(Page 19 of 32) - Gaining traction: How Do Frontrunners Surface Before the Presidential Primaries? authored by Dowdle, Andrew. and Steger, Wayne.

Perhaps the most interesting result in Table 4a is that the lagged Gallup variable does not

have a significant relationship to candidates standing in national Gallup polls during January—
when it matters most for the outcomes of the primaries. In multiple permutations of these
models, there is no significant relationship between national Gallup poll standing of Democratic
standing during the pre-primary year, and their position in national Gallup polls during January
of the election year. This represents the stark disjuncture in Democratic nomination campaigns
between what is happening during the pre-primary year and the period immediately preceding
the caucuses and primaries. The considerable volatility of public support for Democratic
presidential candidates continues right up until the time for voting begins in most of these
nomination cycles. Even in the two years with a relatively clear front-runner in Democratic
polls, 1984 and 2000, the front-runners’ support in national Gallup polls varied considerably.

The patterns are considerably more stable across the quarters of the pre-primary year and

January of the election year for Republican nomination races. Candidates’ home state
constituency size has a significant effect only in the first quarter of the pre-primary year, and
fades thereafter. Network campaign coverage of the candidates is significantly related to
Republican candidates standing in national Gallup polls only during the first and second quarters
of the pre-primary year and again in January of the election year (the same as in the models for
the Democrats). Perhaps most indicative of what is going on in the Republican races, the lagged
Gallup variable has very large, positive relationships to Republican candidates’ standing in the
current quarter/January. Indeed, the effect is almost exactly even for the 3

quarter—meaning

that there was practically no movement in Republican national Gallup poll results between the
second and third quarters. Overall, the results indicate a very high degree of consistency across
quarters in the public support of Republican presidential candidates. These nominations appear
to be largely determined by the 2

quarter of the pre-primary year.

Robustness of the results

In addition to the models reported here, we analyzed numerous variations of these

models, to assess the effects of a other variables identified in various studies as being potentially
relevant to the outcome of the nominations. Specifically, we find that length of time in the race
during the pre-primary period has no effect on the APV or on Gallup poll standing prior to the
primaries. In part, this may be a function of how to measure entry in the race since candidates
may be actively campaigning prior to entering the race publicly or officially. As likely, however,
is that entry in the race is a strategic decision for some candidates. Some wait while others jump
in early to get a head start on the competition. After 1976, there is no significant relationship for
doing so.

We also estimated a series of variables reflecting candidates’ prior positions including

dummies for current and former governors, senators, representatives and candidates having other
appointed government experience or none at all (e.g., Burden 2002; Steger, n.d.). No variable
aside from the vice presidency has any significant relationship with the final APV or national
Gallup polls. We also find little support for the distinction between policy advocates versus
career politicians (e.g., Schlesinger 1966, 1975; Aldrich 1980; Steger, 2003). Beachler (1996)
and Adkins and Dowdle (2001) found that Southern Democrats have an inherent advantage in
presidential nominations owing to their unique ability to appeal to voters in these states. We find