NEW BRAUNFELS, Texas (KXAN) — Clarence Crawford recalled feeling he was in potential danger when a New Braunfels police officer flashed his lights and sounded a siren behind him on Interstate 35 last January.

Crawford, a 43-year-old Black man, had not been speeding or swerving or breaking any traffic laws that he knew of, he later said. He exited the highway and pulled into a medical clinic parking lot for the traffic stop.

His arrest was swift.

Officer Kaleb Meyer stepped out of his patrol car with his handgun drawn and yelled at a passerby to move away. Meyer immediately pointed his gun at Crawford through the open driver’s side window and ordered him to place his hand on the wheel. Crawford argued with the officer about putting his hands on the wheel and getting out of the car but he ultimately complied when the officer pulled out a seat belt cutting tool. The officer grabbed him around the wrist as he got out, video shows.

It wasn’t until after Crawford was pulled from his white Pontiac sedan, pushed face down on the pavement, kneeled on, shocked twice on the back of his right knee with a stun gun, and handcuffed that he learned the reason for his arrest: a dirty rear license plate.

Crawford was charged with two misdemeanors, evading arrest and interfering with public duties, according to Comal County records.

Meyer’s body and dashboard video cameras captured the arrest, including Meyer’s justification to his supervisor for initiating the stop and why it so quickly boiled into a violent confrontation. Crawford had taken two minutes to stop, missed two highway exits and several driveways, kept his hands down when Meyer approached the car and didn’t cooperate with orders, Meyer said.

Ten months after the arrest, New Braunfels’ city manager released a statement saying Meyer had not followed his training. Meyer resigned, and Crawford’s version of the incident has cast further doubt on the reason it happened in the first place, according to court records.

Crawford said Meyer pulled up next to him at a stoplight before they got on I-35, indicating Meyer knew Crawford’s race. Crawford’s tag was supposedly too dirty, a violation of state law, but Meyer was able to relay the plate number correctly to dispatch on his first try. Perhaps above all, though Crawford was argumentative and yelling at times, police video shows Meyer did nothing to deescalate the situation.

Now, the city of New Braunfels faces a federal lawsuit alleging Crawford was racially profiled and his constitutional rights were violated. Charges against Crawford were dropped months later, but he was traumatized and lost his job two weeks after the arrest, according the federal lawsuit.

Deep flaws in reporting racial profiling

Other recent cases of Black individuals being stopped for minor traffic violations that escalated to violent confrontations garnered national attention. In 2015, an Austin Police officer stopped 26-year-old Breaion King, a Black elementary school teacher, after he said she was speeding. The officer picked up and slammed King down against the pavement in a mall parking lot during the encounter.

That same year, a state trooper in Waller County stopped 28-year-old Sandra Bland after he said she didn’t signal before changing lanes. The state trooper aimed his stun gun at Bland and screamed “I will light you up.” Bland was booked into jail and died days later. The medical examiner ruled her death a suicide.

Those high-profile incidents, and Bland’s in particular, spotlighted Texas’ racial profiling reporting law. That law requires all departments that conduct traffic stops to report their stop data and analyze it. Through these racial profiling reports, if done properly, a department can find indicators of racist policing and work to address it.

“We can’t necessarily get inside the head of the officer and study that case by case, but if we look at tens of thousands, or millions or hundreds of thousands of observations, we can see whether there’s any statistical patterns that suggest that Blacks and whites and Hispanics might not be treated equally by law enforcement,” said Frank Baumgartner, a professor at the University of North Carolina who specializes in statistical analysis of traffic stop outcomes.

Texas’ racial profiling law has been in place for two decades, but a KXAN investigation found flaws so deep in the state’s collection, oversight and enforcement that experts say the available information is essentially unusable.

“The data that are presented by the state of Texas are completely unhelpful, and you can’t reach any conclusions about the very purpose for which the data was collected. So, it’s useless,” Baumgartner said.

Laws not being followed, or enforced

Texas’ racial reporting laws have two vital reporting requirements. First, every agency that makes traffic stops must compile that data and submit it annually in a report to the Texas Commission on Law Enforcement, also know as TCOLE. The law requires those reports include an array of information, including a breakdown of stops by race and whether vehicle searches were conducted.

KXAN found TCOLE’s database is missing data from more than 250 law enforcement agencies in the past four years.

Second, law enforcement departments are also required to submit a “comparative analysis” of that racial profiling data to TCOLE. However, when KXAN requested those comparative analyses, TCOLE could not produce any for the past 10 years. TCOLE later admitted to KXAN it has not mandated that part of the law.

Furthermore, TCOLE has not taken a single official enforcement action against any law enforcement agency in the past five years for these failures. Texas law sets fines for non-compliance at $5,000 per violation.

In an interview with Kim Vickers, TCOLE’s executive director, KXAN asked if his agency has collected the comparative analysis annually.

“That’s a very good question, and it’s a difficult question,” Vickers responded. “Because what exactly is a comparative analysis? And, that can take many different forms.”

‘Like a car without gas’

The requirement to submit a comparative analysis has existed since Texas’ racial profiling law came into existence in 2001.

“Data without the analysis is like a car without gas; you simply cannot make anything out of it. It is just numbers that are absolutely meaningless,” said Dr. Alex del Carmen, a racial profiling report expert and professor of criminology at Tarleton State University in Fort Worth.

Between 2001 and 2009, the law required law enforcement agencies to submit their comparative analysis to their local governments.

A 2009 revision to the law, authored by former Irving Republican state Rep. Linda Harper-Brown, made it a requirement to submit the reports to TCOLE as well. That should have created a central, statewide repository.

“At that point we really didn’t have any means of collecting data to know who was being stopped, why they were being stopped or what crimes were being committed — and analyzing it,” Harper-Brown said in an interview with KXAN. “That was a big part of the bill. We wanted someone to analyze the data and tell us what was happening.”

TCOLE’s website provides explicit directions to carry out this law. To satisfy reporting requirements, a department must do two things: submit its data in an annual online report and also upload “a separate PDF comparative analysis document containing a statistical analysis of its motor vehicle stops compared to the gender and ethnic population of the agency’s reporting area.”

Vickers acknowledged that TCOLE has not collected that comparative analysis and that it should be a separate document from the data report.

“I get your point and I see exactly what you’re saying. The raw numbers are components of it, but somebody needs to look at it and try to interpret those,” Vickers said. “My issue is, you’re going to an agency of five officers and nobody there has an idea, even a clue, of how to do that. I’m not saying that excuses it, I’m just saying we need — the state needs to — and I guess that goes through us, to try to see if we can facilitate that.”

Harper-Brown said TCOLE had “let down the public” with its lack of effort in collecting the racial profiling analysis.

“It is very disappointing to me because, as I said, this is a protective measure for everyone involved,” Harper-Brown said. “This takes it out of the realm of hearsay or anecdotal evidence and puts it into actual data gathering and risk analysis to determine whether there is an issue within a law enforcement agency or an issue with a particular officer or a particular group of officers and to determine what measures need to be taken to correct the problem.”

One day after his interview with KXAN, Vickers said he had sent a message to every law enforcement department in the state explaining the requirement to submit a report and comparative analysis, including that a department’s chief could face action against his license or an agency could be hit with a civil penalty for failing to submit those records.

Scattershot compliance

Lacking a specific template for a comparative analysis, and with no enforcement of the provision, law enforcement departments across Texas have be left to independently decide if and how they will conform with that part of the law.

TCOLE’s most recent report database from 2019 contains over 2,000 law enforcement agencies. More than 1,500 of them reported traffic stops. Without TCOLE acting as a central clearinghouse, a member of the public, or a government agency or an advocacy group would need to file a request with each individual agency to get that record.

Since TCOLE could provide no records of any comparative analyses, KXAN requested the past five years of them from 34 Central Texas law enforcement departments. Some departments are doing them, but the varying responses show scattershot compliance with the law.

Of the departments that responded to KXAN, eight were able to provide an analysis and 13 could not.

Some pointed to the aggregate numbers in their annual reports and asserted those constituted the analysis. However, Vickers and del Carmen both said the annual data report alone did not qualify as an analysis.

Muddying the situation further, some departments told KXAN they had submitted an analysis to the TCOLE. But, in response to multiple requests, TCOLE could not locate or provide any analysis to KXAN.

“TCOLE has not been mandating the submission of the comparative analysis until now, but we have been collecting them from those agencies who have chosen to submit them,” said Gretchen Grigsby, TCOLE’s director of government relations. She added that if a department had mailed or emailed their analysis it would not have been “independently uploaded.”

Some departments, such as the San Marcos Police Department, did an analysis on their own. San Marcos created an eight page analysis in 2020. Others, like New Braunfels PD, hired del Carmen to perform their analyses, which was 81-pages long last year.

Del Carmen said it was “hard to believe” and “shocking news” to him that TCOLE had not collected the comparative analyses he has been preparing for years for numerous Texas agencies he counts as clients.

Counted among those departments that did choose to independently conduct their own analysis is Austin’s police department.

What an analysis can reveal

Evidence suggests when a department does commit to analyzing its traffic stop data, the results can be revelatory. Take, for instance, the City of Austin.

Austin’s offices of Equity, Innovation and Police Oversight issued a joint analysis of APD’s police stops in January 2020. The report found, from 2015 to 2018, “Black/African Americans and Hispanic/Latinos are increasingly overrepresented in motor vehicle stops from 2015-2018. White/Caucasians are increasingly underrepresented during the same time period.”

The report recommended APD “acknowledge that racial disparity exists and is worsening,” and said that previous methods of racial profiling analysis were “incomplete” and “resulted in a perception that a trend of disparity did not exist.”

A subsequent analysis issued by the city in November 2020, which lumped in 2019’s data, found “between 2015 and 2019, racial disparity persisted and, in many cases, grew worse.” That second report found Black and Hispanic people were overrepresented, compared to their share of the population, in practically every measured category of traffic policing.

Provided with a baseline of data from those reports, Austin officials have set a goal of zero racial disparity in traffic stops, citations, arrests, use-of-force incidents and deaths at the hands of officers by 2023, according to the November report.

Austin took it upon itself to analyze and investigate APD’s traffic stop data. However, with no mandate from TCOLE, many law enforcement departments have not.

Zero enforcement

The law gives TCOLE authority to discipline and fine a department for failing to comply with racial profiling rules, but KXAN found that has not been happening.

In addition to making TCOLE a central repository for racial profiling reporting, Harper-Brown’s 2009 bill gave the agency power to fine an agency $1,000 for failing to submit a report. That fine increased to $5,000 with the passage of the 2017 Sandra Bland Act, a sweeping piece of legislation that led to significant changes in the racial profiling law, including more detailed reporting.

According to records obtained by KXAN, TCOLE has not issued a single formal enforcement action or fine against a department for failing to submit a racial profiling report in the past five years.

“It is disturbing to me, and it is shocking to me and statistically improbable that over 2,000 police agencies in the state of Texas for the past years have not been identified — not one has been identified —  as an agency that has purposely violated the law, as it relates to racial profiling,” del Carmen said.

Missing data

Vickers initially said his agency has had “100% compliance” with the submittal of racial profiling reports.

However, KXAN found 250 instances over the past four years where departments have zero traffic stop data reported in TCOLE’s database.

After KXAN provided a list of those agencies to TCOLE, Grigsby acknowledged reports were likely missing.

“There are several others that likely should have submitted reports. As Chief Vickers said in his interview, we are always looking to improve this process and appreciate you bringing these to our attention,” Grigsby added.

It is not clear exactly how many of those 250 departments may not be required to report. Texas has hundreds of tiny law enforcement agencies that may not make traffic stops. Those small offices could potentially be exempt from racial profiling reporting requirements; however, they were not labeled exempt in the available data.

“Prior to the Sandra Bland Act taking effect, there was much less clarity that exempt agencies needed to submit confirmation to TCOLE affirming their exemption. That change began to be realized with the 2018 reports,” Grigsby said. “As you can imagine, most city marshals, fire marshals, water districts, DA’s offices, and ISD police departments do not regularly conduct traffic stops and are therefore exempt, but were not strongly pressed to submit a report of exemption prior to (the Bland Act).”

But, not all the agencies KXAN identified were small, fire departments, courts or one-man operations. Many were police departments serving cities with thousands of people. Some were sheriff’s offices overseeing entire counties.

KXAN found TCOLE had no data for 65 law enforcement departments with at least a dozen peace officers. Here’s a breakdown by year of those:

Are they checking the accuracy of their data?

The Bland Act also contained a provision making police agencies responsible for auditing their data to ensure accuracy. In responses to KXAN, many law enforcement agencies in Central Texas said they have no auditing procedures in place.

KXAN found 16 departments in Central Texas alone that said they have not conducted an audit in the past year.

In 2018 and 2019, Vickers said TCOLE did not check if an agency audited its own data. TCOLE has revised its 2020 reporting form, which should be submitted by all departments by March, to include a line requiring a signature affirming the department followed “statutory data audit requirements.”

Del Carmen said he audits data by taking a sample and checking it against the information in actual citations to make sure it was entered correctly. If a high percentage of the sample is correct, that likely indicates the overall data is reliable.

“If this mechanism is not in place in a police department, and they’re simply reporting data at the end of the year that they have not verified, there is a likelihood that the data that they are reporting is not accurate,” del Carmen said. “If it’s not accurate, any analyses, any inferences, that can be made on that data set are simply going to be null and void.”

TCOLE does not have the authority to go into racial profiling reports and verify the data, Vickers said.

KXAN asked Vickers if his agency could certify that the data the public has access to on TCOLE’s website can be relied on as accurate and truthful.

“The basic answer to your question is, ‘No,’” Vickers said. “I can’t tell you, from a TCOLE point of view, that these numbers are totally accurate because I have no way of knowing that. I have to accept what they send in and that’s what they’re held accountable for.” 

Micro-level data can be a solution

Baumgartner says one solution, and what is being done by several states already, is releasing data on individual stops. What Texas currently has available now is essentially “useless,” he said.

“The best solution is just release the spreadsheets of the micro-level individual traffic stops, one stop at the time. Let independent scholars, activists, people who support the police, want to reanalyze the data, to question and analysis by some outside actor,” he said. “Don’t hire a crony to do that analysis and don’t release an annual summary of the data that then can’t be analyzed.”

North Carolina has released over 25 million records of every traffic stop, and other states, like Illinois and Maryland, provide detailed records of stops that can be analyzed, he said.

Adding Texas to that list would likely require a legislative change to the law. But, as del Carmen said, it takes more than what Texas currently requires to find indicators of racial profiling.

“The cumulative or aggregate data only shows you patterns of the agency, which if you have an agency of 50 police officers and one of them is racist, that person is going to be somewhere along the way hidden in the midst of the other 49 because that data will never search if the police agency is not looking inward,” del Carmen said.

That look inward could tip a police department off to a problem and perhaps avert a case like Bland’s, or King’s, or Crawford’s in New Braunfels.

Crawford’s lawsuit remains pending in federal court, and his problems linger. His charges were dismissed, but they remain on his record. He lost his job, and he remains unemployed. All of that, a result of driving home from work with a dirty license plate.

Investigative Photographer Ben Friberg, Graphic Artist Rachel Garza, Editor Eric Lefenfeld, Director of Investigations & Innovation Josh Hinkle, Digital Executive Producer Kate Winkle and News Director Chad Cross contributed to this report.