By Gregory Korte
The Cincinnati Enquirer

Social scientists who study racial profiling have been confounded by a problem even a fifth-grade math student can appreciate.

They call it "the denominator problem."

If the question is, "Do police stop minority drivers more often than white drivers?" the answer lies in a mathematical equation.

The numerator - the first number in a long-division problem - is the number of black and white drivers stopped.

The denominator is the number that should be stopped if there's no bias in policing.

Figuring that denominator has proved to be an inexact science, fraught with variables. The neighborhood, time of day and the driving behaviors of individual motorists can all affect that number.

University of Cincinnati researchers addressed this problem by grouping the stops by time of day. During the day, the race of those working in the neighborhood was given more weight; at night, census numbers indicated what the residential population was. During rush hours, they factored in likely commuters.

They buttressed those numbers with observations from teams of trained UC students who observed the race of drivers at 126 thoroughfares in the city. Usually, those teams agreed on the race of drivers. In a few cases, they did not.

Researchers continue to hone that methodology, but they acknowledged the numbers are still an educated guess.

"I would suspect that if we came back here in 10 years, we would still have a denominator problem," said John E. Eck, a UC criminal justice researcher and one of the authors of the Cincinnati report.

Samuel Walker, a professor of criminal justice at the University of Nebraska at Omaha, is less optimistic.

"Apart from an observational study - which is very expensive - we're not going to get to the bottom of this," he said.

While the Eck report and similar studies still have merit, he cautioned that people "should recognize the limitations and not try to draw conclusions not warranted by the evidence."

Walker advocates a different way of looking at the problem, by making officer-to-officer comparisons of patrol officers assigned to the same shift and district.

"I don't believe all officers are biased. We need to identify those that are and then correct their behavior," he said.

The Boston Globe took a different approach, eliminating the denominator problem by looking at how drivers were treated once they were stopped.

Young white women in Massachusetts were more likely to get off with a warning, the Globe reported in July, while young African-American and Latino men were most likely to get a fine.

E-mail gkorte@enquirer.com