The Rules - Engaging HVTs & OBL

[jmm99]

as evidenced in The MMRMA Deadly Force Project, MMRMA = Michigan Municipal Risk Management Authority, the insurer for many municipalities.

As Legal Adviser Koh's statement makes clear, OBL was targeted because he was part of a force declared hostile (see this post for SROE ref). PID (that he was he) was sufficient to justify "shoot".

The Deadly Force Project involved a multi-scenario simulation geared to civilian law enforcement. I can't claim any sort of personal experience with those MI departments (Grand Rapids Police Department, Ingham County Sheriff’s Department, Kent County Sheriff’s Department, Livingston County Sheriff’s Department, Livonia Police Department and the Southfield Police Department):

Each participating officer/deputy was randomly subjected to a total of three scenarios, with one of each of three types of scenarios:

• Robbery-In-Progress
• Burglary Alarm-In-Progress
• Mugging-In-Progress

Of these three scenarios listed above, all officers were also randomly exposed to three different outcomes of each scenario [JMM: all started with subject's back to officer and hands at waist level of subject's front; then an about face to confront officer]:

• Suspect empty-hand surrender
• Suspect surrender with (non-weapon) object in-hand
• Suspect shoots at officer

To further determine whether outward appearance might influence a participant’s tendency to use deadly force, we also embedded random variations of suspect attire in the scenarios. In all scenario varieties, suspects were filmed in both “dress” and “punk” attire. “Dress” attire was usually a shirt and tie, often with a dress jacket or coat as an outer garment. “Punk” attire varied from individual to individual, based on what actors brought to the film session. Clothing ranged from blue jeans, hooded sweatshirts and leather jackets.

An important finding from the study was lag-time, as well as a definite speed up in the officer response time depending on how serious the situation (robbery > mugging > burglery) was perceived by the officer:

As one might anticipate, the lag-time between when the officer made his/her decision to fire and when the initial shot is fired has critical implications. If an officer finds him/herself behind the reactionary curve in a rapidly evolving situation, the lag-time associated with decision-making can allow the suspect to fire one or more shots at the officer before fire can be returned by the officer. Another critical lag-time consideration becomes manifest when an officer fires at a person who ultimately becomes known to be an unarmed person. As suspects turn toward the camera (officer) they may appear to be armed as the turn is initiated - the suspect’s hands may be clenched and/or positioned at or near waist level. However, since the officer’s decision to fire at the suspect predates the subject being shot .25 seconds or more, the officer can (and easily does) shoot the suspect as he/she is raising his/her hands into a “surrender” position. This was a frequent and somewhat unanticipated outcome in many of the shootings that involved “unarmed” suspects; suspects getting shot while “surrendering.” The officer typically has 1/3 of a second or less (from a critical juncture in each scenario) to decide whether or not to employ deadly force, and then to apply that force, before he/she risks being “shot.”

Those officers who managed to shoot armed suspects before the suspect was able to fire seemed to have elected to use deadly force before it could be clearly determined that the suspect did, in fact, have a handgun. This tendency to employ deadly force “preemptively” was at the core of our objective to quantify when such actions were objectively reasonable. As anticipated, most officers found themselves firing after the suspect fired his/her first shot at the officer.

Officer’s Reaction to When Suspect Fires - Frequency
Officers shoots after subject shoots - 61.44%
Officer shoots before subject shoots - 38.56%

As one might expect, officers seemed more vigilant and situationally postured when they were informed that they were responding to a forcible felony-in-progress. Participants typically had their handguns drawn sooner in the robbery scenarios and were more inclined to utilize the cover that was afforded to them. When handguns were drawn and used in conjunction with a verbal challenge in the robbery scenarios, participant muzzle dispositions were notably higher (i.e., more elevated toward the suspect). And yet, even this higher state of situational readiness seemed to offer little more than a statistical “dead-heat” to the suspect’s first shot fired.

The unarmed shoots became the centerpiece of the study's analysis:

Inter-Agency Shooting Variations

The 307 officers/deputies participating in this study shot a total of 117 unarmed suspects – a frequency of 38% (0.3811). However, there was a significant difference between one agency on the low end of the frequency scale and agencies on the higher end. It should also be noted that the agency with the lowest frequency of participants shooting unarmed suspects and the agency with the highest frequency of unarmed suspects shot were both largely classified as being “urban” agencies. (See table below)

AGENCY, PARTICIPANTS, UNARMED SUSPECTS SHOT, FREQUENCY
1 76 18 24%
2 46* 18 39%
3 51 20 40%
4 38* 16 42%
5 54* 24 44%
6 43 21 49%
*Indicates that individual participants shot more than one unarmed suspect

The question will undoubtedly arise; “What noted differences were there between the agency with the lowest frequency of shootings (of unarmed suspects) and those with the highest frequency?” The answer, simply put; “It was a difference in training.” (See “Training Implications”)

The aggregate 38% unarmed being shot, and the aggregate 38.6% officer shoots first, correlate.

The range from 49% unarmed shoots to 24% unarmed shoots can be superimposed on a rough legal standard - not addressed by the study as such. The 49% unarmed result just accords with "preponderence" (more likely than not, or 50 yards and a nose). The 24% unarmed result reaches into what I would call "clear and convincing evidence". However, that result was a definite outlier. Why ?:

As was previously noted (“Inter-Agency Shooting Variations”), the distinct inter-agency differences relevant to shooting unarmed suspects seem directly attributable to training. There were no substantive differences in agency policies pertinent to the use of deadly force.

The one agency that required its personnel to complete a “Use of Force Report” whenever unholstering their handguns had a 44% rate of frequency in engagement of unarmed suspects. Its participants did typically unholster their handguns more slowly than participants from other agencies, but that didn’t seem to influence their overall judgment in the research scenarios.

The agency with the lowest frequency of unarmed suspects shot (24%), judging from all informal participant debriefs, had the most rigorous scenario-based training regimen. Virtually every participant interviewed from that agency stated that he/she had had one or more force-on-force training sessions in the last 12 months. In itself, this might not seem evidence adequate to suggest that training was the most influential factor, but it is the only factor that clearly stood out from all others.

We were impressed by the overall professionalism exhibited by participants afforded to us by all agencies. Scenario-based training was evident to some degree in all participating agencies. However, in all but one agency, it seemed much more intermittent rather than routine.

All of this accords with e.g., Mark Martins' articles (from the 1990s to today) that ROEs must be taught in the context of tactical-based scenarios - and cannot be "legislated".

The policy implications of the study should be obvious - in the following, "immediate" roughly equates to a "clear and convincing" > "beyond a reasonable doubt" stuation, and "imminent" equates to a "preponderence" > "clear and convincing" situation:

As many aforementioned findings suggest, many police shootings are “gray-area” events that often reflect split-second differentials between in-policy and out-of-policy shootings. The almost universal embrace of the “imminent threat” standard, in our opinion, is warranted and court-defensible.

For clarification purposes, “immediate threat” scenarios are “must-shoot” incidents, and therefore generate the least amount of post-incident controversy. When an officer responds to an immediate threat, he/she is either reacting to someone lunging with an impact or edged weapon, looking down the barrel of a suspect’s gun, or dodging bullets that have already been fired.

We don’t require that officers wait until a threat has manifested to that level before he/she can react with deadly force. An insightful operational definition of what “imminent danger” is was expressed by the New Jersey Division of Criminal Justice in 2000;

Imminent Danger:

• Threatened actions or outcomes that may occur during an encounter

• Threatened harm does not have to be instantaneous

This concept affords officers a large degree of latitude in their interpretation of events – as does the Graham v. Connor “reasonable officer” standard.

Koh: "...bin Laden continued to pose an imminent threat.."

The MMRMA Deadly Force Project was the most comprehensive found by me after some InterWeb searching.

Regards

Mike