FBI AUDITS OF CRIME STATS NOT WORTH THE PAPER THEY ARE PRINTED ON

Audits fail to detect deliberate manipulations; Milwaukee misreported thousands of violent assaults as minor crimes; Oklahoma City, Philadelphia and Seattle have never even been audited

FBI CRIME-REPORTING AUDITS ARE SHALLOW, INFREQUENT, an article by Ben Poston published in the August 18 issue of the Milwaukee Journal Sentinel, takes the FBI to task for its spotty and unreliable reporting of the nation’s crime statistics.

Milwaukee’s police chief and other officials say that reporting problems in Milwaukee are due to human and computer error and not the result of any manipulation of data. Yeah right! If you believe that, I have some prime oceanfront property in Arizona that I’ll sell you for a good price.

Here are some excerpts from Poston’s articvle:

The FBI's crime reporting program is considered the final word on crime trends in the United States, but the agency rarely audits police agencies providing the information and when it does its reviews are too cursory to identify deep flaws.

In each of the past five years, FBI auditors have reviewed crime statistics at less than 1% of the roughly 17,000 departments that report data, a Journal Sentinel examination of FBI records has found. In all, they've audited as many as 652 police agencies during that time, or less than 4% of the total.

And a Journal Sentinel survey of police departments in the 30 largest U.S. cities found that nearly two-thirds have not been audited in the past five years.

Of those, six departments - including Oklahoma City, Philadelphia and Seattle - have never been reviewed by the FBI since the auditing program began 15 years ago.

That lack of scrutiny allows cases of undercounting of crimes, such as in Milwaukee where thousands of violent assaults were not included in the crime rate since 2006, to go unnoticed and gives the public a false sense of the true level of crime, criminal justice experts said.

"It would be more candid to not do any (audits)," said Eli Silverman, professor emeritus at John Jay College of Criminal Justice. "This way, at least you're not offering any pretense of checking on the validity of the stats. If you are going to do that little, then why do it? You either do it systematically, or you don't do it at all."

He called the audit process a "useful fig leaf."

The FBI this year did its first-ever audit of the Milwaukee Police Department, even though it's the largest law enforcement agency in the state, generating about one-quarter of FBI index crimes in Wisconsin. That just-released audit, conducted at the request of Police Chief Edward Flynn, examined 60 incidents - a number experts say is too small to draw conclusions from.

"If they only pull a (small sample), the likelihood that they will find assaults that were downgraded is very low," said James Alan Fox, criminology professor at Northeastern University, of the FBI review. "Their ability to identify systematic misclassification is limited by the total volume of cases they check."

Indeed, the audit failed to detect the scope of problems already identified.
A Journal Sentinel investigation in May found more than 500 serious assaults - including stabbings and beatings - over a recent three-year period had been misreported as minor crimes. Another 800 assaults followed the same pattern. Those findings came from a partial review that compared Milwaukee police crime data with nearly 60,000 cases referred to prosecutors from 2009 to early 2012.

After the Journal Sentinel story, the Police Department launched its own targeted review of more than 34,000 cases, and in June released an initial audit report that showed police underreported more than 5,300 aggravated assaults since 2006.

FBI crime data is often cited by police chiefs and elected officials to give residents a measure of safety in their communities. In Milwaukee, Flynn and Mayor Tom Barrett have touted four straight years of declining crime.

But the information receives little outside scrutiny and is susceptible to manipulation by local police departments, Silverman said of the overall system.

"(Crime data) is a tool that politicians and police leaders use, yet the system is so incentivized to cast a favorable light and there is very little checks and balances to make sure it's accurate," he said.

Faulty crime data has far greater implications than just numbers on a spreadsheet, said John Eterno, director of the graduate criminal justice program at Molloy College in Long Island, N.Y. For example, police departments use the statistics to develop crime-fighting strategies and make hiring decisions.

"If you're saying your crime numbers are really low, but meanwhile they are really high and you aren't hiring cops, you are going to exacerbate the problem of criminality," Eterno said. "You need to have the correct numbers to do the right thing."

Eterno and Silverman have studied crime reporting practices at the New York City Police Department and other law enforcement agencies. Their research has identified widespread problems with performance management systems, most notably CompStat, which is a tool used by Milwaukee police and hundreds of other police departments across the country.

CompStat is used to hold police commanders accountable for crime trends in their districts - they must answer to the police chief and high-ranking officials at monthly meetings about performance metrics such as arrests, traffic stops and crime figures.

Eterno said performance management systems like this can unintentionally provide motivation for police supervisors to downgrade crime.