As I mentioned previously, I’ve been working on a new way to analyze the data the NPMSRP gathers in order to create some sort of metric that would help predict where police misconduct is likely to trend upwards or downward in the future and give some indication as to why police misconduct rates might trend as certain way in one area but not another.
Of course, such an metric will take quite a while to develop, test, and refine before it’s finalized. A case in point is that I’ve already changed it a bit since I introduced the effort a few days ago, hence the new post… but I still want to share what I’ve developed so far and plan to test if the NPMSRP is able to continue past the end of this year. Mostly to see if there is any good feedback about the idea or any interest in it.
The Police Misconduct Index is an experimental measurement that analyzes NPMSRP statistics to determine where police misconduct rates are more or less likely to increase by determining how often police misconduct complaints result in disciplinary action, criminal charges, and criminal convictions under the assumption that the less often police misconduct is acted upon the more likely it is that police misconduct rates will increase.
For example, the following chart ranks states by projected Police Misconduct Rates as published in the NPMSRP Semi-Annual Statistical Report and further breaks down those statistics to determine how many, as a percentage, were followed by disciplinary action, criminal prosecution, and convictions. These statistics are then weighted and results in a PMI ranging from 0-10 which indicates how likely it is that misconduct rates will increase. (0 being least likely, 10 being most likely).
The PMI chart is color coded to indicate which areas have higher misconduct rates (red), lower misconduct rates (green), and average rate (white) as well as which areas discipline or prosecute misconduct less often (red), more often (green), or near the average (white).
PMR p100k indicates the projected police misconduct rate per 100k officers. In other words, if that area had 100,000 police officers, this is how many would be involved in a reported incident of misconduct within a year.
PMDPR indicates how many officers, as a percentage, out of those reported will end up being disciplined or criminally charged in relation to that incident.
PMCPR indicates how many officers, as a percentage, out of those reported will be criminally prosecuted in relation to that incident.
PMConR1 indicates how many officers, as a percentage, out of those prosecuted will be convicted on the charges made in relation to that incident.
PMConR2 indicates how many officers, as a percentage, out of those reported who will be convicted in response to the reported incident.
Since it is assumed that not all incidents reported are true, these rates are calculated on an average within each category which helps reduce the likelihood of false charges affecting the PMI.
The PMI charts not only give us a picture of how often misconduct is acted upon by disciplinary action or criminal charges to help determine where police get away with misconduct more or less often, but also helps determine where misconduct rates may be more or less affected by under-reporting. When PMR rates are low along with a corresponding low disciplinary rate this would indicate that there is a potential under-reporting issue due to state laws that restrict oversight or freedom of information releases for misconduct information.
For example, California has a low PMR, but also has low disciplinary rates. California also has a restrictive LEO Bill of Rights law that prevents police departments from releasing information about officers involved with alleged acts of misconduct and what happens as a result of that misconduct. This causes a problem with under-reporting and increases the likelihood that misconduct rates will increase and that currently reported misconduct rates are artificially low in that state.
The PMI can be used at a local level so long as the number of officers employed by a given law enforcement agency are significant enough to allow for accurate comparisons. Calculations for departments with fewer than 100 employees would be more susceptible to error due to a small sample size. However, when examining large departments, such as those with over 1000 sworn employees, the PMI can be used similarly as shown below when applied to the 10 worst departments with 1000+ sworn officers by PMR as ranked in the NPMSRP Semi-Annual report.
When we look at the results in this localized PMI table and compare this with recent events documented in the media, we can see the two corroborate each other surprisingly well. For example, it’s well documented that a problem in Dallas Texas is that, while the police chief there has been aggressive about responding to police misconduct, state laws and local union contracts have stifled his ability to effectively act against misconduct, as reflected in a high PMDPR but low PMCPR and low conviction rates.
Conversely, Denver Colorado has been in the news for having a stunning 0% disciplinary rate for brutality complaints and a similar 0% conviction rate for officers criminally charged with serious acts of misconduct, but a high payout rate for misconduct civil suits. In Atlanta Georgia, we can see how the department there is under a lot of fire, but how a particularly high-profile incident has pushed them to begin addressing misconduct issues… though it’s still uncertain whether those efforts will ultimately be successful.
So, even though it’s still under development, the PMI and the data analysis used to generate the PMI, appears to be a pretty useful way to look at the data the NPMSRP has gathered so far. Not only because it shows misconduct rates in comparison to disciplinary rates, but also because it can act as an indicator as to how misconduct might trend in the future for a given agency or region.
As always, especially since the PMI is under development, I am interested in any feedback about how useful you think this information is in comparison to the current NPMSRP statistics made available on this site and if it should be included in our future reports as part of the NPMSRP statistical reports.
Let me know what you think in the comments below or send me an email at packratt@injusticeeverywhere.com
Sorry, there are no polls available at the moment.
I think the info is very useful. I’m always linking people to this site in various blogs, FB or Myspace comments, etc.. I think that the more people that see these statistics, the more their eyes will be opened to the misconduct that goes on within law enforcement. How you present those statistics is very important.. this is good.
I was never big on statistics and don’t remember the formulas. But I’d be willing to bet that in some small suburban towns or counties with a high PMR p50k that the odds of falling prey to a criminal cop beat the odds of being victimized by a civilian criminal. I bet some of your formulas would bear this out. Might make for some interesting commentary on the state of policing in America.
Well, the problem with reading too much into the PMR and PMI of small departments is that a single case can put those departments above average for PMR/PMI. After all, a single case in a single officer department gives that agency a 100% PMR.
This skewing makes it harder to say that a high PMR is an issue of systemic problems or just an aberration. This is why I tend to focus on departments with larger populations since it takes several cases to put them over average and that tends to make it easier to suggest that this above-average statistic is an indication of deeper problems.
Hope that helps.
It does. With hiring standards being almost nonexistent, and given the type of people Law Enforcement attracts, it will be interesting to see if those slaps on the wrist make any difference. Good point about California. Do you think you get less news stories from CA because of that officers bill of rights?