Texas police data of little use
Bland Act fails to fix 20-year effort on racial profiling
Traffic stops are how most citizens encounter the police, so over the past 20 years Texas law enforcement reform advocates have pushed to analyze a basic measure of the fairness of their work: whether officers treat black and white motorists differently during traffic stops.
Researchers say the information required to do that isn’t complicated. A department needs to report why each traffic stop was made, whether the driver was searched and what if anything was found — separated out by the race of motorists. The most recent update to the effort, 2017’s Sandra Bland Act, was supposed to do exactly that.
But the Texas Commission on Law Enforcement, the agency in charge of collecting the information, left out one part — the driver’s race.
“It’s too bad we lost out on the racial part of the racial profiling bill,” said Scott Henson, director of Just Liberty and a longtime Texas criminal justice analyst.
The result: After two decades and three separate laws, efforts to identify which of the state’s nearly 2,000 law enforcement agencies have glaring racial disparities in their policing have largely failed, yielding only a patchwork of numbers with widely varying degrees of useful
ness.
Experts say the statewide information gathered by TCOLE provides little utility to researchers, elected officials or the public.
“That’s like a crime against public records requests,” said Frank Baumgartner, a political science professor at the University of North Carolina and author of numerous racial profiling studies.
How well agencies track and analyze their policing practices by race is much more than a matter of good public record-keeping. The death of George Floyd after a Minneapolis police officer knelt on his neck for nearly nine minutes has prompted yet another national soul-search into the question of how and why law enforcement agencies continue to treat minorities differently than whites.
Studies across the country have shown that police stop, search, arrest, use force on and kill black people in numbers disproportionate to their population representation. Police say numbers alone don’t tell the whole story, and often there are sound law enforcement explanations behind what appear to be biased behavior. Officers may be directed to more aggressively patrol high-crime areas where minorities live in greater concentration, for example.
Researchers also agree aggregate numbers showing unbalanced treatment by race in a department don’t necessarily prove bias. Yet at the very least, they say, identifying which agencies seem to be unfairly targeting minority motorists is a crucial starting point.
“Changes depend on somebody shining a clear light on the patterns in a way that helps the broader public to understand the patterns and their significance,” said Charles Epp, a professor at the University of Kansas School of Public Affairs and Administration and author of “Pulled Over: How Police Stops Define Race and Citizenship.”
Texas, however, “seems almost intent on not allowing an examination of that question.”
What happens after the stop
Houston’s population is less than a quarter African American. Yet more than a third of the traffic stops police made last year were of black drivers.
That may seem like clear evidence of racial profiling — policing based on the assumption certain races are more likely to be criminals. Yet police say — and researchers agree — that simply comparing the two percentages is not a good measure of whether a department disproportionately targets minorities during traffic stops.
Typically, more than half of cars pulled over by police departments are out-of-towners, making comparisons with the local population meaningless. Officers also say they seldom know the race of the driver before a stop. In 2018, the first year Texas police departments were required to report the information, most agencies said officers were aware of drivers’ skin color prior to pulling them over less than 10 percent of the time.
Studies have raised questions about their truthfulness. In May, Stanford University researchers found the number of black motorists police stopped dropped after daylight — “when a ‘veil of darkness’ masks one’s race, suggesting bias in stop decisions.”
The study observed the pattern at the Texas Department of Public Safety, whose officers last year reported knowing the race of drivers in only 2 percent of their stops. In a written response, the agency said it took profiling seriously and pointed to an audit by researchers at the University of North Texas concluding the department did not profile drivers by race.
Even if they don’t report seeing a driver’s skin color, however, police can anticipate it, Baumgartner said. “We live in segregated cities,” he said. “So police can sit on the black side of town and pull everyone over, and not go at all to the white side of town.”
Houston NAACP President James Douglas said police concentrated their patrols “in Fifth Ward, Third Ward, Sunnyside — places heavily populated by African Americans. If you spend most of your time in areas populated by African Americans, it stands to reason you will arrest them more frequently.”
Many researchers, therefore, have concluded that the best way to test racial bias is to examine what occurs after a stop, when police have no doubt of a driver’s race. Clear differences in how people are treated on the side of the road could signal bias.
“How many of those traffic stops turn into a ticket?” Houston City Council Member Letitia Plummer asked. “And then how many of those turn into a more serious issue? How many of those stops end up putting people in the criminal justice system?”
Baumgartner said the most reliable test is a comparison of how often motorists are searched, a decision often left up to officer discretion. Last year in Paris, northeast of Dallas, black motorists were searched at double the rate of whites. In Houston and Travis County, outside of Austin, sheriff ’s deputies searched black drivers at about 2½ times the rate of white motorists.
The disparities mirror a 2005 study, which found two-thirds of Texas police departments searched African Americans and Hispanic drivers at higher rates than white drivers.
Police and their supporters have argued no useful conclusion can be drawn without knowing what the searches uncover. Even when black drivers are searched at a higher rate than whites, if police discover contraband at the same rate on all the searches it demonstrates solid police work, not bias.
As a result, analyzing search “hit rates” by race — comparing what the searches turn up — has become “the meat and potatoes of the evaluation,” said Alex del Carmen, a criminal justice professor at Tarleton State University who trains police how to comply with Texas law.
The Sandra Bland Act was intended to provide this final piece of missing information. Bland, a 28-year-old African American woman, was stopped in 2015 by a Department of Public Safety trooper for failing to signal. The stop, captured on the trooper’s dash camera and Bland’s cellphone, escalated into a verbal, then physical confrontation. Bland committed suicide at the Waller County jail.
Among other changes aimed at keeping police more accountable, the new law was intended to compel departments to report on what their officers found when they stopped and searched motorists, said Rep. Garnet Coleman, D-Houston, who sponsored it. While some Texas departments had done it on their own, many did not.
Criminal justice reform advocates celebrated.
“Starting next year, we’ll find out how often these consent searches are productive and whether the results are racially disproportional,” Henson wrote at the time.
No race data from searches
But in the time between when the Sandra Bland Act passed and went into effect, TCOLE determined not to ask police departments to separate the search and contraband information by race — the key to any analysis.
“I wish I knew what happened,” said Dwight Steward, an Austin economist and researcher who has studied racial profiling in Texas police departments. What the commission produced, he said, is useless for measuring search disparities by race — and thus for identifying departments whose patterns suggest profiling.
Coleman said he, too, was baffled. “I don’t know why they didn’t do it the way we wanted it done,” he said. “We were pretty clear. (TCOLE) should have known exactly what we meant.”
Gretchen Grigsby, in TCOLE’s Office of Government Relations, said the law enforcement commission did its best to interpret the intent of the Sandra Bland Act “based on the information we had at the time.”
“We are certainly willing to work with the Legislature on any changes that need to be made to the report,” she added.
Until then, many Texas police departments will continue to escape a meaningful analysis of their work as envisioned by the law. The state’s reporting requirements yield only hints of behavior in need of further examination.
In Tyler, nearly a third of all vehicle searches are discretionary — a high number, said Epp, the Kansas researcher. Yet who is searched, and what is found by race is unavailable. A spokesman for the department said its only racial profiling analysis was what TCOLE requested.
In 2018 in Brooks County, between Laredo and Corpus Christi, 4 out of 5 motorist searches conducted by sheriff’s deputies were discretionary — yet fewer than 10 percent of the searches yielded any contraband, according to its TCOLE report. How that divides up among different races, however, is not required information. Sheriff Urbino Martinez did not respond to a request for comment.
Former Department of Public Safety commander Patrick O’Burke helped oversee reforms after the 1999 Tulia scandal, in which police in the Panhandle town arrested dozens of black residents based on a single informant later shown to be lying. Detailed traffic stop profiling data “could have been very helpful, but it didn’t ever achieve its full potential,” he said. Instead, “it’s just a perfunctory exercise.”
Other racial profiling components of the Sandra Bland Act also have had little impact. While the new law requires departments to make it easy to file a racial profiling complaint, departments differ in their approach. A review of reports that publish complaint information shows a tiny number with scant detail.
“A Caucasian officer conducted a traffic stop on a vehicle with a Hispanic driver and Native American passenger due to the vehicles expired registration,” according to the Plano Police Department’s 2019 report. “The officer greeted the two individuals and when the driver asked how the officer was doing he stated, ‘good, as long as you don’t call your friends to come shoot us up.’”
“If two white women were in the car instead he might not have made that comment,” the complaint said. The officer was cleared of racial bias.
There is no single place to analyze racial profiling complaints, but a review of individual department reports yielded none that was upheld. Advocates and legal aid organizations said they were unaware of any sustained traffic stop racial profiling complaint in Texas.
Disparity or bias?
Experts say even when armed with the best data demonstrating a police department conclusively treats minority drivers differently, proving bias is a challenge. A department’s numbers may be skewed by the behavior of a few officers.
“It is not difficult to measure whether there is disparity between racial/ethnic groups in stops made by police,” said Don Arp, executive director of the Nebraska Commission on Law Enforcement, which collects and analyzes state data. “The difficulty comes in identifying the causes for the disparity and whether or not it is based on race.” Over the past decade, no Nebraska racial profiling allegation has been upheld, according to the reports.
Still, Matt Simpson of the ACLU of Texas said comprehensive numbers at least offer a starting point for deeper examination of a local department’s policing. “People think this is going to be used for litigation,” he said. “But we’re really talking about adjusting policies and training.”
Grigsby, the TCOLE spokeswoman, said it wasn’t the agency’s role to analyze or act on the profiling data it collected. That’s left to local elected officials, whose interest varies widely.
In Austin, the Office of Police Oversight last year asked the police to perform a deeper analysis than required by the state. It showed clear racial disparities across stops, searches and arrests — none of it a surprise to the city’s minority residents, said Director Farah Muscadin. “They said, ‘You’re telling us what we experience.’”
Following the report, the oversight office invited citizens and police to meet. It later presented a dozen recommendations to the city manager, starting with an acknowledgment “that racial disparity exists and is worsening” and that previous methods of measuring the gaps were inadequate. It proposed statistical benchmarks to eventually close the stop and search disparities.
“The report they had been doing was only what the state required,” Muscadin said. “We needed to ask for more to really look at what the data said.”