Devlin Barrett has an article in the WSJ with the above headline in the printed paper. The online version, as of this writing, is headlined "Inadequate Data Hampers Law Enforcement in Fight Against Rising Crime." I preferred the original.
Another serious deficiency with official crime counts is that they are only "crimes known to the police." A crime committed but not reported does not show up in the official "crime rate."
This deficiency is particularly serious because there may be a toxic interaction between criminal justice policies and the statistics we use to measure the results, as SF Chron columnist Debra Saunders noted last week.
As law-enforcement officials struggle to cope with a sudden, unexplained rise in violent crime in many cities, they find themselves hampered by an outdated system for gathering national crime data that leaves them blind on such basic questions as how many murders happened last month.The article notes two deficiencies -- the long lag between crime and the official statistics and an undercount due to counting only the most serious crime in each incident.
Another serious deficiency with official crime counts is that they are only "crimes known to the police." A crime committed but not reported does not show up in the official "crime rate."
This deficiency is particularly serious because there may be a toxic interaction between criminal justice policies and the statistics we use to measure the results, as SF Chron columnist Debra Saunders noted last week.
Why would a victim of crime not report it? The National Crime Victimization Survey asks people this regularly. A study of annual surveys published in 2012 found that the relative numbers of people not reporting because it wasn't important enough and those not reporting because the police could not or would not do anything have shifted substantially.
Could the watering down of criminal sanctions for crime, particularly property crime, increase the perception the police will not do anything, thereby increasing the number of victims who do not report, thereby masking an increase in crime that results from the watering down?
Lest anyone think that is a feverish right-wing conspiracy theory, Saunders asked Magnus Lofstrom, an expert from the Public Policy Institute of California, an organization that regularly pumps out reports intended to boost the case for softening criminal sanctions.
Interaction between the phenomenon observed and the means of observing is a deep problem in many areas of science. In physics it is related to the Heisenberg Uncertainty Principle, a limit on how well we can know anything. In social science we have the Hawthorne Effect, the reason why we do "blind" studies and why studies that aren't or can't be blind are suspect. There may be an interaction of that kind at work here.
The NCVS is a needed and important cross-check on the "known to police" numbers, but it only gives us coarse data on a national or regional level for broad categories of crime. For all the reasons noted in the WSJ and the Chron, crime stats must be taken with a grain of salt. Reports from the field cannot be ignored or brushed off merely because they are not consistent with or not supported by official statistics.
In 1994, the percentage of unreported violent crime victimizations that were not reported because the victim believed the crime was not important enough to report (21%) was more than double the percentage that went unreported because the victim believed the police would not or could not help (10%). From 2005 to 2010, the percentage of victimizations that went unreported due to the belief that the police would not or could not help increased from 7% to 20%. In 2010, a greater percentage of unreported victimizations was not reported because the victim believed the police would not or could not help (20%), than was not reported because the victim did not think that the crime was important enough to report (15%).Note that is for violent crime. For "household property" crime (burglary, vehicle, theft), 60% of offenses are not reported, and "police would not or could not help" is the number one reason. The report did not track trends on this item, though.
Could the watering down of criminal sanctions for crime, particularly property crime, increase the perception the police will not do anything, thereby increasing the number of victims who do not report, thereby masking an increase in crime that results from the watering down?
Lest anyone think that is a feverish right-wing conspiracy theory, Saunders asked Magnus Lofstrom, an expert from the Public Policy Institute of California, an organization that regularly pumps out reports intended to boost the case for softening criminal sanctions.
I don't know that I trust these crime statistics, given some retailers' failure to apprehend shoplifters. As Chronicle columnist C.W. Nevius reported last weekend, security staff at a local Safeway often don't intervene in cases of in-your-face shoplifting. A security guard told Nevius that if someone swipes merchandise worth less than $1,000 (actually $950), "it is just a misdemeanor. We don't usually do anything." If retailers are letting people shoplift, then they're not reporting thefts to the authorities and property crimes are more prevalent than statistics show. "I'm an economist," PPIC's Lofstrom told me. "The incentive to report this particular crime, shoplifting, is decreasing over time."But it's not limited to shoplifting, of course. As the NCVS report rate shows, this is a serious, across-the-board problem and likely getting worse.
Interaction between the phenomenon observed and the means of observing is a deep problem in many areas of science. In physics it is related to the Heisenberg Uncertainty Principle, a limit on how well we can know anything. In social science we have the Hawthorne Effect, the reason why we do "blind" studies and why studies that aren't or can't be blind are suspect. There may be an interaction of that kind at work here.
The NCVS is a needed and important cross-check on the "known to police" numbers, but it only gives us coarse data on a national or regional level for broad categories of crime. For all the reasons noted in the WSJ and the Chron, crime stats must be taken with a grain of salt. Reports from the field cannot be ignored or brushed off merely because they are not consistent with or not supported by official statistics.
Leave a comment