Is Negative Campaigning Demobilizing the Middle?

Steve Pearlstein:

There is a vigorous academic debate over whether negative advertising depresses or increases voter turnout. I suspect it does both, depressing turnout among moderates and independents while stimulating it at the ideological extremes. In that process, what has changed is the composition of the turnout rather than its overall level.

From the Lau, Sigelman, and Rovner meta-analysis of studies of negative campaigning:

We did find some support for the idea that whereas negative campaigns stimulate partisans to get out and vote, they are more likely to turn independents off on voting; however, too few studies (only nine) were involved in this hypothesis test to achieve the power necessary for conventional levels of statistical significance (t = 1.6, p < .07, one-tailed).

I’m not sure how many further studies distinguishing independents from partisans (or moderates from ideological “extremists”) have been published since the meta-analysis.  And there is always the question of how large this two-sided effect really is, if it exists.  My guess is that if negative advertising demobilizes the middle but mobilizes the extremes, it does so mainly at the margins.


7 Responses to Is Negative Campaigning Demobilizing the Middle?

  1. Andrew Gelman April 23, 2012 at 4:57 pm #


    I don’t understand your last sentence. What do you mean by “mainly at the margins”?

    • John Sides April 23, 2012 at 6:41 pm #

      All I meant was that the role of negative campaigning has relatively small effects. Not that such a phrase is much more specific!

  2. David Lauter April 23, 2012 at 7:11 pm #

    The weakness of Pearlstein’s thesis is that overall turnout, as least at the presidential level, has been up, not down. That seems hard to square with a line of argument claiming that tons of money leads to huge amounts of advertising, which mobilizes the base while turning off the center. In Ohio in 2004, it’s true that the Bush campaign succeeded in turning out a huge number of Republican voters. But so did the Democrats. It’s not true that GOP negative advertising suppressed Democratic turnout — at least not in that state.

  3. Rhonda April 23, 2012 at 7:39 pm #

    Prof. Sides: I don’t understand how your response to the Pearlstein quote addresses his main. Is there a vigorous academic debate about the effect of negative advertising on turnout? Your quote and discussion of Lau et al doesn’t really address this issue. Or maybe I’m slow. Thanks.

    • John Sides April 23, 2012 at 7:59 pm #

      Rhonda: Sorry, I’ve written so much on this that I unfairly assumed that readers would have seen it. Happy to elaborate. I think the preponderance of evidence is that negative advertising does not demobilize voters as a whole. It may, as Perlstein suggests, demobilize independents or moderates, but actually mobilize partisans — two effects that may cancel each other out when you look at the electorate as a whole. The quote from Lau et al. is meant to address Perlstein’s specific thesis. The Lau et al. meta-analysis does speak to the broader debate as well.

      • Rhonda April 24, 2012 at 2:57 pm #

        Helpful. Thank you.

  4. carl pinkele April 24, 2012 at 12:39 pm #

    Perlstein’s is provocative and that in itself is a big plus, add to that it appeared in a major public newspaper. In some earlier unpublished/unfinished papers some colleagues and I took a stab at deconstructing the reactions to negative adds and the whole politics is messy business. We tentatively found that gender, ideology and age were positively associated with reactions to negative campaigning AND broader attitudes about the nature of politics. In brief, conservatives who were male tended to like negative campaigns as did conservative women; the more liberal — including moderate people were, especially women — the more likely they were to self-demobilize when confronted by negative campaigns. So, negative campaigns, etc. do not fall evenly upon the political attitudes landscape.