Everyone “knows” that presidential campaigns have grown increasingly uncivil over the years. But nowhere in the vast research literature on negative campaigning can one find hard evidence that what everyone knows is true. This is because researchers (including me) have focused single-mindedly on “negative” campaign messages, i.e., messages critical of the opposition. Such ads clearly have been on the rise, but they aren’t necessarily uncivil. As John Geer, among others, has argued, there’s nothing objectionable about negative campaigning per se; indeed, criticizing one’s opponent’s qualifications and issue positions is a vital function of political campaigning. (On the utility of campaign ads, see John Sides’ post about another recent study.)
Have presidential campaign ads actually become less civil? My co-blogger David Park and I have just completed a paper in which we try to answer this question. For each of 954 candidate- or party-sponsored ads in the 1960-2000 presidential campaigns, we calculated scores on two related but distinct dimensions of incivility: unpleasantness and nastiness (the latter being, in essence, a more “active” version of the former). To tap these two dimensions, we used the Dictionary of Affect in Language computer program, which scores the emotional tone of a text.
What we find is not at all what the conventional wisdom about the tone of presidential campaigns would lead one to expect. True, the language in these ads actually has grown more unpleasant over the years. In the early days, Republican ads were more unpleasant than Democratic ones in the early days. By the end of the century, though, the GOP’s ads were no more unpleasant than they had been during the 1950s, but the Democrats’ ads had grown progressively more unpleasant. Thus, by the end of the century the Democrats’ unpleasantness surpassed that of the Republicans. Even more strikingly, on the nastiness dimension no upward shift was perceptible for either party; nor, for that matter, was there or is there now any real difference between the two parties.
If these findings are to be believed—and they are based on a measurement technique that has proven valid and reliable in a wide variety of contexts—then what everyone “knows” about campaigns seems to be in need of some fundamental rethinking. To be sure, the Republicans have aired some extremely memorable attack ads, but by our reckoning it is the Democrats who have been responsible for the increasingly unpleasant tone of presidential campaign ads. And the nastiness of the two parties’ ads at the end of the century was no more pronounced than had been the case at midcentury. Finally, comparison of the overall unpleasantness and nastiness scores for these campaigns indicates that they are very similar to benchmark scores that have been compiled for contemporary American English texts more generally. So if we are currently in the throes of a political “incivility crisis,” perhaps the finger of guilt should be pointed at American culture writ large rather than at the culture’s political outcroppings.
For a copy of our paper, click here.