Expert Commentary

Narrow victories and hard games: Revisiting the primary divisiveness hypothesis

2010 study from the University of Wisconsin on how hard-fought presidential primaries unfold and their implications for the general election.

In the world of political science scholarship, there has been extensive research on the question of how hard-fought primary battles affect candidates in the general election. Do they hurt by polarizing the party faithful — the “sour grapes” idea — and providing ammunition for general election rivals? Or do they help by raising the media profile of the eventual nominee in a useful way? The narrowness of the primary race and how aggressively candidates attack each other are two additional variables, as scholars from the University of Wisconsin note in a 2010 study.

Published in American Politics Research, “Narrow Victories and Hard Games: Revisiting the Primary Divisiveness Hypothesis” uses campaign advertising data from the 2008 Democratic presidential nominating contest between Barack Obama and Hillary Clinton to measure degrees of negativity and, ultimately, to test how this affected Obama in the general election. The researchers “use the tone of campaign ads to measure the divisiveness of primary contests and examine the extent to which our measures of divisiveness overlap with standard measures of competitiveness.”

The study’s findings include:

  • The competitiveness of the race in a given state was correlated with a greater volume of ads but not necessarily more attacks: “Contrary to the expectations and assumptions in the divisive primary literature, we did not find that more competitive nominating contests were more divisive than those won by a safe margin. In 2008, Clinton and Obama were actually less likely to ‘go negative’ in tight races, and we find no support for the assumption that ‘the closer the primary, the more divisive it is for the party.’ “
  • The “inverse relationship between competition and our measure of negativity appears to be driven by the campaign strategy of the Clinton camp. Overall, Clinton’s ads were more negative (13% of all airings) than those aired by the Obama campaign (6% of all airings)…. The Clinton campaign was more likely to ‘go negative’ in areas where Obama won fairly easily … as well as in places where she won handily…. On the other hand, both candidates avoided running negative ads where the race was close.”
  • Overall, there is “little evidence that the tone of the Obama and Clinton campaigns harmed the Democratic Party in the general election. Indeed, it appears that the more vigorously contested the nomination contest, the better Obama performed in November.”
  • Further, “Obama was marginally helped by close nominating contests. After controlling for previous Democratic performance and including state-level random effects, a one percentage point increase in relative competitiveness was associated with approximately three tenths of a percentage point increase in Obama’s general election vote share. This finding was the same whether we used our measure of relative competitiveness or one based on the Democratic winner’s margin of victory.”
  • A comparison of advertising between the Democratic and Republican campaigns in 2008 also yielded an interesting result: “Turning to the tone of the Democratic nomination campaign, negative ads accounted for an average of 6.8% of all primary ads aired. The Democratic contests were more negative than their Republican counterparts, with only 3.5% of Republican ads deemed negative.”

The authors note that their “analysis was limited to 2008 and that the positive effects found for primary competitiveness may be conditional on the unique circumstances of the historic nominating fight between Clinton and Obama. Nevertheless, our results challenge the assumption that competitiveness implies divisiveness.”

About The Author