An IJJ Special Report
Racial IDs in Crime Stories: When Are They Fit to Print?
By Sally Lehrman, IJJ Senior Fellow

In a suspected home invasion robbery last fall in Sacramento, two attackers gunned down Sean Aquitania Sr. As they took off, they stopped briefly to kill the young father’s 7-month-old son. Outraged residents heard descriptions of the attackers on television and on the radio: The light-skinned Hispanic man wore a white shirt with green sleeves and a green Oakland A’s cap; his black companion, a multi-colored, hooded sweatshirt.

At the Sacramento Bee, an editor following guidelines on relevance trimmed out all the racial references. Some readers fumed at what they saw as political correctness gone nuts. “We got pounded,” says Ricardo Sandoval, assistant metro editor at the Bee. Readers, he recalled, wanted to know, “Why are you guys hanging back?”

Sally Lehrman IJJ Senior Fellow

Prompted by pressure from competition and their own audiences, many news outlets are reconsidering longstanding policies against naming race in crime suspect stories. The Raleigh News & Observer recently loosened requirements that allowed for racial descriptions only in violent crime stories. The Arlington Heights (Ill.) Daily Herald began contemplating a change after reporters argued that cutting race out made a description worthless. The Sacramento Bee has convened a review committee.

At first glance, the impetus to include race when describing an attacker or burglar seems reasonable. It’s a journalist’s duty to provide immediate, relevant information that might protect public safety. Whether racial identifiers actually serve this purpose, though, is worth questioning. How trustworthy are eyewitness or police accounts of race? And when they’re wrong, what’s the damage?

In this country, people seem at ease in classifying a person by race. But a closer look finds our definitions a bit pliable and hazy, changing with geography and time. The history of the Census reflects these transitions – identifying people mainly according to their free or enslaved status early on, later including a mix of national origins with what we now would call “race.” Today, our views are shifting once again.

Until recently, race was one of the more trustworthy recollections by witnesses, according to Christian Meissner, an experimental psychologist at the University of Texas, El Paso, who studies identification of suspects. But immigration and internal migration have touched every pocket of the nation and the multiracial population is increasing. With this “increasing variability of raceness,” as Meissner put it, our terminology has become less reliable.

Certain markers of race, researchers recently discovered, also can affect what police or witnesses “see.” A team from the University of Texas, El Paso, and the University of Northern Iowa showed participants two identical faces with the same, racially ambiguous features. When they added a stereotypical “Afro” hairstyle, viewers perceived darker skin and different facial features than the same illustration with a slicked-back “Hispanic” hairstyle.

Overwhelmingly, the participants put each version into a racial category consistent with the hairstyle.
Most people think that memory is like a video recording, stable and recalled at will. Faces are the easiest to remember, the popular theory goes. In fact, most of us are terrible at recording and recalling events, explains a summary of recent eyewitness research in the journal Psychology, Public Policy and Law. It’s easy to imagine how lighting, setting, and neighborhood might influence racial perceptions. In addition, we narrow our attention to what matters in the moment. Then, along with other distortions, we tend to remember events in a way that makes personal “sense” to us and that is consistent with our expectations. Many times, we’re wrong.

How trustworthy are eyewitness or police accounts of race? And when they’re wrong, what’s the damage?

Mistakes like this by eyewitnesses, of course, can send police -- and news audiences -- off on the proverbial wild goose chase. Based on these findings and cases in which searches have veered wildly off course, some argue that law enforcement should get rid of racial identifiers altogether in suspect descriptions. In an April 2003 article in the Columbia Law Review, Essay and Review Editor Bela August Walker condemns race as useless because of “a dual problem of overbreadth and inappropriate narrowness.”

In Santa Monica, Calif., for instance, police investigating a series of 19 armed grocery store robberies went on the hunt for “two African American males, aged 20-30, one tall (6’ to 6’2”) and 150-170 pounds, and the other short (5’5” – 5’7”), and about 170-190 pounds.” Acting on the description, police officers stopped a magazine photo editor and a bank analyst one night as the two African American men from New York – one short, one tall—pulled into the parking garage of their hotel. The evening they had spent at an L.A. Dodgers game ended in humiliation as about seven officers with a police dog surrounded them, ordered them out of their car at gunpoint, hand-cuffed and frisked them, and searched their car.

The officers jumped to conclusions based on a summary that was, in fact, “exceedingly vague,” wrote Ninth Circuit Court of Appeals Judge Stephen Reinhardt, upholding a ruling that the stop violated the men’s civil rights. There were no key details, such as facial hair or scars. “Other equally general descriptions could serve as the basis for similar demeaning treatment of many other African-Americans,” Reinhardt wrote.

Reinhardt pointed to evidence that police often unconsciously lower their standards for satisfying even a detailed description when a minority person is involved. “The burden of aggressive and intrusive police action falls disproportionately on African-American, and sometimes Latino, males,” he wrote.

Race and gender do tend to trump any other descriptors in our minds, UT’s Meissner agrees. Rather than heightening awareness, looking for “race” can overwhelm our ability to perceive finer distinctions. These two “tend to drive your expectations of lots of things about that face, interpretations about behavior,” he explains.

When we use race in crime descriptions, then, there’s a high probability that we’re repeating inaccuracies. Furthermore, we may inadvertently help replicate the disparities in our criminal justice system that Reinhardt laments. Just as dangerously, cognitive research has found, we contribute to social attitudes and perceptions that it’s unlikely we would endorse.

Like the hair styles in drawings found to call up racial groupings, the words and context journalists choose can conjure unconscious attitudes about race in our audiences. Social scientists have explored these implicit reactions in depth and continue to study their implications.

In an analysis by collaborators at UCLA and Stanford University, researchers studied the effects of local crime news. In studios set up at two malls and UCLA, the team showed 2,331 Los Angeles volunteers identical news stories, except that in some cases, the perpetrator was identified as an African American male; in another he was a white male, and in a third set, no identification was offered. A control group didn’t see any crime news story.

For another discussion of police descriptions and race, see Criminal Justice Journalists.

More than half the time, when faced with a crime story without a guilty party, viewers made one up. That person, 44 percent of the time, was black. When the perpetrator was white, one-quarter of white study participants didn’t remember any suspect at all.

Political scientists Franklin G. Gilliam and Shanto Iyengar probed more deeply into racial attitudes and the ways they were affected by crime news. All of the crime news stories influenced racial attitudes, with white participants showing about a 10 percent increase in what researchers called “new racism.” They defined this as a belief that discrimination no longer occurs, that black people make illegitimate demands and don’t believe in the American values of hard work and self-reliance, and the like.

When a black perpetrator was shown, even “old-fashioned” racism seemed to increase among white viewers. Furthermore, watching crime news with a black perpetrator increased white viewers’ preference for punitive remedies. This held up as well in a random sample survey of white Los Angeles residents. Our choices in crime reporting, Gilliam and Iyengar conclude, not only deeply influence views about race, but also can racialize public policy more generally.

The power of racial descriptions, according to these and other social scientists’ research, reaches far beyond normal newsroom considerations. The Associated Press is adhering to a strict policy on racial descriptions, says Brian Schwaner, news editor for Louisiana. “When you can’t provide enough detail that really helps you identify a suspect, why put it in?’ he asks. At the Arlington Heights paper, editors are considering policies ranging from one that would use any “reliable” description, including race, to another that would omit the entire identification when it was vague, such as “a white man, 30- to 40-years old, medium build.”

Context, the Bee’s Sandoval has concluded, is key. Is the description detailed and useful? Is it relevant and likely to be accurate? Does it feed preconceptions or does it add information?

Considering the potent social environment we report and write into, the answers to these questions are not straightforward or simple.