Risk-targeted activity to negativity ratios and/or percentages.
Early Warning Systems: Part 3
By: Doreen Jokerst and Randy Means
Ed. Note: Readers are best served to read Parts One and Two of this article before approaching the information in this final part. Go to www.hendonpub.com, click Resources, and then click Article Archives. This article is written in terms of field law enforcement; however, most of the concepts involved could also be applied to an early warning system for correction officers. The data input would differ but early warning theory is the same.
The only real solution to the problems described in Part Two of this article series involves more work, that is, the formulation of appropriate risk-targeted activity to negativity ratios or percentages. The use of raw number thresholds as triggers for early warning and intervention can be seriously counter-purposeful, misleading and even dangerous.
Citizen Contact Data
There are several activities that agencies can use according to their needs, organization type and size. For example, an agency could track complaints, lawsuits and uses of reportable force in ratio to stops and arrests (contacts) as those are the activities that tend to spawn those negativities.
This is a use of percentages. Officer A charges people with assault or battery on him/ her 28 percent more frequently than any other officer in the agency. Question: Why do so many people want to hit Officer A? A more sophisticated version of the same idea is to compare officer ratios. Officer A is assaulted in 10 percent of his contacts. No other officer in the organization is assaulted in more than 5 percent of their contacts.
Of course, the difference could arise from Officer A being more inclined than other officers to charge the crime, which could signal a need for more guidance and greater consistency on when and if the organization wants those charges made. Historically, officer discretion has largely been the guide on that question. Maybe that decision should be revisited, at least to an extent.
Using the same conceptual example, an agency could track resisting or obstructing or delaying charges or even disorderly conduct charges in terms of ratios or percentages of contacts. Why do people get so disorderly when Officer A is around? Arrestee injuries could be tracked as a percentage of arrests and compared to other officers’ percentages, seeking to learn why Officer A’s arrestees get hurt so much more often than other officers’ arrestees?
This could be similarly but further assessed analyzing officer injuries or accidents as a percentage of stops and arrests and compared to other officers’ percentages. Why does Officer A get hurt so much more than other officers, especially given lower activity levels (if that’s the case)? Additionally, assuming brandishing a firearm is a reportable involvement (which it should be), a targeted ratio could be firearms brandishments and discharges as a percentage of stops and arrests. Why does Officer A need to bring the gun to bear so much more often than other officers?
The risks or possible problems targeted by the applications above would include poor interpersonal communication skills, lack of emotional intelligence, power problems, propensity to use force and other constitutional rights violations. As discussed in length in Parts One and Two of this article, the agency must also ensure it is comparing apples to apples and not apples to oranges. Ratios are most useful and appropriate when comparing officers working the same shift, assignment, division/district, etc. Some assignments will produce far different numbers than others. Here, again, misuse of data can be counter-purposeful.
Tracking vehicle operation data, the same concept regarding comparable ratios and activities could and should be used. Raw number thresholds could be extremely misleading. So, driving complaints could be tracked as a percentage of other officers’ driving complaints. Why does Officer A get complained on regarding driving 32% more often than any other officer? Using GPS technology, an agency could learn every time an officer drives more than 85 mph. If that officer wasn’t on a freeway or in close pursuit of a violent felon, 85 mph is dangerously fast and a matter of early warning concern.
Vehicular accidents and pursuits as a percentage of other officers’ numbers could be tracked using the same analysis as that above. The same type of data could be used to identify broader organizational issues. In all of this work, the risks targeted would be injury and death, especially of innocent people, and also property damage. If an agency didn’t employ needed GPS technology, it could still evaluate “too good” response times and available pursuit data, including that flowing from in-car cameras.
Even though most agencies use their systems primarily to track and identify behavior of those working the patrol function, they could and should be used more broadly. In respect to case performance, the agency could assess cases rejected for prosecution as a percentage of other officers’ numbers.
Everybody has cases rejected, but why are Officer A’s cases rejected 26 percent more frequently than any other officers’? Why do prosecutors hate Officer A’s cases so much worse than they hate other officers’ cases? Although this data area might also be driven mostly by patrol activity, it could involve other officers, detectives and even crime scene personnel. Using the same analysis just mentioned, an agency could track cases dismissed by judges as a percentage of other officers’ numbers in that regard.
Evidence suppressions could and should also be looked at as a percentage of other officers’ numbers using the same analysis. Also, if a suppression order includes an evidence-based judicial finding that an officer has violated someone’s constitutional rights, an internal investigation should be launched.
It is understood that judges make mistakes in such matters but other complaints are routinely investigated without first knowing their legitimacy. Certainly judicial allegations should receive the same favor. In any event, it is dangerous in terms of downstream civil liability exposure that judges would make such “accusations” publicly and officially and that law enforcement would just ignore them.
It is sadly true that many officers are unaware of evidence suppressions in their cases or the reasons for case dismissals. Some officers don’t even know their cases have been dismissed. Officers should keep up with what’s happening to their cases and be required to report evidence suppressions as they do uses of force. Court systems aren’t always helpful in this regard, but obviously case outcomes should be tracked. The risks targeted in the case performance area are constitutional rights violations, poor report writing, and weak courtroom testimony skills.
In addition to all these ratio and percentage-based triggers, there are a multitude of other issues, somewhat more “personal” in nature, which need consideration, and which will require thoughtful methods of discovery and intervention. This will help agencies identify inappropriate relationships, stress, lack of rest, substance abuse, lack of critical knowledge and/or skills, propensity to use force and excessive force, temptations, desperation, and potential corruption.
Risk management is a high priority throughout law enforcement and is the job of every supervisor at every level, from sergeant to chief or sheriff. An identification and intervention program is an essential part of managing risk and improving accountability to standards. Used right, an early warning system can help employees identify and solve problems, improve themselves and their circumstances, and better succeed. It can make an organization smarter by improving its self-awareness and vision.
But if alarm thresholds and triggers are chosen poorly and intervention methods are selected unwisely, early warning systems can backfire and produce dangerously counter-purposeful consequences. Einstein contended that “the true sign of intelligence is not knowledge but imagination.” Most early warning systems are not very imaginative. They tend to identify the wrong officers, reduce positive activity, and make officers less safe. Logic and caring demand better.
Possible Top 20 List
1. Informant problems, including inappropriate fraternization and relationship
2. Association with criminals
3. Loss of equipment
4. Attendance problems
5. Sick leave use and abuse
6. Lack of rest indicators
7. Changes in work habits
8. Deterioration in physical fitness level
9. Changes in appearance
10. Attitude changes
11. Lack of “outside” involvements
12. Frequent address changes
13. Calls for service to officer’s residence
14. Lifestyle and financial changes (up or down)
15. Alcohol or prescription drug abuse
16. Stressors like divorces, deaths in family, children in school problems, etc.
17. Performance in situational/contextual training
18. Feedback from trainers
19. Results of post-incident reviews, including use-of-force reports
20. Citizen input, like street names (consider “Thumper”)
Lt. Doreen Jokerst is a 14-year veteran of the Parker Police Department, a suburban Denver-area agency. She currently heads its Professional Standards Unit and operates its early warning and internal affairs tracking system.
Randy Means is a career police legal advisor, consultant and trainer, and a nationally recognized expert in police systems and author of the book, The Law of Policing. He may be reached at email@example.com.