Did the 2008 John McCain campaign deliberately manipulate the color of Barack Obama’s skin in TV ads, trying to evoke racial biases? Several major media outlets have made this allegation in stories about our recent research. If true, this would be further evidence that American political campaigns have, as some observers allege, long been making subtle but intentional appeals to racism.
But what our study found was more nuanced. We’d like to explain.
President Obama’s face did indeed appear darker in the most negative attack ads of the 2008 campaign. And subtle differences in skin tone can bring to mind negative racial stereotypes among some viewers.
But we do not know the McCain campaign’s intentions.
Possible explanations for the different shades of Obama
Dark imagery is commonly used to make negative ads feel more foreboding, regardless of a candidate’s race. Images of Obama may have been digitally altered as a part of that more general effort to darken ads. Or the McCain campaign might not have manipulated the images at all, but rather chosen darker images from the many existing pictures of Obama.
Or the campaign might have chosen to use darker images unconsciously, as a product of implicit cognition — not intentionally to appeal to bias, but because of an impulse to associate darker skin with negativity, and even criminality, in the process of producing a sinister ad without thinking it through.
Our work still suggests that darker portrayals of black candidates can have harmful consequences, whether deliberate or not. Darker depictions of candidates can bring to mind negative racial stereotypes.
How we studied darkness in campaign ads
We analyzed Obama’s skin complexion in 126 political ads that aired during the 2008 presidential campaign and found that darker images of Obama were more common in the most negative attacks ads, especially in those linking Obama to crime. We also found that the darkest images were more likely to air more frequently as the election approached.
Does it matter that darker images appeared in negative ads? Past studies have shown that in the lab, negative stereotypes come to mind when you show someone pictures of unfamiliar African Americans with darker skin. But it’s quite different to show, as we have, that darkening the complexion of a well-known and accomplished political leader can bring to mind the most negative stereotypes about blacks.
We set out to explore this effect with a simple experiment. First, we showed people either a lighter or darker image of Obama. Afterward, we asked those people to fill in the blanks to complete partial words, and we counted how often they provided stereotype-consistent completions of the following prompts: L A _ _ (LAZY); D _ _ _ Y (DIRTY); and _ _ O R (POOR).
People who had seen the darker image were on average 36 percent more likely to use these stereotype-consistent words than were people who saw the lighter image. (In this part of the experiment, we weren’t asking them to describe Obama; we were just asking them to complete those prompts).
In other words, darkening a candidate’s complexion could trigger racial stereotypes, wherever candidates of color are running for office. The “dark skin penalty” holds true across a variety of racial and ethnic groups.
How our research can be used
Even if darker images are chosen to create a foreboding atmosphere rather than in deliberate appeals to racial bias, this research shows one systematic way that associations between darkness and negativity continue to haunt social and political life in America.
One way to address this phenomenon is to track advertisements throughout the campaign season, a project undertaken by organizations, such as the Political Communication Lab at Stanford (the data source for this study) and the Campaign Media Analysis Group (CMAG). Researchers who want to follow up on our work might use these sources to analyze campaign advertisements — including skin tone — in 2016.
Solomon Messing is director of Data Labs at the Pew Research Center. Ethan Plaut is a postdoctoral fellow at Stanford University. This research was conducted when they were PhD candidates in communication at Stanford University.