Numbers can Vary:  Trust Your Perceptions
[May 22, 2008]

            Don't read this if numbers bore or intimidate you...  You shouldn't have to delve into the meanings behind the numbers.  At least, you shouldn't if you have confidence in those presenting them.

            Last year, 2007, we had 125 Murders in Oakland. I mentioned 136 in the fifth paragraph of my previous essay regarding "Oakland's Shock and Awe."   Since there appears to be such little shock and awe, perhaps the harmless error would have gone unnoticed for many.  Please see the new attachment which makes the correction.

            I recall my thinking that the murder average for 2006 - 2007 was 136.  I wanted to use that number first, as a more secular trend justification, but I didn't have the year-to-date number for 2006.  Had I used it, the shock would have been that Oakland could expect almost 200 murders in 2008.  While perhaps more sensationally justifiable, I chose to base my math on the 2007 murder total of 125 and not include the 2006 total of 146.

            This is another point to keep in mind.  The "official numbers" vary and are sometimes difficult to pin down.  The murder number for 2007 was officially reported to be 145, 146, 147 and 148.  Since just one employee is shuffling through the various OPD reports for Uniform Crime Reporting purposes, and thus also the Daily Crime Report, the numbers to-date are often understandably lagging.  There are also other number altering phenomena such as "adjusting the numbers," "recalibrating" them, and other hopefully benign reasons. 

            There are some things we should keep in mind with respect to the numbers game.

1.  The numbers will fluctuate year-to-date.  Some will crow when the numbers are fewer than last year, and hold frigid silence when they are higher.  On January 31, it was great to expect that the murders were thus far down 30%, and Felony Assaults were down 3%.  Past the middle of May, they are now up 45% and 9%.  The longer the trend line reviewed, the better they represent what is going on.  Yes, a year's experience can be meaningful, so it would be better for everyone to see a 12 month "trailing average."   

2.  Administrative pressures can skew the numbers.  Everyone now senses the tremendous pressure to "reduce the crime rate" in Oakland.  The Area Command Captains know they are being held personally "accountable" and they are transmitting that desperation to each Lieutenant, Sergeant and every officer on the street.  I viewed a Memphis police officers' site and read the consequences of such pressures there.  After all, they are Number 3 to our Number 4 in national violent crimes.  They talked about answering calls that would ordinarily be considered a serious Part 1 Felony, and being pressured to downgrade them.  This can occur from the Radio Room dispatcher to the cop on the Beat.  As example, a burglary might be side-tracked as a "suspicious circumstances."  A felony assault might be downgraded to a simple dispute (as vehicle accident victims might agree not to make a report).  The dispatcher might let a "hummer" remain un-serviced and eventually not reported.  Triaging calls is the first sign of potential influence on the "numbers."

3.  The numbers can be influenced by administrative redefinition of what the offenses are.  While the FBI and the Department of Justice strive to make the reporting consistent, human nature looking for loopholes can tempt police chiefs to find ways to lower the numbers.  Looking at the UCR data over the years, there are instances of large Departments not reporting certain offenses, or even all of them, for a particular year.  Oakland's Auto Theft number for 2007 was originally reported as being 10,549 by Crime Analysis, and then somehow reduced to an even 8,000 (I have copies).

4.  The numbers are obviously influenced by relative under-reporting.  Some crimes cannot be under-reported.  Looking at all Part 1 Serious Offenses, with the criteria remaining essentially the same, Oakland reported 44,117 such crimes in 1969.  In 2007, Oakland Police reported only 30,014 to the Department of Justice.  This is unlikely.  Auto Thefts, which have a high rate of reporting, were 2,742. In 2006, they were 10,549 for an increase of 285%.  That makes sense.  However, Burglaries in 1983 were 11,647 and in 2006 they dropped to 5,070?  That's not plausible.

5.  The numbers are also affected by the integrity of the administration.  Before anyone gets too indignant, let it be said that "corruption" of numbers also includes incompetence, lack of attention, and other examples of misfeasance.

6.  Lastly, the greatest influence on numbers, in all facets, is the quality of the police leadership, the first hint of such a misguiding leader is a litany of excuses where things will be better when we have more money, when we have more manpower, when we have more facilities, when we defeat the police union, and so on.  Crime, as speeders on city streets, is most affected by "police presence," and more so by efficiency in operations, and the motivation of all personnel.  Look to cities where the crime rates have stayed below the prevailing trends and you will find examples of positive leadership.  Look to cities that have crime rates consistently worse that prevailing trends, and the degree of leadership incompetence can be measured by the deviations from normative standards.

            In conclusion, the public and their policy makers shouldn't have to look deeply into the esoteric of crime statistics.  They should trust their experts.  After all, it takes a good expert to understand all the nuances of the numbers to translate their meanings into operations.  But that's the whole point, if the operations aren't effective, and so ineffective as in Oakland, then the nuances of the numbers are obviously consternation to the Chief in charge and confusion to the public at large.

ronoz