Is “Necessity” The Answer To Stop Police Murder?

Yale lawprof Jed Rubenfeld, with a couple of law students, proposes an alternative test to the “reasonable cop” metric for determining when a police officer should shoot.  His avowed purpose is both lofty and curious:

IN police shooting cases like that of 12-year-old Tamir Rice in Cleveland, grand jurors are typically asked to focus on one question: At the moment the officer fired, did he have reason to believe the person he shot posed a grave and imminent threat?

This question fails to consider whether alternatives to deadly force were available and how race might have affected the officer’s perception of the threat. Under the prevailing legal standard, those questions aren’t relevant. Until we make them relevant, American policing will continue to have a tragic excessive force problem and, specifically, a race problem.

The disparate impact view of police killing leaves little doubt that cops perceive blacks as more of a threat than whites.  Empirical studies back up the obvious.

But racial bias can affect what seems reasonable. Individuals of all races in America perceive black people as more aggressive and dangerous than white people. Studies show that black people are seen as being physically stronger and less prone to feeling pain than people of other races, and black children are often perceived to be older than they are. When faced with an armed black target, shooters are both more likely to shoot and quicker to shoot than they are when faced with an armed white target.

These biases can affect the way we think, judge and act. As a result, force that may seem unreasonable if used against a white person may seem perfectly “reasonable” when used against a black person.

What makes this a curious approach when seeking to propose a new test is that it suggests that race be a factor in determining whether a shoot is righteous or not.  While it may be true that police are more apt to view a black person as a potential threat, is the answer to shoot more whites to even out the score?

Obviously, that’s not what he proposes. Rather, the inclusion of the bias piece is meant to question the reasonableness test.

Police officers in the United States are taught that they can use deadly force if they reasonably believe an individual poses a grave, imminent danger to themselves or others. Superficially, this “reasonableness” rule looks unobjectionable.

The use of “reasonableness” is one of those neat law tricks that people find invariably appealing. After all, how can reasonableness be wrong? It’s so, well, reasonable.  But Rubenfeld doesn’t accurately explain the existing test.  If it were merely “reasonable,” the problem would be substantially different than it is in practice.  Rather, the test is derived from Graham v. Connor, which was a § 1983 case that’s been adopted for use in criminal prosecutions for homicides by cops.

The distinction between the Supreme Court’s supposed “objective reasonableness” test and its application in real life is that it’s cop reasonableness, because non-cops can’t understand or appreciate the secret, magic voodoo that gives rise to a police officer’s perception of a threat that appears completely non-threatening to mere mortals.  Of course, cops demand this nuance, because they’re special, making split-second decisions while doing a dangerous job under life-threatening situations where their special training enables them to see things that no one else, no juror, no judge, no lawprof, no one but a person trained and experienced in the intricacies of police work, can see.

Without acknowledging this critical aspect of the analysis, Rubenfeld proceeds to his fix.

One critical and common-sense change would be to adopt a necessity rule. The difference is simple but crucial. Even when the police have a reasonable belief that a person is dangerous, the necessity standard does not permit deadly force if non-deadly or less deadly alternatives are available and adequate to meet the threat.

Unworkable? Rubenfeld notes that this is already the DoJ policy, which putatively requires “the use of the least or even a less deadly alternative so long as the use of deadly force is reasonable.”  It’s not law, of course, but discretionary policy. He adds another requirement as well.

Only eight states legally require their police officers to give a warning, where feasible, before using deadly force. The necessity rule would make this mandatory. The officer who killed Tamir Rice fired less than two seconds after arriving on the scene. If Ohio law had required the police to first try warning Tamir, which they could have done from inside their vehicle, he, too, would probably still be alive.

Well, in the Tamir Rice case, Loehmann claims to have issued three commands. And even if he didn’t, would his failure have meant that he was obliged to stand there and be shot if his perception of the threat had been credible?

Rubenfeld recognizes the obvious criticism of his solution:

The difference is not semantic; it can be a matter of life or death. A serious necessity rule would require all officers to carry nonlethal weapons and deploy them when feasible.

That it can be a “matter of life or death” doesn’t make his solution more than mere semantics.  As reflected in the wiggle words he uses at each level, such as “when feasible,” these fine ideas are subject to the same dreaded “reasonableness” as he condemns.  So a cop contends that a warning wasn’t feasible? Boom, dead black kid. So a cop contends that he didn’t shoot because the kid was black, but because he reached for a gun? Boom, dead black kid. So the other person isn’t black? Boom, dead white kid.

Problem solved? Not quite.

There is no easy fix to compel police to wait until a threat is real, imminent and certain, before killing. It happens along a spectrum of threats and excuses, feelings and senses, doubts and the certainty that once a cop sees the flash of the muzzle, it’s too late. No rule can require a cop to allow himself to be killed before acting to defend his life.

The variables are too great, human perceptions and actions too frail, to construct requirements that will mean anything more than the magic words to be spoken to justify a police shooting.  While Rubenfeld’s solution sounds good on paper, it’s a steaming pile of horseshit on the street, where cops will shift their justification from “I was in fear of my life” to “I was in fear of my life, perceived no opportunity to use lesser force or stop the bullet by issuing a warning.”  And there will be an expert cop witness to explain why the officer’s shots were objectively reasonable.

And therein lies the rub.  As long as the Graham v. Connor test, the objectively reasonable cop metric, precludes a jury from finding against the steaming pile of horseshit presented to them from some cop-expert whose livelihood and goal in life is to rationalize why every trigger pull by a police officer couldn’t be helped, there will be no consequences for a needless killing.

It’s hardly a perfect solution, but the perfect solution would require police who choose not to kill when it’s unnecessary, when they can use lesser force or simply wait until the potential of a threat of harm pans out to the point where it’s real.  But cops are free to do that now, yet don’t.  At least their actions can be judged through the eyes of society, the people who pay for their existence, who decide whether the police are meeting the expectations that justify their wearing a shield, rather than being constrained to subjugate their view of reasonableness for that of police apologists on the witness stand.

This too is a very imperfect solution, since no one ever suggested that jurors are any better at ignoring prejudice toward blacks, or giving cops far too much latitude in killing. But at least it offers a chance to hold police accountable, rather than merely make a minor shift in the rhetoric.  As for the 20-20 hindsight, so what? That’s how every trial happens, and if it’s a good enough reason to put a non-cop in prison for life, it’s good enough to be applied to a cop as well.

30 comments on “Is “Necessity” The Answer To Stop Police Murder?

  1. bmaz

    Should be noted that what is “already the DOJ policy” has been wholly ineffective at regulating and addressing FBI shootings. The FBI never finds fault in their shootings (was an NYT article a couple of years back with details.
    So the mere incantation of the DOJ standard will not do squat.

  2. Mike

    Acknowledging that informed jurors are a rarity, the jury could choose not to believe the bullshit coming from the expert witness about the cop’s reasonableness.

  3. Andrew Case

    Not sure where he gets the idea that only eight states require officers to give warnings before using deadly force. Garner made warnings required, when feasible, before shooting at a fleeing suspect (and then only when the fleeing suspect poses harm to others). The Ninth Circuit expanded the doctrine to require a warning, when feasible, before any use of deadly force in Idaho v. Horiuchi, 253 F3d 359, 372 (9th Cir. 2011). I count nine states in the Ninth Circuit. Connecticut requires a warning when feasible (CGS Sec. 53a-22(c)). The NYPD Patrol Guide (not a state law, true) requires a warning when feasible (PG 203-12). I’m sure there are more but don’t have time to look.

    For those who are interested, the most insightful discussion on whether a warning is “feasible” is found in Hulstedt v. Scottsdale, 884 F. Supp. 2d 972, 998 (D. Ariz. 2012), issued by Judge Murray Snow more than a year before he found that Joe Arpaio engaged in racial profiling and installed a monitor.

    1. SHG Post author

      A bit of a petty detail, not worth my time to check its accuracy (except Garner did not make warnings required, but only “where feasible”), and offered in a needlessly and inappropriately pedantic fashion. Don’t be that guy.

      1. Andrew Case

        Apologies for tone; I’ll try to be informal. My point is only that most of what he proposes is already required of officers — if only by patrol guide regulation and not by state law. In disciplinary interviews (all my experience is in PD discipline, not criminal charges or grand juries), where they have to justify what force they use (or whether a warning was feasible) they do just what you say they would do. Lesser force is usually pepper spray. They usually can’t use the pepper spray because they walked into the room with the gun drawn instead. So I guess I was trying to agree with you.

        1. SHG Post author

          It’s all there, whether in policy, law (wherever feasible) or opportunity. As we know, all that changes nothing as there is always a ready excuse why it just can’t be done without risking the slaughter of a wonderful officer. And no judge or jury wants to be responsible for a cop dying.

  4. pavlaugh

    I haven’t worked on a case where a cop was a homicide defendant, but I’ve never thought there is a different “test” for police officers in the criminal context. My understanding is that we would resort to the general self-defense statute, which at least in my state (TX) requires the actor to “reasonably” believe force is “immediately necessary.” I don’t think there would be some special jury instruction in a “cop” case compared to a regular old murder case. And “reasonably” generally implies that we consider the facts and circumstances surrounding the situation.

    So as you suggested, a trial would just come down to whether the jury believes the expert witness’s opinion. I’m pretty confused why everyone on the Internet talks about Garner and Graham, Section 1983 / 4th Amendment cases.

    1. SHG Post author

      I spent about 10 emails with a reporter trying to explain why Graham (Garner has nothing to do with it) isn’t precedent in a cop defendant case, but that it de facto becomes the test because of how it plays out in the trenches. They’re not good with ambiguity. You’re absolutely right, the law is no different for a cop than any other shooter, but it still plays out under Graham. Because reasons.

    2. Mark

      SHG may have seen it in Amnesty International’s DEADLY FORCE POLICE USE OF LETHAL FORCE IN THE UNITED STATES, about which he wrote previously. The report states,

      “Only eight states require that a warning be given (where feasible) before lethal force is used, however no state meets the requirement for a warning under international standards: Connecticut; Florida, Indiana; Nevada; New Mexico; Tennessee; Utah and Washington.”

      1. SHG Post author

        You prolly meant this as a reply to Andrew Case, but he wasn’t referring to my saying so, but to Rubenfeld’s saying so in his op-ed. Either way, it’s a little too law-nerdy to worry about, as it doesn’t change anything about his argument.

          1. SHG Post author

            Not your fault. I get to watch stuff spiral out of control here all the time. At least your law-nerdiosity was informative.

          2. Mark

            (I’m just nerdy). The politically difficult solution would be a statute governing use of lethal force and incentivizing the police to use less than lethal force prior to its use. The status quo makes it pretty easy to create a story afterwards to fit the facts after the killing is done.

  5. EH

    Perhaps it’s best to take the Hawaii tactic: deterrence works when it is frequent and predictable (even if it’s very “light”) better than if it is infrequent and harsh.

    The question of when a cop should be convicted of criminal murder is a tricky one, given cops’ unusual situation and necessary discretion to shoot people. And it seems to have few legs practically speaking: nobody really wants to blame individual cops.

    But if we make it relatively simple for the family of a wrongfully-killed person to obtain a civil judgment against the city (not the individual cop or the department, to avoid as much defendant bias) by imposition of an appropriate civil standard, then the departments have a much greater incentive. Obviously it isn’t really the appropriate thing for a murder–but nonetheless it would probably be a lot better in terms of results. And in the context of civil liability for police shootings you can add a presumption of liability, e.g. for deaths in custody, while executing an improper warrant, while unarmed, etc.

      1. EH

        Do you think police are immune to financial incentives?

        Police departments are money hungry and there’s no evidence that they won’t respond to expensive, predictable, ongoing liability the way that everyone else does: reduce liability-causing behavior. OTOH, if the only risk is the super-rare 1:100,000 chance that the cop will actually be filmed and/or convicted, the incentive is gone.

        1. SHG Post author

          Doubling down on cluelessness is no way to go through life. I realize you can’t control yourself, so have someone pull the plug on your computer before you do yourself more harm.

          Edit: Now I feel bad for treating you like the idiot child, so I will explain, briefly, why your notion is moronic. Cop couldn’t care less if the city pays out awards. It happens all the time, in huge numbers, and means nothing to the cop on the beat. It is not a financial incentive. It doesn’t affect the police department. It means nothing.

          Had you suggested the elimination of qualified immunity for police officer (something often suggested), it would at least have been in the ballpark of viable possibilities. But a 1983 without a Monell requirement? BFD. Absolutely nothing. Wouldn’t change a thing.

          1. EH

            Eliminating qualified immunity would be great from a incentive perspective, and I support it, but I think it’s a non-starter from a political perspective.

            In any case, you and I disagree about the incentive effects of departmental budgets. The city doesn’t currently pay out much in proportion to the damage they do, largely because it’s still incredibly hard to get judgments–even civilly. If that changed, which I think would be more politically viable than changing criminal or 1983 law, it’d be some serious money.

            1. SHG Post author

              Payouts for police killings are enormous. They have no impact on police behavior. This isn’t a matter of your feelings on the subject, but fact.

    1. JAV

      Aside from the point that there have been more than a few large settlements with no change to the problem, the idea of trying to pass off the financial and procedural burdens of seeking justice to the family of the victim through a civil suit because prosecuting the officer is too hard is just wrong.

      It a full feelz moment here, but I think it’s wrong to create a system where a police shooting can never be a criminal offense.

      1. SHG Post author

        Criminal prosecution and civil action serve different functions. Putting aside that EH’s grasp of the financial incentives is completely wrong, as is conclusively proven by vast experience, and would fail if the person killed has no family (orphans get murdered for free?), can’t find a lawyer, gets a shitty lawyer, etc., society has a direct and concrete interest in prosecution cops who commit crimes. There is a hard problem with prosecuting cops that demands fixing, and there is no substitute for addressing the problem.

    2. Patrick Maupin

      But if we make it relatively simple for the family of a wrongfully-killed person to obtain a civil judgment against the city…

      then the city has every incentive to increase the length of the automatic vacation the cop gets, to reduce the chances of him killing again.

      Eliminating qualified immunity [is] a non-starter from a political perspective.

      And firing, or even disciplining, cops is often a non-starter from a legal perspective, due to the lopsided union contracts.

      If we ever come up with the political will to fix either of those problems, it’s not inconceivable that the other one will go along for the ride.

      1. SHG Post author

        Why do you take EH’s comment seriously? It was utterly clueless and idiotic. Resist the urge to go down that path.

  6. RollieB

    Not being a lawyer (there’s one in the immediate family however) and evidently not having a legal mind (I’ve been sent to Reddit repeatedly), isn’t the real problem, and therefore the solution, the prosecutorial system? Aren’t prosecutors acting as defense council for the cops, rather than prosecutors, a significant part of the problem? (Reddit here I come, again?)

  7. Pingback: Impunity in Tamir Rice killing intensifies demands for systemic reform - Big Sky Headlines

  8. DaveL

    They could, but not having the judge explain to them that cops are special snowflakes with superpowers they, as mere civilians, could not possibly understand, would probably go a long way towards making that happen.

    1. Mike

      Bullshit is bullshit. If the jury is capable of disbelieving it from an expert witness then they should be equally capable of disbelieving it from a judge. There is a lack of caring and understanding generally that keeps the population uninformed.

Comments are closed.