|
Philadelphia Inquirer
Saturday, October 18, 1997
Phila.'s crime data discarded by FBI
The city won't be part of national crime surveys till next year,
after flaws in its reporting methods have been corrected.
By Clea Benson and Craig R. McCoy, INQUIRER STAFF WRITERS
Inquirer staff writer Howard Goodman and graphic artist Matt Ericson
contributed to this article.
The FBI, citing problems with how Philadelphia police count crime, is
purging Philadelphia statistics for the last year and a half from its national
surveys of crime.
The decision to discard the data followed meetings here between top Police
Department officials and FBI representatives to discuss flaws in the city's
system for tallying crime and how to correct them.
Both sides agreed to toss out the data reported by Philadelphia police from
the start of last year until mid-1997 so that the department can submit
corrected numbers next year, city and FBI officials said.
``We are basically removing those numbers for Philadelphia from the data,''
said Harlin McEwan, an FBI specialist in crime reporting who met with police
brass earlier this month. ``We would suggest . . . that those numbers should
not be relied upon, and they know that.
``It was a joint decision,'' he said. ``Clearly, we aren't going to use
numbers that aren't accurate. They don't disagree.''
McEwan said that there was no reason to believe police were under-reporting
crime, and that the corrected numbers may be close to those originally filed.
He said that other major cities have had crime statistics removed from the
FBI's Uniform Crime Reporting program, but usually because of a change in
methodology.
He said he knew of no case comparable to Philadelphia's, in which data were
tossed because of a longstanding counting system that turned out to be
fundamentally flawed.
For at least the last 30 years, the numbers reported to the FBI by
Philadelphia police haven't reflected the crimes that actually occurred in a
given month or year, but rather those that police logged into their records in
that period.
Because crime reports often take months to process, a lag developed. One in
every five crimes that police report in a given year actually occurred the
year before.
Police Commissioner Richard Neal said the decision to pull the data from
the FBI's voluntary reporting program was his. The department is instituting a
new, technologically advanced system that will allow it to conform to FBI
standards by next year, he said, and he wanted to start the new method as soon
as possible.
``I contacted the FBI and asked them to come here so we could discuss
exactly how this data was to be gathered,'' Neal said. ``We had a very good
meeting. Because we are revising our system, we decided we would re-calculate
the data to conform to a month-to-month basis.''
Mayor Rendell last night praised Neal for taking action to change the
system.
``I think it's very commendable of him to have the desire to get it right
not only in 1996 and 1997, but for the future,'' he said.
LONG-TERM TRENDS
The police say they have used the old tallying method consistently for
many years, so the numbers are valid for following long-term crime trends.
Questions about the department's reporting methods arose last month in the
midst of a debate over whether Philadelphia has made the same technological
advances as cities like New York, where up-to-the minute computerized crime
tallies are at the heart of successful policing efforts.
At a Sept. 2 hearing in which Philadelphia lawmakers discussed policing
strategies, Rendell countered critics of the department by unveiling
statistics showing that crime in Philadelphia dropped 17 percent in the first
half of 1997.
It later came to light that those figures were not a clean count for the
first half of the year and included some crimes from late 1996. Rendell said
it was ``embarrassing'' and ``unconscionable'' that police had given him the
numbers without explaining the peculiarities.
The department's method has resulted in a skewed picture of crime in
Philadelphia. One problem is that comparisons with other cities are difficult
because the time periods don't match. Another issue is that the city's
breakdown by types of crime is slightly out of whack.
For example, the city reported to the FBI that 7,155 assaults occurred in
1995. An Inquirer review found that there were actually 6,561 assaults in
1995, meaning that police overstated the number by 8 percent.
The Inquirer's review used computer files provided by the Rendell
administration that show actual dates on which thousands of crimes occurred.
The analysis of the 1995 data showed that the inaccuracies were not large for
most crimes.
UPDATED TALLIES
The Police Department said it would submit updated tallies of 1996 and 1997
crime to the FBI in time for the spring release of the preliminary national
data for 1997.
Neal and Rendell said they believed the recount would vindicate their
claims that crime has been dropping in Philadelphia.
Because of the recount, though, Philadelphia will not be included in next
month's planned release of national figures from the first half of 1997. The
FBI last month released its final report on 1996 crime, including figures for
Philadelphia, but federal officials also will no longer give out that
Philadelphia data, McEwan said.
McEwan said it was impossible to tell what impact leaving out Philadelphia
would have on the national crime rate tallied for the first half of 1997. A
review of recent data shows that Philadelphia has accounted for slightly less
than 1 percent of the nation's crime.
The department has been reporting its crime data every month, but Neal said
that would change under the new system. Police will now report every six
months, as cities such as New York do, so they will have more time to go back
and correct their figures.
------------- ------------------ --------------------- ------------------ --------------------- ----------------- ----------------- -------------------
|