Dallas PD's Crime Statistics Come Under Review
Dallas police Officers T. Schultz (left) and D. Torres chased down a theft suspect March 10 in Deep Ellum. Courtesy Dallas Police Department
The Dallas Morning News via YellowBrix
April 27, 2010
DALLAS – A city audit of how the Dallas Police Department collects crime statistics found that the department “makes a good faith effort” to comply with federal guidelines and that its numbers “appear substantially correct,” according to a report released Monday by the city auditor.
But several samplings of police reports cited in the audit appear to show large rates of error in how the department classifies various types of crimes.
The audit, for instance, found that in a sample of 60 “randomly selected” police reports that the Police Department did not include in its 2008 crime statistics, 11 should have counted as thefts.
One private auditor contacted by The Dallas Morning News said such an error rate seemed large.
“That would basically suggest that one out of six of things was misclassified, if their data is in fact correct,” said Jeff Weyandt, a certified public accountant and auditor for the Dallas accounting firm Fox, Byrd & Co. “That’s a pretty high error rate in any kind of a sample.”
That sample was among several taken by city auditors from 2008 crime statistics. Though it remains unclear how the samples were selected, the error rates appeared high. For example, among another sample of 60 “randomly selected” offenses not counted as crimes by the police, 13 should have been classified as car burglaries, the audit report says.
Other inconsistencies were evident in the audit’s examination of whether the Police Department correctly classifies crime for reporting to federal authorities. But while City Auditor Craig Kinton admitted that there appeared to be errors in his office’s report, he stood by the overall assertion that the Police Department has largely categorized crime correctly.
Although it is not mentioned in the official audit, Kinton said his office found that none of the crime categories were underreported by more than 5 percent.
The city audit was initiated before a Dallas Morning News investigation last year found that the Police Department lets hundreds and perhaps thousands of offenses slip from the city’s crime tally each year by not following federal guidelines.
The city auditor, who reports directly to the City Council, is primarily responsible for appraising the efficiency and effectiveness of city departments and services. The office also investigates reports of fraud and accounting irregularities.
Police are supposed to classify crime and collect statistics in accordance with the Uniform Crime Reporting Program, which is run by the FBI. Law enforcement agencies are not required to participate, but almost all do.
While the FBI strongly discourages comparing one place with another based on the raw statistics, many people do. The data is also used by the federal government to allocate certain federal grants.
Last year, Dallas shed its distinction, based on the data, as having the highest crime rate of U.S. cities with more than 1 million people. San Antonio went to the top of the list, and Dallas fell to No. 2.
The City Council has set a goal of being out of the top eight spots by 2013.
The audit report was presented Monday to the City Council’s Budget, Finance and Audit Committee on a decidedly positive note.
“Our overall conclusion is that DPD makes a good faith effort to comply with the UCR guidelines,” Kinton told council members.
“Our testing indicated that while we did note some errors … offenses reported for 2008 appear to be substantially correct.”
Committee members said the results should clear up doubts about the department’s crime reporting policies.
“It’s good to have fresh eyes come in, take a look and then be able to say that the numbers and what’s being reported is accurate,” said the committee’s chairman, Jerry Allen.
After an interview Monday afternoon with The News, Kinton admitted that portions of the audit report may not have clearly supported his findings. He asked for time to review issues raised by The News before responding by e-mail to some questions.
In the sampling of undercounted thefts, Kinton said that his office’s report may have mischaracterized the sample by calling it randomly selected. That portion of the report “may not clearly convey how the test results support the conclusion that offenses reported … appear substantially correct,” Kinton said in an e-mail.