The point made below by Craig Allen Smith has probably been made many times. But Smith’s presentation socked me in the jaw:
Because journalists report “stories” they are necessarily guided by the logic of “good stories,” including the tension associated with unexpected outcomes. Dramatic logic therefore thrives on erroneous predictions, which foster dependency on narrators and undermine independent thought. The paradox of narration is that seeming correct enhances the narrator’s perceived expertise, whereas seeming incorrect heightens the drama and the audience’s need for authoritative narration. Right or wrong, the narrator’s role is enhanced. Conversely empirical logic relies on observation to test the narrator’s projections. Here, seeming correct enhances the narrator’s credibility, whereas seeming incorrect undermines both the narrator’s credibility and the explanatory power of the narrative. (p. 520)
The problem is not so much that campaigns are “unpredictable.” The problem is that incorrect predictions make for better journalism when “journalism” is understood in a certain way.
As journalists constructed their explanation of presidential campaigns through 1988 they used a dramatic logic that is empirically illogical. Those who interpret election laws and procedures, voting histories, and opinion polls as mileposts in a “horse-race” misinform their audiences about the democratic process. The problem stems not from a lack of data or information. The problem is that unwarranted expectations can never be exposed by “story journalism” because unwarranted expectations are a sure-fire source of dramatic surprises. When observations contradict mythic expectations, they lead to even more fanciful melodrama. (p. 527)
Both quotes come from Smith’s “The Iowa Caucuses and Super Tuesday Primaries Reconsidered: How Untenable Hypotheses Enhance the Campaign Melodrama”, published in Presidential Studies Quarterly in 1992 (JSTOR link). Smith’s contrast between “dramatic logic” and “empirical logic” also relates to recent posts on using the scientific method in journalism (see Robert Niles, Matt Thompson).