FLORISSANT • If there's a better way for the state to measure racial profiling in traffic stops, Florissant Police Chief William Karabas would like to see it.

Last year, a state report showed his city to be among Missouri's worst places to "drive black." The figures indicated that Florissant officers were nine times more likely to stop a black driver than a white.

That's according to a yardstick the state uses, which compares the race of drivers stopped with the racial makeup of the community. It's a method that has remained unchanged since 2000, when state law began requiring police to report their stops to the state attorney general's office for annual tracking.

The law put Missouri in the forefront of identifying, and presumably addressing, harassment of minorities.

But pretty much everyone involved in the process has come to agree that the methodology is meaningless. One social scientist calls it "inept"; another says it's the "most untrusted, worst method."

Now some officials are beginning to look for a fairer, more relevant approach. Social scientists say there are better options out there — one of which is used in Illinois.

The problem in Missouri is a presumption that the driving population mirrors the driving-age census population. But highways and major destinations — such as malls, schools and large employers — can skew that dramatically.

"When it comes to traffic enforcement, we're dealing with a lot more than just the static population of the community," Karabas said. The chief noted the wide disparity between Florissant's African-American population of 25 percent and the Ferguson-Florissant School District's enrollment of blacks, which is 80 percent.

It isn't just police stung by high disparity ratios who complain.

"The benchmark that the (attorney general) uses has pretty much been discredited over the past decade," said Brian Withrow, associate professor of criminal justice at Texas State University-San Marcos. "I wouldn't call the attorney general a liar — at least not publicly — but there is such a thing as an untruth by omission."

Attorney General Chris Koster, who did not hold the position when the counting system was devised, declined an interview on the subject. His office issued a statement that said: "The reporting method was designed more than ten years ago and is the product of the law passed by the General Assembly in 2000. Certainly, we are willing to consider improvements in collecting information, keeping in mind the important value of year-to-year comparisons and historical data."

When Koster's office releases the numbers every year, it cautions that they are not proof of racial profiling but merely a chance to highlight potential problems and prompt dialogue.

Redditt Hudson, program associate for the American Civil Liberties Union of Eastern Missouri, has joined with the St. Louis and St. Louis County police chiefs and officials from the NAACP, Missouri Highway Patrol and Anti-Defamation League in seeking a better way.

"We know what the problem is, but what do you do about it?" asked Hudson, a former St. Louis police officer. "It may be time for the attorney general's office to turn over the responsibility for managing those statistics to a new entity that is more focused on accountability."

GATHERING DATA

When Missouri became only the fifth state to attempt to address racial profiling legislatively, there was plenty of debate — locally and nationally — on how best to do it. Most experts agree there is still no gold standard.

Denise Lieberman, then legal director of the ACLU, said tracking police stops seemed like "a good first step."

"The statistics themselves don't stop the problem, but they give you a way to quantify the problem so people can pursue remedies," she explained.

Richard Rosenfeld, a criminologist at the University of Missouri-St. Louis, one of three researchers who analyzes the data for Koster, said the legislation offers little flexibility.

It requires officers to record each driver's race, what prompted the stop, whether there was a search and what was found. Those stops are then compared with the community population by race to compute a "disparity index." A race scoring higher than 1.0 is considered overrepresented, under 1.0 is underrepresented.

Rosenfeld admits that it creates "a huge problem" that is "grossly inflating the disparity" for some communities.

Withrow, the Texas State researcher, came to the same conclusion when Ladue officials asked him to look at their data this fall. The state said Ladue officers were almost 21 times more likely to stop a black driver than a white driver.

Withrow, a former state trooper who has testified before Congress on racial profiling, was immediately struck by this: Police estimate that 93 percent of people they stop don't live in Ladue, where census figures say less than 1 percent of the population is African-American.

"There's no way the residential population of Ladue is indicative of the general driving population," Withrow said, elaborating in a scathing report that is now posted on the city's website.

Rosenfeld suggests a simple solution: Add a line to the checklists to show where the driver lives. "It wouldn't take us all the way, but it would take us a long way down the road," he said, noting that a conversation about it with Koster's office is ongoing.

Withrow, in his report for Ladue, suggests estimating a community's driving population by using accident data, which better indicate who is actually on the road.

Another option is found in Illinois, where the Department of Transportation has collected traffic stop data since 2004.

Its researchers, from the University of Illinois at Chicago's Center for Research in Law, do the same thing as Missouri but with a twist. They not only compare a police department's traffic stops with the racial makeup of its jurisdiction but also compare them to the racial makeup of its county and use whichever index number is lower.

Project director Alexander Weiss said Illinois lets police agencies challenge their numbers. That's happened about 30 times since 2004; two-thirds were changed as a result.

LAW STOPS AT NUMBERS

Missouri's statewide numbers — which experts judge to be more credible — show that since the study started, black drivers have been stopped at an increasingly disproportionate rate.

In 2001, the first full year of data, black motorists were 35 percent more likely to be pulled over than white drivers. By 2009, it was 70 percent more.

Researchers and activists say it shows the limits of a law that withholds funding from departments that don't submit data but does nothing to hold them accountable for results.

The legislation also calls for sensitivity training on racial profiling, but there's no indication the state follows up to make sure it's done.

"It's not apparent to me that very many jurisdictions have really changed the way in which officers are trained, counseled and supervised in regards to this problem," Rosenfeld said.

The Consortium for Police Leadership in Equity, an independently financed group of social scientists, is trying to fix that. It works with 15 metropolitan police departments across the country, including in St. Louis, to examine profiling and other race or gender equity issues.

Statistics are part of it. But researchers also seek access to personnel files and ask officers to volunteer for attitude and behavioral studies. Then they suggest reforms.

It represents an unusual pairing of social science and law enforcement, marked most significantly by the latter's willingness to open its doors.

"The chiefs that want us to come in are not afraid to hear the truths about their department," said Phillip Goff, a University of California-Los Angeles social psychology professor. He co-founded the coalition in 2009 with Denver Deputy Police Chief Tracie Keesee.

"This is going to tell you a lot more than any other analyses have, and we don't know how it's going to go," he said. "There's a lot of eyes on us waiting to see if this is going to be a successful enterprise."

St. Louis Police Chief Dan Isom said the state report — which showed his officers were 88 percent more likely to stop a black driver than a white driver in 2009, up from 7 percent since 2001 — is only a starting point.

His department looks at traffic stop data across police beats and flags potential problems for review by the district captain. Isom said he now wants a deeper look.

Goff and his team were here for two days in late February for an initial assessment, interviewing police commanders, aldermen and civil rights leaders.

Isom had the option of ending it there. But he chose to proceed with a full study, which will give researchers access to department records and the right to publish their findings (without identifying specific officers). It could last months, even years.

The chief pledges to publicly release their final report.

"In terms of credibility," he said, "we need to come back to you, we need to come back to the people they spoke to ... so they see this is not just window dressing."