Editor’s Note: This post was written by Jack Massaroni.

Political scientists have devoted scores of articles and books to the effects of negative advertising. Attack ads have seen a consistent rise in popularity over the past few years, culminating in a rate of nearly 65% of ads in the 2012 election. The effects of attack ads on supporters have been well documented, and generally speaking, they can have real effects. According to Richard Lau, Lee Sigelman, and Ivy Brown Rovner, political scientists are mixed in their assessments of the effects of negative advertising, but they do appear to have some effect on the target of the attacks. However, viewers often penalize the sponsors of such ads, so a sponsoring candidate often receives no net gain from releasing an attack ad. That is, while the target of the ad does lose some support, the sponsor of the ad can lose even more support because they chose to “go negative.” If these repercussions are real, why would any candidate even go near attack ads? In 2016, outside groups (such as super PACs) may provide valuable cover for candidates. Because the attack ad is not directly sponsored by a candidate, the backlash effect where the sponsoring candidate loses support may not exist. This means candidates reap the benefits of attack ads but don’t necessarily pay the price.

The 2016 cycle has been a close contest, and candidates have spent immense resources on television ads. As Donald Trump became the GOP front-runner, more and more attack ads emerged. This is in line with research predicting that when trailing, campaigns are more likely to “go negative,” or engage in adverse campaigning. However, the question remains, are these ads working? In early March, over 75% of attack ads aired attacked the GOP front-runner. In this case, it would appear that these ads have neither their desired effect nor has Trump been affected in the way research would predict. It appears that Trump voters are immune to attacks on their candidate. Nearly 40% of those who voted for Trump in the Iowa caucuses had decided on Trump more than a month in advance. In the New Hampshire primary and Nevada caucus, that number is closer to 60%. The mounting number of attack ads leading up to the contests did not change their minds. Attack ads on Trump appear not to sway his loyal supporters, and it seems that most Trump supporters are unwaveringly loyal.

One explanation for this phenomenon is that the one trait that distinguishes Trump supporters from other Republicans is their sense of anti-elitism. This anti-elite sentiment could cause Trump supporters to disregard attack ads as illegitimate. Attack ads from super PACs are exactly the kind of elitism Trump supporters hate. One of Trump’s big selling points early in the campaign was his “self-funded” campaign. While not necessarily true (he is loaning his campaign money, rather than donating), it conveyed a message that Trump could not be bought. However, the effect of these ads on non-Trump supporters has been immense. The attack ads have definitely affected non-Trump supporters, with one study showing that 40% of those polled felt distinctly worse about Trump after watching one ad. The same study shows that Trump supporters felt better about Trump after viewing the ads. The long-term effects of the attacks on Trump have yet to become clear and it remains to be seen whether these attacks will hurt him in the general election.

With research showing that voters tend not to like attack ads, Bernie Sanders has tried his hardest to put out more positive messages or ads closely tied to his policies. In January he ran an “attack ad” that alluded to Hillary Clinton. However, she was not referenced to by name but supporters and reporters knew exactly who Sanders was referring to. This approach may have helped him; Sanders claimed that he would not attack his opponents and voters subsequently judged him a more honest candidate. However, in some states, this honesty was deemed not as important. In Iowa and Nevada, as well as other states that voted for Hillary Clinton, voters ranked electability before honesty in terms of importance. After his losses, Sanders did attack Clinton for a few weeks to remind voters of the differences between him and Secretary Clinton. However, his campaign has touted the fact that Sanders has not once put out an attack ad and all of his ads have had neither Clinton’s face nor name appearing in them. It is only in his speeches that he has discussed her directly, although he has reminded the crowds at his rallies to respect her and quieted their booing once.

The attack ads of 2016 tell a story; the meanest ads show who the candidates fear. Even during the nomination contests, there have been attack ads from Republicans against Democrats, namely Clinton, by super PACs. While the Republican Party is trying to attack Clinton and decrease public opinion of her, their ads have had the opposite effect for Sanders. Sanders is viewed as the weakest competitor for the Republican candidate, therefore a wealthy donor is using ads to raise the support of Sanders so that it will be a Republican landslide come the general election. This shows that candidates and their supporters are trying to get an early start on the general election and people can expect more attack ads as the fall approaches.