete effectively with Bush in the massive set of primaries that followed.

The 2004 Democratic Campaign

At the time of this writing, the first primary of the 2004 campaign is a year away. Nonetheless, one can anticipate that the Republican race will be a one-candidate campaign; George W. Bush seems unlikely to be challenged in the Republican primaries. In contrast, the Democratic race is far more likely to be tumultuous; half a dozen Democrats are running for the nomination. A crowded field is typical for the out party, especially when there is no heir apparent.

For the Democrats, the calendar has strategic implications for the way the campaign is conducted. At this point, the dates of many of the 2004 primaries and caucuses have not been set. If the calendar for 2004 is similar to that of 2000, we can expect (in the Democratic campaign) first Iowa and New Hampshire, then a few other early contests, followed by a massive Super Tuesday. The front-loaded calendar will put great emphasis on early fundraising, mass media strategy, and campaign organization. Much of this will take place during the invisible primary (Buell 1996). In 2004, the invisible primary is likely to be more visible than ever before since candidates will be “competing” earlier and the media is likely to cover the campaign earlier.

Under the Democratic rules approved for the 2004 nomination campaign, primaries and caucuses must occur between the first Tuesday in February and the second Tuesday in June. The party does, however, make special exceptions for Iowa and New Hampshire. Tentatively, Iowa will hold caucuses on January 19 followed closely by New Hampshire on January 27. For some years, South Carolina has held an early and important Republican primary, a “gateway” to the upcoming southern primaries. The South Carolina Democratic Party has decided to hold their primary on February 3, 2004. Depending on how many other states front-load to that first Tuesday in February, South Carolina may loom as an important next step after New Hampshire. Most of the Democratic candidates have already ventured to this state.

Each election year the rules change incrementally and each race brings a unique set of candidates and political circumstances; nonetheless, multi-candidate campaigns have followed relatively few basic scenarios. In the Republican campaigns of 1988 and 1996, one candidate emerged as the dominant front-runner during the invisible primary, that candidate stumbled through some of early contests but then swept virtually all of the remaining contests. In the Republican campaign of 2000, one candidate became the dominant frontrunner during the invisible primary, a strong challenger emerged in the early contests, but the frontrunner swept virtually all of the remaining contests. In the Democratic campaign of 1984 and the Republican campaign of 1980, a single candidate established himself as the dominant frontrunner, a strong challenger emerged in the early contests, and these two engaged in a protracted battle throughout the remaining contests. In the Democratic campaigns of 1988 and 1992, one candidate lead but did not dominate the field in the invisible primary, that candidate enjoyed some success during the early contests, and after some time effectively captured the nomination. This last scenario seems most similar to the 2004 Democratic campaign.

Al Gore's decision not to run in 2004 changed the expected dynamics of the nomination campaign. Several Democrats, including Connecticut Senator Joe Lieberman, former House Minority Leader Dick Gephardt, Massachusetts Senator John Kerry, North Carolina Senator John Edwards, and former Vermont Governor Howard Dean, are pursuing the nomination. (Other potential candidates include Florida Senator Bob Graham, activist Al Sharpton, and former Senator Gary Hart.) As of this writing, no frontrunner has emerged, but that could easily change over the next few months. In the post-reform era, there have been few instances in which there was no clear, dominant frontrunner. Whoever ends up as frontrunner, either before or after the early contests, would hope to avoid a long, divisive campaign. Winning the nomination of a divided party can be a hollow victory. A candidate's chances of winning the general election are better if he or she is able to win the nomination without a long, potentially damaging, primary battle (see Atkeson 1998, 2000).

Each candidate has competitive strengths or weaknesses, and the schedule of contests provides opportunities and hazards for individual candidates. National preference polls (among party identifiers), fundraising success, and positive media coverage in the invisible primary are related to subsequent success in the primaries. Several national polls taken in January 2003 show Lieberman leading with 19–29%, Gephardt and Kerry with about 10–17% each and Edwards with 7–14%; all other candidates were in the 2–7% range. Lieberman's lead may reflect his name recognition as Gore's 2000 running mate. Several candidates have the potential to raise substantial financial resources; the extent to which they realize that potential remains to be seen. Lieberman and Gephardt are both well known and may receive significant support from specific constituencies. Kerry has access to substantial personal wealth. Edwards has been very successful in his fundraising efforts; as of January 2003, his 527 organization (used by candidates as soft money vehicles) has raised far more than that of any other Democratic candidate.

Other factors, such as previous national campaign experience, “charisma,” and popularity in New Hampshire (see Jackson 2002) can also be beneficial. Among the current set of likely candidates, only Gephardt has previously run in the presidential primaries. That could help him avoid some of the slips and gaffes that sometimes plague presidential campaigns. Reports of Edwards' ability to connect with voters are common; this can be a valuable asset in Iowa and New Hampshire. On the other hand, his relative youth and lack of political experience could be detrimental. Kerry's early lead in the New Hampshire polls could cause reporters and thus potential voters to regard him as an important candidate. In addition, his status as a Viet Nam veteran might play well if the campaign focuses on issues related to national security.

A year before the primaries begin, no one candidate has emerged as the clear frontrunner. If this is still the case in January 2004, we can expect a process of “winnowing” to occur, providing one or more candidates with momentum while dashing the hopes of others. In past campaigns, the results of the Iowa caucus have provided cues to the voters of New Hampshire. Typically, the winner and runner-up of the New Hampshire primary have been among the top three finishers in Iowa. Similarly, the results of the New Hampshire primary have provided cues to the voters in subsequent states. In the post-reform era, the two leading contenders for the nomination have generally been the top two finishers in New Hampshire, and the eventual nominee has always been one of those two. (In 1988, Al Gore and Jesse Jackson skipped Iowa and New Hampshire but became major contenders by winning several southern primaries soon thereafter.)

In the weeks before the Iowa and New Hampshire contests, the news media will set expectations for the candidates. Deviations from these expectations will likely be interpreted as indications of the candidates' viability and electability. We can get a sense of the implications of the interactions between expectations and results by considering some possible scenarios based on very early polls. (Of course, much could change between now and next January; a week is a long time in politics.) This is more than mere speculation; consideration of possible scenarios before the fact illuminates potentialities of the process that are more difficult to perceive after the results are known.

The Iowa caucus is tentatively scheduled for January 19. Gephardt is from neighboring Missouri and he won in Iowa in 1988. Thus he is expected to do well there in 2004. This expectation, however, is doubled-edged. It seems likely that the news media will interpret anything less than a first-place finish in Iowa as a setback for his campaign. On the other hand, the candidate who comes in second, especially if second place is “better than expected,” might get a substantial boost in terms of media coverage. Such coverage would cue potential New Hampshire voters that that candidate is a potential winner and thus worthy of consideration.

Kerry is likely to face similar terrain in New Hampshire. Early polls there show him with a commanding lead. A victory would be helpful, since the New Hampshire winner usually fares well in subsequent primaries, but a loss could be devastating. For the other candidates, a second place finish would be helpful; the news media tend to focus on the top two finishers in New Hampshire while ignoring or disparaging the others. Finishing third or worse tends to brand candidates as outside the set of viable contenders. In order to compete effectively in the expensive multi-state Super Tuesday events, candidates generally need both substantial financial resources and positive momentum. The 2004 campaign could quickly develop into a two-person contest between the winner and runner-up in New Hampshire. There are, however, a number of specific factors that could alter this scenario.

First, Kerry will be expected to win New Hampshire. For him, a second-place finish would be detrimental. Second, Dean does not currently enjoy the level of fundraising potential that the other candidates do. Since he is from neighboring Vermont, a second-place finish in New Hampshire is unlikely to provide him sufficient momentum to raise enough money to be competitive in the multi-state Super Tuesday events soon after that. Third, as of this writing, only two states, South Carolina and Missouri, are scheduled to hold primaries right after New Hampshire, on February 3. If that does not change then Edwards would have the opportunity to show his strength in the South. On the other hand, if several other states move up to February 3, then only those few candidates who have substantial financial resources and who have already demonstrated their vote-getting abilities will be competitive in the big multistate events.

The number and identity of states holding primaries on or shortly after February 3 will affect the fortunes of the remaining candidates. If substantial frontloading and “bundling” of primaries occurs, resulting in a quasi-national primary, then the candidate with greatest financial resources and support from party elites is likely to win those primaries and thus the nomination. Once the big multi-state events begin, there will be no time for candidates to build momentum and financial resources as there once was.

In the 1970s and early 1980s, presidential nomination campaigns were characterized by strong momentum effects. Candidates who exceeded expectations in the early contests soon had the popular and financial support to compete effectively in the remaining primaries, which were spread out over several months. Since then, the front-loaded calendar, especially the presence of early multi-state events, seems to have reduced the impact of momentum, leading to relatively brief campaigns dominated by the front-running candidate. The upcoming 2004 Democratic campaign could resemble either of these types, or some combination thereof. If a dominant frontrunner does not emerge in the invisible primary, then expectations and momentum seem likely to play an important role during the early contests. It will be instructive to observe the extent to which momentum developed in the early contests enables a candidate to compete effectively in the remainder of the campaign.