A new analysis of Durham police statistics found that the proportion of black drivers stopped in recent years rose significantly during daylight hours when the race of the driver was presumably more apparent.
The analysis compared 151,701 traffic stops from January 2010 through October 2015 and found that the proportion of black drivers pulled over during daylight hours was 12 percent higher than the proportion during nighttime stops.
Among male drivers only, the odds that a driver was black were 20 percent higher when stopped during daylight than when stopped at night, according to the study by RTI International.
But the study, released Thursday, also found the disparity had largely disappeared by the last two years of the study.
“This improvement may have been the result of changes in Durham Police Department training or policies,” said Travis Taniguchi, a research criminologist at RTI.
The study also found no racial disparity among female drivers stopped by police.
During a news conference Wednesday afternoon at RTI International, the city’s Interim Police Chief, Larry Smith, listed some of the changes the department had deployed to address the issue. They included a greater emphasis in de-escalating potentially volatile encounters between officers and residents, and obtaining a written consent from a motorist before searching a stopped vehicle. The department has also begun to look more closely at any officer who stops large percentages of minority drivers. An officer with 25 or more stops is flagged if 75 percent of those stops involved people of color.
“This will help us pinpoint where the help is needed,” said Smith, adding that the changes are part of “ongoing strategies, training and policy adjustments.”
The RTI study reviewed traffic stops between the 5:30 and 9:30 p.m. “veil of darkness” hours. It used statistical modeling to predict how many black drivers would be stopped compared with white drivers. Absent bias, one would expect the same ratio during daylight as during the night.
But the ratio of black drivers to white drivers stopped during daylight – when an officer was more likely to see the driver’s race – was higher than the modeling predicted, Taniguchi said.
“The difference between day and night is what’s indicative of bias,” he said.
“The bias appears to be apparent,” Durham Mayor Bill Bell said after the news conference. “You can't deny the data.”
Taniguchi said this comparison provides more reliable data than a previous study by UNC political science professor Frank Baumgartner that compared the percentage of black drivers stopped to the percentage of black Durham residents.
Racial differences in traffic stops and searches were among the complaints that led Bell to have a city advisory board look into police practices two years ago. After holding numerous public meetings, the city’s Human Relations Commission found that a pattern of racial bias and profiling existed within the department.
Former Police Chief Jose Lopez rebutted the profiling accusation, saying the higher number of black and Latino drivers stopped by police reflected high-crime areas where residents were asking police for help. Lopez, who led the department for eight years, retired last year after City Manager Tom Bonfield asked him to resign or be dismissed.
The RTI study found greatest racial “disproportionality” when drivers were stopped by the police department’s HEAT or High Enforcement Abatement Team, which focuses on drugs, vice and gang violence. It found no such pattern within the traffic unit, which stops drivers for offenses such as speeding and driving while impaired.
The study found no disproportionate minority contact in Raleigh, Fayetteville or Greensboro traffic-stop statistics. “No evidence of racial bias was found in these cities,” it said.
The study also did not address searches during traffic stops, where Durham police statistics also showed black and Latinos drivers more likely to have their vehicles searched.
In October 2014, the city began requiring written permission for “consent” searches, in which an officer lacks probable cause or evidence to suggest a crime has been committed.
In July 2015, a study found police were searching fewer motorists by consent since the new policy began. Meanwhile, probable-cause searches, which do not require permission, were up.
Baumgartner, who did the analysis along with two graduate students, said there appeared to be a “substitution effect,” a claim Lopez denied.
The latest study was based on data the Police Department gave RTI in an effort to be transparent and to see what impact its police were having.
“It’s essential that we get an objective view of our operations,” Smith, the interim chief, said in a statement, “and in turn be willing to not only accept the findings, but continue to work toward putting the necessary tools in place to correct the issues this analysis revealed – and ensure that bias of any kind is never a part of police operations.”
Smith said before arriving at the early afternoon news conference, members of the police department command staff met with community groups and organizations that have in the past publicly expressed concern about the racial disparities during traffic stops in Durham.
“We heard your reports,” Smith said he told the community representatives. “Today is evidence of that.”
Mark Schultz: 919-829-8950
Some changes implemented by Durham police
▪ HEAT, Patrol and Traffic Services units have been outfitted with in-car cameras to capture interactions during stops.
▪ The department conducts semi-annual traffic stop data reviews, analyzing data pertaining to the initial purpose of a traffic stop, the enforcement action and the potential for being searched during a stop (DPD traffic stop data is also reviewed by the state of North Carolina).
▪ Monthly reviews of in-car camera footage are done by division commanders.
▪ Complaints resulting from traffic stops are investigated by the Professional Standards Division in an attempt to identify patterns.
▪ Signed written forms are required for all consent searches during traffic stops.
▪ Additional training of officers to prevent bias