Opinion by Radley Balko in The Washington Post
Feb. 14, 2019 at 11:57 a.m. PST
Last November, I published a long investigative piece about the Little Rock Police Department. The story centered around Josh Hastings, a former LRPD officer who was hired despite lying about, then admitting to having attended a Ku Klux Klan meeting. Hastings then racked up a stunning disciplinary record before he shot and killed a black teen named Bobby Moore in 2012. As I wrote at the time, the closer you look at Hastings’s career at LRPD, the more you see a department rife with police abuse, racism, questionable shootings, and a lack of proper training and discipline.
Moore’s mother, Sylvia Perkins, sued and won a judgment against Hastings, but the unemployed officer will likely never be able to pay. So the family also sued the former LRPD chief, the LRPD and the city of Little Rock under what’s called a Monell claim. That means they would need to show that the LRPD had engaged in a pattern or practice of poor training, lack of discipline or a general culture of abuse that would have made a shooting like that of Bobby Moore foreseeable. I won’t rehash all the problems at the department; I’ll just sum it up this way: In 15 years of reporting on policing issues, I haven’t seen many departments as bad as the LRPD. As Chiraag Bains, a former official in the Obama administration who investigated civil rights abuses, put it, if ever there was a successful Monell claim, this is it.
And yet.
In 2017, a federal district court judge dismissed Perkins’s lawsuit. Perkins appealed to the U.S. Court of Appeals for the 8th Circuit. Last week, a three-judge panel for that court issued a unanimous ruling: Perkins had failed to show a pattern and practice of abuse at LRPD. But in dismissing the lawsuit, the panel repeated many of the mistakes the district court judge had made. More troubling, the panel also made a factual error that cuts to the heart of why Perkins filed her lawsuit in the first place.
Perkins, represented by attorney Mike Laux, had submitted over 200 pages of documentation of abuses, failure to train, and failure to discipline officers at LRPD. The panel’s ruling was less than 19 pages. The opinion didn’t address the bulk of the seemingly endless examples in Laux’s brief, but instead picked and chose a few, then attempted to explain them away. Yet even here, the opinion was not only unconvincing, it also managed to botch the facts.
Mostly, the ruling reiterates the 2017 findings of federal District Court Judge Brian S. Miller. As Bains and University of South Carolina law professor Seth W. Stoughton said in interviews for my original article, Miller’s opinion sided with the LRPD on every disputed fact. During summary judgment, he’s supposed to consider disputed facts in a manner favorable to the plaintiff. Miller settled factual disputes and credibility judgments in favor of the police.
Ultimately, Miller concluded that Perkins had failed to show a pattern of constitutional violations. Why? Because the LRPD had investigated most of the incidents she cited, and determined that its officers hadn’t done anything wrong. Never mind that one of Perkins’s key claims — one for which she provides ample evidence — is that the LRPD’s internal investigations themselves are part of the problem. Miller’s ruling is a tidy bit of circular logic that, if applied in every case, would mean that any police department could be cleared of engaging in a pattern of constitutional violations so long as the same department managed to investigate the incidents alleged to compose the pattern, and find that its officers did nothing wrong. Put another way, the authority in determining whether a police department is engaging in unconstitutional practices can’t be that very same police department.
When Perkins’s attorney Laux pointed to clear examples of officers who had received inadequate training — some by their own admission — Miller simply pointed to the number of hours of training they had received, as if elapsed time in a training class is all that matters. Even when Laux managed to get former LRPD officials to admit examples in which the department had cleared officers who should have been disciplined, Miller noted that merely meant the former officials had “accepted a set of facts and reached an incorrect decision.”
The appeals court panel ruling repeats most of these errors, but then adds some of its own. When discussing the fact that Hastings once body-slammed a homeless woman, for example, Judge Roger Leland Wollman writes, “Other than his use of the word ‘body slam’ to describe the takedown, the record does not support an inference that Hastings’s use of force was unconstitutional or that [LRPD chief Stuart] Thomas had noticed that the report was false.” What Wollman doesn’t note is that Hastings lied about the incident, and that his superior then falsely reported that Hastings “wrapped his arms around” the woman and “placed her on the ground” — a description clearly at odds with the term “body slam.”
Hastings himself later admitted in a deposition that his supervisor’s description of the incident was inaccurate. In other words, according to Hastings himself, the final report upon which Wollman bases his assessment of the incident clearly contradicts what actually happened, and in a way that greatly reduces Hastings’s culpability. Here again, in adjudicating whether a police department committed violations, a court is deferring to reports issued by the very police department in question.
Perkins and Laux documented numerous other incidents in which police reports contradicted witness accounts or surveillance video — always to the benefit of the police officers. But Miller and the appeals court panel see no pattern, because for each time an LRPD report contradicts how officers are depicted in those witness accounts or videos, the respective courts either defer to the LRPD report or simply don’t bother to discuss the incident.
But the appeals court opinion from last week also made a more egregious error: Wollman’s opinion got a key fact wrong, and in a particularly damning way.
The fact in question has to do with the Early Intervention System (EIS) the LRPD used — or at least was supposed to use — to flag problem police officers. Early warning systems like EIS issue an alert each time an officer exceeds a preset threshold in one of several categories, such as use-of-force incidents, citizen complaints, tardiness, use of sick leave and so on. Studies have shown that, used properly, these systems are pretty good at predicting things like civil rights abuses and unjustified shootings. Intervention in the form of counseling, discipline or mentoring can bring wayward officers into line.
But “used properly” is key. Here’s how the system was used at LRPD:
Between July 2010 and the end of 2015, the EIS issued alerts on potential problem officers more than 700 times. Just two of those resulted in intervention, according to department records. Both were for abuse of the sick-leave policy.
For early warning systems to be effective, supervisory staff need to be well-trained on the system and receive ongoing training as the software is tweaked to suit the needs of the department. In a deposition, Thomas, the former chief, testified that the department often failed to send someone to the annual training conference to learn about updates to the system. The sergeant that Thomas assigned to implement the system later testified that he hadn’t even read the Justice Department manual on how to operate it. Even some instances of deadly force somehow never made it into the system.
“To have just two interventions after over 700 alerts, and for low-level misconduct, is just drastically out of proportion,” said Bains, the former senior counsel for the Justice Department’s Civil Rights Division. “It signifies a deep problem in the department. It’s a sign of a broken system.”
Clearly, the LRPD wasn’t paying much attention to EIS. If it had, the department may well have prevented many of the incidents laid out in Perkins’s court filings, including the death of Bobby Moore. Incredibly, the LRPD’s inattention to the EIS continued right up until the time Perkins filed her lawsuit. In its own court filings, the city of Little Rock claimed that Hastings had been flagged by the EIS just three times. Laux went through the record himself, and discovered Hastings had been flagged 19 times, a staggeringly high number given his relatively short tenure on the force. The LRPD paid so little attention to the EIS, they had missed 16 of the 19 alerts for Hastings.
It would be bad enough if the 8th Circuit panel had merely considered the fact that the LRPD ignored those other 16 alerts and decided it didn’t matter. But the court does something far worse: The court ignores those incidents, too. Wollman writes in his opinion: “Hastings triggered three EIS alerts during his career.”
So that’s that. The rest of the opinion then proceeds on this alternative fact. It sure makes Wollman’s job less difficult. It’s much easier to forgive a police department for failing to remove an officer after three alerts than after 19. And it’s much easier to forgive the department if, in doing so, you don’t bother to mention that the department simply overlooked or ignored those other 16 alerts.
Wollman then considers those three alerts, and explains why each fails to support the Monell claim. After brushing aside the three alerts the LRPD decided to recognize, Wollman then adds, for emphasis, “Hastings did not trigger any further EIS reports during his time on the police force.”
But he did. Lots of them. That the LRPD didn’t seem to notice is precisely the problem. And that the 8th Circuit panel also didn’t notice ought to make us question whether the courts are carefully considering these claims or merely looking for reasons to dismiss them.