Baltimore police underreported serious crime in 1999 by 14.5 percent -- nearly 10,000 cases -- according to a confidential audit report.
The auditors found that police had improperly listed as simple assault 3,276 cases that should have been classified as aggravated assault, a much more serious crime. That means 50 percent more aggravated assaults were committed last year than police had reported, sources familiar with the report say. In another major discrepancy, 4,472 cases that had been reported as "destruction of property" should have been reported as larceny.
Police Commissioner Ronald L. Daniel said yesterday he had not seen the report but is not surprised. He would not blame his predecessor, Thomas C. Frazier, who had been accused by then-City Councilman Martin O'Malley of trying to make Baltimore look safer by underreporting crime.
Frazier, who left the post in the fall before O'Malley was elected mayor and works for the federal Department of Justice, did not return phone calls seeking comment.
"There is a problem, we discovered the problem, and we are correcting the problem," Daniel said. "There was human error involved, there were training issues involved, but I am not surprised at the results of the survey. I will say in the past I believe there has been pressure on district commanders to reduce crime, and I certainly do not approve of the process of reducing crime through creative writing."
Police officials are completing the audit to correct internal department reports before submitting crime data to the federal government by an April 1 deadline.
The underreporting is higher than initial estimates. A month ago, officials estimated the audit would show a 10 percent discrepancy. The report says 9,575 additional serious crimes were committed for the year, 14.5 percent more than the previously reported figure of 67,000. Overall, according to the report, lmost 77,000 serious crimes were reported in 1999.
The underreporting of serious crime in Baltimore far exceeds the figure of about 8.5 percent compiled recently in Philadelphia. The review of that city's crime statistics prompted national attention and a series of reforms.
The Baltimore audit, which was conducted by the department with the assistance of a private consulting firm, the Maple and Linder Group, also found a much smaller number of cases -- 152 -- in which crimes were improperly classified as more serious than they should have been.
In reaching their conclusions, the auditors reviewed randomly selected cases from the first six months of 1999 and then projected the results for a full year.
The review of crime statistics follows a lengthy public debate over the accuracy of city crime data. As a council member, O'Malley did his study of police crime data and accused Frazier of trying to make the city appear safer than it was, a charge Frazier denied.
Although Daniel refused yesterday to place the blame for underreporting on Frazier, he said his opinion could change after he reviews the final report.
"I can have my own opinion but absent direct proof it is difficult to ascertain why someone writes or does things they do," he said.
O'Malley said the audit was not part of a continuing row with Frazier.
"Anytime a new administration comes in, it is recommended that an audit be done to ensure that baseline crime figures are accurate so that any change in the future can be properly measured. We're not looking to make anybody's last year look any worse," he said. "We have to establish a baseline figure."
According to those familiar with the audit, the errors were uncovered in two ways. The audit unit headed by Maj. Walter J. Tuffy Jr. first reviewed police reports. From reviewing the narrative of those reports, they concluded that 1,455 cases had been improperly classified as less serious.
In addition, the reviewers made random calls to 3,959 crime victims, seeking further information about the incidents. For every 10 people called , the auditors found one victim who provided information that warranted upgrading the crime.
Between the calls and rereading the police reports, 1,849 cases were upgraded from a lower to higher category. Projecting those results for the first six months and then for the entire year, the report concludes that another 8,183 crimes had been improperly coded. The statistics were then adjusted downward to account for cases that were improperly coded as more serious than they were.
According to those familiar with the report, one of the most frequent errors came in cases classified under the category of "vagabond," a minor crime that instead should have been listed as larceny. These included some 300 cases involving the department's decoy unit. In these cases, the unit would place a cell phone or another valuable item in a parked car and wait for someone to steal it while undercover police watched. Often, the report found, police would charge those caught in the sting with theft, but then classify the case as "vagabond."
Other common errors were auto thefts classified as unauthorized use of an automobile and burglaries listed as lost property.
Among the cases improperly reported were five cases of rape that had been listed as less serious sexual offenses.
To combat the problems uncovered in the audit, Daniel said he plans to start an inspections unit that will be responsible for reviewing crime reports and other internal department documents. A similar unit was disbanded in 1997, but he said he hopes to have it operating again by June.
"You need to make sure there is a constant review, particularly with the young force we have," Daniel said.
Sun staff writer Gerald Shields contributed to this article.