A still image from the Willie Horton ad (Source: National Political Action Committee)
A still image from the Willie Horton ad. (National Political Action Committee)

The specter of William R. Horton looms over politics. A convicted murderer, he was temporarily released from a Massachusetts prison in April 1987 as part of the state’s weekend furlough program. Horton, a black man, then raped a white woman and brutally assaulted her white fiance.

When then-Massachusetts Gov. Michael Dukakis became the Democratic presidential nominee in 1988, Horton was the centerpiece of attacks on Dukakis’s record on crime. Called “Willie” by the campaign of Republican candidate George H.W. Bush, Horton was most famously depicted in an ad called “Weekend Passes,” produced by an independent group supporting Bush.

The impact of the Willie Horton ad has been described as “devastating to Dukakis.” In a profile of Larry McCarthy, the ad’s creator, the New Yorker’s Jane Mayer described the ad as “the political equivalent of an improvised explosive device, demolishing the electoral hopes of Dukakis.” Dukakis himself now thinks that he was “getting killed.”

As of this week, the Willie Horton ad has surfaced again. According to a New York Times story on Monday, congressional efforts to reform the criminal justice system are threatened by the possibility that supporters could be attacked just as Dukakis was. As one advocate of criminal justice reform, Sen. Richard J. Durbin (D-Ill.), put it: “The Willie Horton story is still part of the legend and lore of American politics.”

But like a lot of legends, the prevailing characterization of the Willie Horton ad — that it cost Dukakis the election — has one important shortcoming: It is largely wrong.

The real story of the Willie Horton and the 1988 campaign is far more complicated. It is about how few people likely saw the ad itself, how it was actually news coverage that brought Horton into Americans’ living rooms and how Horton’s possible impact on voters disappeared before Election Day.

Bush began mentioning Horton in speeches in June 1988, well before the ad aired.  But this generated little news coverage of Horton. In an extensive study of the 1988 campaign, political scientist Tali Mendelberg identified only 25 stories in major newspapers between June and the beginning of October — an average of less than one story a week. Polls from this time period show notable movements only around the period of the two party conventions.


The Republican convention was particularly important. It took Bush from a double-digit deficit at the beginning of August to a narrow lead at the end of that month. After that, Bush never lost the lead. This is something that conventional accounts of the Willie Horton ad rarely mention: Dukakis was already behind when the ad appeared.

The aid first aired on Sept. 7, but could not have had much direct impact. It aired only on cable television networks whose market share amounted to less than 1 percent. On Oct. 3, “Weekend Passes” was pulled and another ad that focused on Dukakis’s record on crime began to air more widely, but this ad did not mention Horton. A third ad, produced by the same independent group, began to air later in October. It featured Horton’s victims but did not include Horton’s image.

In short, not the infamous Willie Horton ad nor any of its successors both included Horton’s image— the very thing that played on racial fears — and could have been seen by large numbers of Americans.

Mendelberg argues that Americans were actually more likely to learn about Horton from news coverage. She documents 19 separate evening news broadcasts between Oct.  7 and Election Day that mentioned Horton, often including his photograph, photos of his victims and/or images from the “Weekend Passes” ad.

On its face, this news coverage appeared influential. Drawing on a large survey conducted during the two months before the election, Mendelberg compared voters interviewed before Oct. 3, when the Bush campaign and news coverage began to emphasize Horton more often, to those interviewed between Oct. 3 and 21.

In this second time period, Bush’s popularity was sharply higher among people who already harbored unfavorable feelings about blacks. And during roughly the same period, Bush’s margin over Dukakis in news media polls more than doubled — from about five points to about 11.

If we assume that this shift is attributable solely to stories about Horton — and not to the debates, other ads or news coverage that aired during this time — then the conventional wisdom seems born out. A racially coded message appeared to benefit Bush by activating some people’s prejudices.

But what happened after Oct. 21? On that day, Jesse Jackson said on CBS News what few had said publicly to that point: that using Horton’s case as a campaign message actually played on racial fears. In so doing, Mendelberg argues, Jackson took an issue that had been only implicitly about race and made it explicit. Subsequent news stories began to mention race and racism when they discussed Horton and the campaign.

Most crucially, voters, now aware that this campaign message potentially violated norms against racism, no longer reacted in the same way. Among voters interviewed after Jackson’s statement —between Oct. 22 and Election Day in Mendelberg’s survey — racial prejudice ceased to predict attitudes toward Bush and Dukakis. Jackson’s argument had neutralized the issue.

In other words, a series of attacks on Dukakis’s record on crime — though not the Willie Horton ad per se — coincided with an increase in Bush’s lead over Dukakis in the first part of October, particularly among white voters with negative attitudes toward African Americans. It is difficult to know whether this increase is due to those attacks or to a variety of other things that occurred in this time period. But the apparent effect of these attacks on Dukakis disappeared after the use of Horton as a campaign message was framed as a racially coded appeal.

Of course, there is always the counterfactual: Could Dukakis have come from behind in October and won the election if not for the attacks on his record on crime? We do not and cannot know. Regardless, the story is far more complicated than phrases like “improvised explosive device” would imply.

The true story of the Willie Horton ad does not mean that race and racism have no potency in American politics, or that racially coded attacks can always and easily be neutralized. For example, white Americans’ views of blacks are related to their views of Barack Obama in myriad ways — to their views of Obama himself, the policies he champions, and even his dog. Race isn’t the only factor affecting views of Obama, or views of our criminal justice system, but it is an important one.

But in the case of Willie Horton, the actual impact of race has gotten lost in the usual mythology that surrounds elections, whereby shadowy and shrewd campaign “gurus” produce ads fine-tuned to activate our worst impulses in only thirty seconds — thereby determining the outcome of the election.

Dukakis calls his slow response to the Willie Horton ad “the biggest mistake of my political career.” He shouldn’t be so hard on himself. And politicians today should look skeptically on the conventional wisdom that has grown up around this ad.