August 9, 1998
By Alan W. Bock, Senior Editorial Writer
(Ed. note: Alan Bock is one of my favorite writers. First, he is so smart that he usually agrees with me. Second, he is a really good journalist, who actually checks facts. Imagine that!

Politicians of both major parties point with pride to declining crime rates, as shown by official statistics, as evidence that their enlightened policies are working. There are reasons, however, to doubt that those statistics really reflect reality.

By the time they are compiled, the statistics are older than is usually acknowledged. The data in California’s 1997 crime report, for example, were compiled by local agencies and reported in 1995. Yet they are sometimes used to tout the wisdom of policies put in place after their compilation.

Nobody forces police agencies to get their reports in to the state, so there is no consistency from year to year in the number of police agencies reporting.

Criminologists believe an average of 30 percent of cities in California never report. In California, Oakland hasn’t reported for several years. Has there been no crime in Oakland? The upshot is that it is virtually impossible to compare crime statistics from year to year with any reliability.

A fingerprint card is supposed to accompany felony arrest information sent to the state. When those fingerprint cards do not accompany the records, those crimes are not included in the report. Some criminologists estimate this variable to be as high as 40 percent to 60 percent of the records without fingerprint cards.

The criteria for the seven serious crimes included in the national FBI report have changed over the years. Arson has been dropped and added again, the minimums for serious property crimes changed from $200 to $400. It makes it even more difficult to discern valid year-to-year trends.

The FBI does not maintain a uniform Uniform Crime Report, which is based on reports from state governments, most of which are at least 2 years old by the time the FBI gets them, and all of which have approximately as many anomalies as are found in the California reports.

The crime reports do not take into account demographic factors like the number of males aged 18-25 (the most crime-prone sector) as a percentage of the general population.

The California report uses sampling to create its estimates -analyzing 45 percent of reportable crimes in 1997, a larger amount than the previous year. Sampling can be sophisticated and might be necessary, but it reduces the reliability factor.

According to an FBI Victimization Survey released in September 1997, based on door-to-door surveys in sampled neighborhoods, only three of 10 crimes are ever reported to the police. Perhaps most of those unreported crimes are considered too minor to report, but nobody really knows.

Political pressure to show success at reducing crime may be leading to fudging. So far this year, as New York Times writer Fox Butterfield recently reported, there have been charges of falsely reporting crime statistics in New York, Atlanta and Boca Raton, Fla., resulting in the resignations of high-ranking police commanders. "In Boca Raton, for example," Butterfield wrote, "a police captain ... systematically downgraded property crimes like burglaries to vandalism, trespassing or missing property, reducing the city’s felony rate by almost 11 percent." Philadelphia has withdrawn its crime figures for 1996, 1997 and the first half of 1998 because of sloppiness, downgrading and under-reporting.

Most of the criminologists I talked to are aware of most of these shortcomings, but believe that murder is a fairly reliable indicator (since there is usually a body and the victim usually has relatives) and murder rates are down. So perhaps crime really is down.
(Ed. note: Remember that the Drug Czar said that the Dutch murder rate is twice ours when it is less than one fourth ours.)

On the other hand, it is possible that since 40 percent to 60 percent of felony reports to the state do not include a fingerprint card, some of those felonies might be murders, so the murder rate might be somewhat higher than state reports suggest.

And the fact the Oakland hasn’t reported to the state in three years means at least some murders don’t show up in the state reports.

I would love to believe that serious crime is finally declining. For reasons I outlined a few weeks ago, I doubt if the Three Strikes law has had much of an impact on crime rates, but it is just possible that various factors - a reduction in the percentage of young males, the peaking of the crack cocaine epidemic, economic growth finally having an impact on the propensity to do crime rather than go to work - have led to a reduction in crime.