Since Michael Brown’s fatal shooting on August 9, news coverage of the event and its aftermath has been followed by a second wave of analyses scrutinizing those initial stories for racial bias. When news outlets ran a particularly ominous photograph of Brown, wearing a red jersey, fingers splayed in possible gang signal (friends say it was his characteristic peace sign), African Americans took en masse to Twitter, using the hashtag #iftheygunnedmedown to post personal photographs—one sweet and one sinister—to show how the media skews narratives of black men.

The New York Times, for one, received pushback after publishing a complicated profile of Brown’s spirited childhood describing him as “no angel,” laid out in print and online beside what many perceived to be a gentler profile of the cop who shot him. Public Editor Margaret Sullivan described the use of “no angel” as “a blunder,” as well as the decision to pair the two profiles in a way that “seemed to inappropriately equate the two people.”

Would Brown’s history of drug use and petty theft have been less prominent in coverage of his slaying if he had been white? These accusations hit at the same profound question raised in the case: If Brown had not been black, would he have been shot in the first place?

While charges of racial bias in the media have, at the moment, been limited to discussion of how Brown and other black victims are portrayed, broader issues of bias are revealed when looking at which crimes journalists choose as newsworthy. In a survey of broadcast news published Tuesday, Media Matters for America found that television coverage crime suspects’ race doesn’t match up to the raw data of who is actually arrested—black suspects receive disproportionate coverage for their alleged crimes.

Researchers for the group watched New York newscasts on WCBS, WNBC, WABC, and WNYW, counting the percentage of suspects revealed as African American, either by a photo aired on the newscast or a verbal description. Over a three-month period, the parade of potential perps were overwhelmingly black: Eighty percent of theft suspects were African American, as were 73 percent of assault suspects and 68 percent of murder suspects whose cases received airtime. But when Media Matters compared the numbers to arrest statistics from the NYPD, the racial breakdown showed a much lower percentage of black suspects. “African-American suspects were arrested in 54 percent of murders, 55 percent of thefts, and 49 percent of assaults,” they wrote.

The comparisons on the study weren’t perfect: Media Matters could only get aggregate data from the last four years of NYPD arrests, while they surveyed only three months of New York news. Still, the results fit into a longstanding pattern of the media covering black suspects more often—and often more harshly—than white suspects of similar crimes.

In a study of the Chicago broadcast media, a research team found that black defendants were more likely than defendants of other races to be shown through a mugshot rather than a personal picture or none at all. Another study of television coverage found black suspects are twice as likely as white suspects to be shown on camera under police restraint. While it’s difficult to pinpoint whether a particular suspect is being covered more harshly because of their race, taken in tandem this data points to a dangerous precedent: Black men are easily perceived as criminals, disproportionately to the rate they may be committing crimes.

It’s a similar framing to “Missing White Girl Syndrome,” a name coined to reflect the deluge of coverage when a young, affluent, white female goes missing—and the dearth of coverage when children of color disappear. Entities like the “Black and Missing Foundation” and shows like TVOne’s Find Our Missing are attempts to fill in the gap publicizing missing children who are not white. Scanning the faces of the missing children on these sites comes closer to the reality of the racial breakdown of missing persons: About 34 percent of missing persons overall, and 37 percent of missing minors, are black, according to FBI statistics from 2013. Who media coverage chooses to cast as victims, and what victims—like Brown—don’t fit neatly into the role, has a powerful effect on the mindset of the public who watch these reports and take with them a subliminal idea of what someone who might be victimized looks like.

Alexis Sobel Fitts is a senior writer at CJR. Follow her on Twitter at @fittsofalexis.