Advertising, Iowa and Newt: It’s Not Over Until The Big Head Croaks

Did Newt Gingrich make a mistake in not responding in kind to the onslaught of negative advertising directed at him in Iowa this past month?  Most political observers are blaming Gingrich’s decline in the polls – his support has dropped by about half in the span of a month – to his decision to take the high road despite the media blitz targeting him. Figures released by the Campaign Media Analysis Group show that 45% of all the ads aired to date in December have been negative attacks on Gingrich.  Much of that advertising has been funded by SuperPacs who operate on behalf of candidates, but without – in theory – any direct connection to those candidates. In contrast, Mitt Romney largely escaped a similar fate; CMAG figures indicate that negative ads against Romney comprised only 20 percent of all television ads in Iowa this month.

Pundits are claiming that this more than 2-1 ratio in negative ads is the primary reason that Gingrich’s lead over Romney has evaporated. But are they correct?  Political scientists’ views toward the effectiveness of campaign advertising have evolved through the decades.  Initial studies in the 1940’s and 50’s suggested that advertising had “minimal effects” on voters’ attitudes and behavior, but more recent studies have painted a more nuanced picture, suggesting for instance that advertising may have substantial effects on voters preferences and turnout. But this is an evolving field of research, and much of the work on media effects in general is based on research conducted in simulated settings. For reasons of research design (and money), it is much harder to gauge the real world impact of paid advertising.

If there is growing evidence that campaign advertising does have an impact on voters’ preferences, it’s still not clear why.  How does negative advertising work?  One theory is that by providing new information regarding a candidate, negative ads cause voters to update their assessments of that candidate.  So, when watching the debates, audiences were impressed by Newt’s command of issues and his policy pronouncements. However, when told by a multitude of advertisements that Newt lobbied for Fannie Mae and Freddie Mac, voters began to reassess whether he was a true conservative who believed in smaller government and who really opposed these government-backed mortgage lending giants.  Similarly, ads linking him to Nancy Pelosi and supporting global warming, or advocating for the individual mandate, had the same impact – they provided new information that contradicted the story he was telling.  In short, negative ads have durable effects because they change people’s opinions of candidates.  This is true even though studies indicate that viewers forget the specific content of the ad soon after watching or hearing it.  No matter – the damage has been done.

A second theory, however, suggests that negative ads activate particular memories or emotions that then become a major part of how an individual assesses a candidate at any particular moment. (Whether they activate hidden cues, or simply lead individuals to weight certain cues more heavily, is not clear.) So, when Iowans see an ad that juxtaposes Newt’s claims that he’s a small government cultural conservative with evidence regarding how he has behaved – lobbying for Freddie Mac, multiple marriages, etc. – it evokes a particular emotion – say, Newt is a serial hypocrite.  And that becomes the primary cue by which individuals decide whether to support Newt or not.   In effect, negative ads prime voters to think of Newt through a particular cognitive frame.

Now, these two theories might seem like academic hair splitting. No matter which theory is right, both suggest the obvious: that the barrage of negative ads altered how Iowans evaluated Newt’s candidacy.  And that is consistent with the more recent political science research that does find substantial effects for some type of advertising. However, the two theories have potentially different implications regarding how Newt might have responded to these ads.  According to Theory I – let’s call it the rational model – the negative advertising barrage has fundamentally changed many Iowans’ views toward Newt; he is now viewed much less positively, making it less likely that caucus-goers will support him come January 3.  In short, the advertising has had a durable, lasting impact on voters’ attitudes toward Newt and in the absence of countervailing evidence; these opinions are not likely to change.

Theory 2, however (we’ll call it the priming model) suggests that some of these advertising effects may be more transitory – that the traits or attributes associated with Newt that the advertisements invoked may fade over time.  From this perspective, in the absence of further priming, voters may fall back to their prior views toward Newt – assuming they have strong prior beliefs.  Interestingly, there is some fascinating research done by the team of Alan Gerber, James Gimpel, Don Green and Daron Shaw on campaign advertising during Rick Perry’s 2006 reelection campaign for Texas governor that suggests that the impact of television advertising, while substantial, is also relatively short-lived, consistent with the priming model.

Extrapolating from that and similar research to Newt’s case in Iowa, however, is fraught with difficulty. For instance, the Perry study looked primarily at positive advertising.  Moreover, there may be differences in how individuals react to symbolic ads designed to evoke a particular emotion versus a more information-based ad that provides new evidence by which to judge a candidate. Simply put, there’s a lot we don’t know about the role that advertising plays in campaigns.

Keeping this uncertainty in mind, one could argue that, following the logic of the priming model, Newt calculated that the initial impact of negative advertising might be substantial, but that it would lessen over time.  From this perspective, staying positive and trying to ride out the initial storm may have seemed quite logical. Of course, he likely underestimated the unprecedented volume and the duration of the negative advertising directed toward him. Still, it provides at least a plausible explanation for Newt’s response to the negative advertising.

If this model is correct, however, it suggests that Newt – and his campaign surrogates – have no time to waste if they want to stop the bleeding.  The priming theory indicates that what really matters for Newt’s fortunes is whose advertising goes on the air last, in the final days before the actual caucus. It is the impact of those ads that voters will bring with them into the voting booth.  This means   that Newt’s shadow SuperPacs may still impact this race if they start airing negative ads against his opponents during these remaining three days before Tuesday’s primary.  For what it’s worth, there’s evidence that Newt’s surrogates are pursuing this strategy, beginning with this negative mailer directed at Romney that went out recently.  Meanwhile, Newsmax is funding  this half-hour infomercial, hosted by Ronald Reagan’s son Michael, to run on Iowa television during the next several days.  At the same time, it appears that other candidates are now beginning to target Romney.

Will this be enough to change the dynamics of this race?  In 2008, fully 30% of Iowans made up their mind regarding who to vote for in the last three days of the campaign.  This suggests we may yet be in for more surprises.

4 comments

  1. Matt, when the folks who sell tooth paste and hamburgers on TV are creating ads that are really changing votes, there is something wrong with the way the democracy is functioning. Maybe P. T. Barnum was right.

    Furthermore, when religious or social issues are the most important factor in who wins in Iowa, and Citizens United permits anonymous “groups” to say almost anything without identifying who is paying for what they are telling the voters on TV and radio, “something is rotten” in the way we pick candidates and elect presidents.

    The law of unintended consequences? We need a 28th amendment.

  2. Jack,

    One thing to keep in mind is that the surveys are showing that even evangelical voters aren’t very concerned with social and religious issues this election cycle – they are more worried about economic ones. the real divide is between those who worry about job creation versus those who focus more on cutting government spending and shrinking the deficit.

    As for citizen’s united, we’ll see if total spending on campaign ads is up over 2008- but I agree with you regarding the transparency aspect. The real anomaly, however, is how much of the negative ad targeted Gingrich. My guess is he is going to start fighting back. And that may be the lesson here – we need more advertising, not less!

  3. Matt. This is a really interesting post. When you were talking about the durability of Gingrich’s polling arc, I wanted to raise the issue of framing effects. Clearly Gingrich is a different candidate when he is the “ideas guy” as opposed to when he is the “serial hypocrite.” The framing that is primed at the forefront of voters minds clearly matters in this regard.

    My read of the media effects literature is consistent with the priming model. Framing effects clearly matter, but their impact is moderated by intervening variables: most importantly education levels and the degree of public deliberation. Education helps determine the strength of the priors that voters have and deliberation serves to moderate and reduce the importance of framing effects. If we think Iowans are a particularly deliberative bunch, then these ads shouldn’t matter too much. If we think most voters are “bowling alone” and divorced from social deliberation, then the ads should have a great impact. Right now it seems to be the latter, although I view the caucus process as somewhat more deliberative, so Gingrich might do better than polling suggests. If the electorate is relatively uneducated and lacking in deliberation, then Gingrich is toast.

    See: Druckman, James N. . 2004. “Political Preference Formation: Competition, Deliberation, and the (Ir)relevance of Framing Effects.” The American Political Science Review 98 (4):671.

  4. Orion,

    Thanks for the thoughtful comment. As I think more about it, I do wonder how useful the Gingrich polling arc (and I note that it seems to have stabilized in recent polls) is for testing our campaign advertising/media effects theory, in that the disparity in advertising was so great – with such a preponderance targeting Gingrich, and only Gingrich – that it may be sui generis. But it will be interesting to see if his advertising counteroffensive, even at this late date, arrests his slide in the polls.

Leave a Reply

Your email address will not be published. Required fields are marked *