Black women were 17 percent more likely to get pulled over by a patrol officer before sunset than at night in the Raleigh Police Department’s Southwest District, according to a new study of citywide traffic stops between 2010 and 2018—a measure considered an indicator of possible racial bias. 

Even so, the Research Triangle Institute study found “no evidence of disproportionality at the city level.” The disparities came only with that specific confluence of variables: women in the Southwest District (one of eleven RPD districts) who were stopped by patrol (as opposed to non-patrol) officers. 

The RPD presented the report’s findings last Thursday night in a Raleigh Convention Center conference room, where the sixty or so officers outnumbered members of the public two-to-one. Police chief Cassandra Deck-Brown, who has argued that her department does not need a civilian oversight board, told those in attendance that the RPD “had quite a bit of homework to do,” but “I do believe [the report is] a story worth telling and a story worth discussing.”

RTI research analyst Brian Aagaard explained that, while the RPD made more than a half-million traffic stops during the study’s eight-year window, RTI only examined the seventy-two thousand that took place between 5:30 p.m. and 9:00 p.m. 

The “veil of darkness” approach limits the analysis to periods when officers would have more ease (before sunset) and difficulty (at night) visually determining the race of a driver when deciding whether to pull them over. This method has been used in studies of traffic-stop disproportionality in cities like San Diego, Minneapolis, and Durham.

The idea is that if black drivers were pulled over disproportionately when it was sunny but not at night, the cops are probably biased. If the rates don’t change regardless of light, the cops probably aren’t biased. 

And except for black women in the Southwest District, that’s what RTI found.

But there’s evidence that the primary assumption behind the veil-of-darkness methodology—the driver’s race is difficult to discern at night—may not always hold up. 

A 2015 study of traffic stops in Syracuse, New York, that appeared in the Review of Economics and Statistics used the same veil-of-darkness approach as RTI’s study but refined its definition of darkness to exclude areas lit by streetlights. 

After all, if streetlights show officers the race of a driver, that undermines the test’s fundamental premise. 

When using a more refined analysis that took streetlights into account, the Syracuse study found evidence of racial profiling in traffic stops between 2006 and 2009; ignoring the presence of streetlights in that same data set, however, yielded no evidence of profiling.

“How this difference arises is hard to say,” the study’s authors write. “It could be due to differential police behaviors in poorly lit areas, but it could be due to differential driving behaviors or some other unobservable features of poorly-lit areas.” 

The bottom line: “Accounting for heterogeneity in nighttime ambient lighting may be important for the veil test.”

RTI didn’t factor in the presence of streetlights in its report.

In an interview, Aagaard acknowledges that ignoring streetlights could obscure bias, but he also suggests that ambient nighttime light could actually impede officers’ ability to determine a driver’s race because of the glare on windshields and other clear surfaces.

“It’s something that warrants further examination,” Aagaard says.

After the initial slideshow by RTI representatives Thursday evening, no one mentioned the disproportionate level of traffic stops of black women before sunset in the Southwest District. 

Comment on this story at 

Support independent local journalism. Join the INDY Press Club to help us keep fearless watchdog reporting and essential arts and culture coverage viable in the Triangle.