When Unconscious Bias Bears Arms – Part 2
Yesterday I returned to a theme I have considered on several occasions: the universality of human bias, specifically unconscious racial bias. I discussed the impact of the universality of unconscious racial bias on policing and police shootings, and I concluded with an observation that, given that racial bias exists in all of us, we need to be especially careful about how we go about selecting among candidates for police officer positions. Today I comment on one aspect of New York City’s system for selecting police officers.
We should start by considering what we expect from our police officers: intelligence, judgment, high integrity, honesty, calm in the gravest crises, courage under fire, physical fitness, marksmanship, interpersonal sensitivity, detailed knowledge of criminal law and procedure, including all-but-instant recall of every subtlety of Fourth Amendment jurisprudence, a high level of accuracy and detail in reports and testimony, and ability to handle the full wide range of humanity with unflagging equity and respect.
We expect police officers to ensure our safety, and to do so without bias or error, let alone wrong-doing.
And, in fairness, we should consider what we have been willing to spend to buy this skill set. In New York City, we pay starting cops a base salary of $42,819 a year. Maximum pay after many years of service is $78,026. By comparison, a nearby Suffolk County police officer hired since 2012 starts at $42,000 and goes up 8.6 percent a year for 12 years, maxing out at $111,506. (Suffolk County officers hired before 2012 max out at $139,234.) A Nassau County police officer starts at $34,000 but goes up to $107,319 after nine years.
Studies have consistently found racial bias in our criminal justice system generally, and in policing in particular. I recently noted one especially good study that found racial bias in the use of physical force against civilians at every level except the most severe – police shootings. I expressed skepticism that racial bias would be manifest at every level of force lower than shooting, but would just vanish at the highest level. I wonder if the sample size of non-African-American shooting victims was just too small to yield statistically significant results.
Every New York City police officer applicant must pass a written test. We as a society place a lot of confidence in written tests, but a moment’s thought should give us serious pause about the value of a pencil-and-paper examination as a qualifier for armed street patrol. How well can a multiple choice exam select for street judgment, integrity, honesty, and interpersonal skills?
But for better or worse, an exam preference is written into the New York State constitution (article V, section 6), and for an applicant pool the size of the police officer applicant pool there is no cost-practical alternative to a standardized written exam. A candidate who passes the written exam must then undergo a background investigation, a medical exam, a written psychological exam, an oral psychological exam, and a character investigation.
As far as I can tell, the NYPD does not subject police officer applicants to any of the computerized tests for race-based bias that I discussed in yesterday’s post. The NYPD does use the Minnesota Multiphasic Personality Inventory, which is a widely used and highly regarded psychometric assessment of psychology and psychopathology.
The MMPI does not directly test for unconscious racial or other bias, but it does test for a wide variety of psychological features like anxiety and depression, anger, aggressiveness, excitability, “people orientation,” paranoia, cynicism, and antisocial behavior. The NYPD does not disqualify candidates based on their MMPI scores alone, but uses problematic scores to identify areas for special attention in the screening process.
Given the state of the science regarding unconscious racial bias, given the serious risk that such bias accounts for much of the systemic bias in policing, and given the ready availability of easily administered tests for racial bias, it is disturbing that the New York Police Department apparently does not directly test for bias.
Unfortunately, there is reason to believe that even if the Police Department did try to screen out highly biased applicants, New York City bureaucracy would frustrate the effort.
Like any other civil service candidate who is disqualified, a police officer candidate who is disqualified can appeal to the City’s Civil Service Commission. Overall, the most common grounds for civil service disqualifications that are appealed to the Commission are insufficient education or experience and failure of the medical exam. But for police officer applicants, the most common grounds for disqualifications that are appealed to the Commission are psychological and character. Together, psychological and character appeals account for 225 of 252 police officer candidates’ disqualification appeals reported by the Civil Service Commission on the City’s open data portal.
The Police Department’s psychological evaluations are done by its Psychological Services Section, which is staffed primarily with psychologists who have doctoral degrees. Civil Service Commissioners are not psychologists, and none has served as a police officer. Only one of the current five, Charles McFaul, is a career civil servant. Civil Service Commissioners are political appointees, named by the Mayor, without confirmation by the City Council, or by anyone else. The only check on the Mayor’s appointment authority is that no more than three commissioners can be members of the same political party.
On the face of it, frankly, it is disturbing that a lay commission can and regularly does overrule the Police Department on questions of psychological and character fitness of police officer candidates. Looking deeper into it does nothing to allay the concern.
During fiscal year 2016, the Civil Service Commission decided 121 police officer candidate psychological disqualification appeals on the merits, meaning not on procedural grounds. Of those, the Commission affirmed 108 disqualification decisions and reversed 13. During the same period, the Commission decided 98 character disqualification appeals on the merits. Of those, the Commission affirmed 91 disqualifications and reversed seven.
That may not seem all that important – just 10 percent of these decisions were reversals. But that means that, in just one year, the Commission made 20 people eligible to be put on the street under arms despite the Police Department’s considered opinion that those officers lacked the psychology or character required of a police officer.
It’s important to note that the Commission only hears appeals from candidates who have been disqualified. If a candidate is found to be qualified, there is no appeal. Therefore, the incentive for Commission action is inherently one-sided. If the Commission affirmed all of its disqualification appeals, serious question would arise whether the Commission’s function adds value commensurate to its cost. The Commission can only demonstrate that it adds value to the disqualification process by reversing a disqualification appeal.
* * *
The Civil Service Commission’s process for determining psychological disqualification appeals is remarkably unregulated. Neither New York State nor New York City statutes spell out any procedures the Commission must follow, and the Commission’s own rules include just one section, with 11 paragraphs, that covers disqualification appeals.
Two of those paragraphs delineate eligibility to file an appeal, three paragraphs provide deadlines for filing appeals, and three paragraphs state deadlines for action by the Commission – although all three of the Commission’s deadlines are effectively voided by the open-ended qualification “or as soon thereafter as practicable.” One paragraph provides that the candidate can buy a transcript of the Commission’s proceedings.
Only two paragraphs meaningfully constrain the Commission’s appeals procedure. One paragraph says that the Commission may “affirm, modify, reverse or remand” the disqualification decision, which isn’t much of a constraint, and isn’t complied with by the Commission anyway – the Commission dismissed 88 appeals out of 209 decided during fiscal year 2016, which is a perfectly appropriate disposition except for the fact that the Commission’s rules don’t provide for it.
The other paragraph says that the Commission will decide an appeal based on the written record unless it determines to convene oral argument to allow the candidate “to make an explanation and to submit facts in opposition to the action or determination of the City Personnel Director. At such proceedings, the City Personnel Director will be permitted to defend his/her action or determination.”
As a matter of judicial nomenclature, the proceeding described by the Commission’s rules is not an oral argument, but an evidentiary hearing. The Commission’s decisions routinely refer to the hearings as “de novo,” which is Latin for “anew” – meaning that the hearing is held not to review the Police Department’s decision, but to come to a decision fresh, as if the Police Department had never issued a decision. Thus the Commission’s “appeals” are not actually appeals, but trial-like decisions of dubious legal validity under statutes that provide for “appeals.”
It is unclear why a candidate is entitled to submit new evidence on appeal that the candidate did not submit in the first place. In fact, in two of the seven character disqualification reversals, applicants failed to submit any evidence at all in response to the Police Department’s notices that it was planning to disqualify them. This is an open-and-shut case of what lawyers call waiver – failure to assert a claim or a defense at a critical point forecloses assertion of that claim or defense thereafter. Yet in both cases, the Civil Service Commission, completely without explanation, allowed the applicants to appeal Police Department decisions that the applicants had not contested at the Police Department itself.
It is also not clear why, if the Commission thinks the facts in the record are not sufficient, the Commission does not simply rule against whichever party bears the burden of proof – which is, as any lawyer will tell you, the customary way to handle a case where the evidence is insufficient to carry the burden of proof. Nor is it clear why, if the Commission thinks the facts in the record are not sufficient, at the least the Commission does not remand the appeal back to the Personnel Director with instruction to hear and respond to the additional evidence.
The Commission’s rules say nothing about how the Commission decides whether to schedule oral argument – no statement of a legal standard, no listing of factors or considerations to be taken into account. It simply provides for the Commission’s unconstrained discretion.
In the Commission’s character disqualification decisions, the Commission includes a rote reference to the Commission’s “statutory authority” to hold hearings, but there is no such statutory authority. Whatever authority the Commission has to hold hearings comes exclusively from its own rules. A first-year law student knows that a statute is enacted by a legislature, and that a statute is legally superior to a rule promulgated by an executive agency. It is an example of the disconcerting degree of legal sloppiness that pervades the Commission’s disqualification decisions that the decisions’ authors and signatories do not evidently know the difference between a statute and a rule.
Moreover, the Commission’s rules do not allocate the burden of proof on the appeal. Does the candidate bear the burden of proving that the Personnel Director’s decision is wrong and therefore that she is qualified, or does the Personnel Director bear the burden of proving that the candidate is not qualified? Ordinarily on an appeal, the person taking the appeal must prove that the decision appealed from was incorrect. Ordinarily, an appellate body grants broad deference to the fact-findings made in the decision that is challenged on appeal. The Commission’s rules discuss none of this.
Nor do the Commission’s rules specify a required level of proof. Is the question whether the Personnel Director’s decision was arbitrary and without “substantial evidence,” the standard that ordinarily applies in appeals from administrative determinations? Or is the question whether there is a “preponderance of the evidence” for one side or the other, a standard that would apply in a trial such as the Commission’s “de novo” hearings?
Indeed the Commission’s rules are so meaningless they are not even published on the Commission’s Web site.
The Commission’s appeals decisions add little clarity to any of this. The Commission’s decisions routinely assert that most of its appeals are decided without evidentiary hearings – but every single decision that reversed a police officer disqualification involved a hearing before the Commission. By way of explanation for its supposedly discretionary decision to hold a hearing, the Commission in no case offered more than a rote recitation that there were “certain issues that needed to be resolved,” a legally vacuous statement.
The Commission’s decisions on psychological disqualifications include no meaningful discussion of the burden of proof. The decisions typically include identically worded and legally empty statements that “the record does not support a disqualification.” Some of the Commission’s decisions reversing the Police Department’s character disqualifications do state, correctly I think, that the burden of proving character fitness lies with the applicant.
But the burden of proof is only half the issue: the Commission’s decisions, like its rules, give no clue of the Commission’s views on the standard of proof, which lawyers would regard as a critically important appellate issue.
* * *
In addition to being largely unregulated, the Commission’s disqualification appeals process is disturbingly secretive.
The Commission withholds from public scrutiny the names and identifying information of candidates taking psychological appeals, quite reasonably arguing that disclosure of this information would constitute an undue intrusion into the candidate’s personal privacy under New York State’s Freedom of Information Law. If a candidate is identifiably associated with a psychological diagnosis or problematic character trait, public disclosure of that diagnosis or trait could be damaging outside of the context of the candidacy.
I submitted freedom of information requests to the Commission for copies of all of its decisions issued over a one-year period in which the Commission overturned police officer disqualification decision made by the NYPD based on the psychology or character of the applicant. Initially, the Commission took the position that the disqualification appeals process is almost fully confidential, and that no part of the Commission’s disqualifications appeals decisions can be made available to the public. After I appealed, the Commission gave me redacted copies of 13 decisions in which the Commission had overruled psychological disqualifications and seven decisions in which the Commission had overruled character disqualifications.
The redactions showed the tenacity of the Commission’s preference for secrecy. The police officer candidates’ names were redacted, which is fine, but the redactions extended to other areas entirely outside any interest in the privacy of the candidates. For instance, the Commission redacted the expiration dates of the police officer “eligible lists” that the candidates were on before they were disqualified – even though at the end of the decision the Commission recited the approximate amount of time before the eligible list expired. The Commission even redacted its own appeals docket numbers – index numbers that the Commission assigns to appeals for case tracking purposes.
The Commission posts no data on its Web site about appeal outcomes, and it posts a grand total of one data set on the City’s open data portal – a single data set listing the disqualification appeals by type of disqualification, job applied for, employing agency, some dates, and whether the case is closed or remains open. The Commission does not post any information about the “closed” dispositions except the fact of closure – whether the appeals were granted or denied, much less why. The data set is apparently updated only annually – the current data ends last December.
The Commission’s open data posting does include one curious disclosure: the Commission’s posted data set includes civil service examination numbers and expiration dates of resulting eligible lists, even though that information was redacted from the responses to my freedom of information requests on the grounds of personal privacy.
To make matters worse, for those candidates whose disqualification appeals are successful and who become police officers, there is no way for the public to look at those officers’ records later and correlate them with any kind of problems on the job. This is because the “personnel records” of police officers, correction officers, and firefighters are protected from disclosure by New York State law – section 50-a of the state Civil Rights Law, to be specific. Such a correlation would help the public understand how effective the Police Department’s psychological and character screening is, and whether and to what degree the Civil Service Commission’s disqualification appeals function undermines that screening process.
There is currently a small debate going on about reforming section 50-a, but I expect that police unions will retain sufficient influence in the next session of the New York State Senate to block anything meaningful.
* * *
All of the discussion so far has been about process; let’s consider the outcomes of the disqualification appeals. I obtained and reviewed redacted copies of 20 Civil Service Commission decisions reversing psychological and character disqualifications pursuant to the Freedom of Information Law. Many of the decisions were disturbing, in both their results and their rationales.
A large share of the psychological disqualifications that were appealed had to do with excessive alcohol consumption. Police officers are permitted to carry firearms off-duty, and most do so – including, for one disturbing example, while drinking. Whereas the Police Department appears to take past misuse of alcohol seriously, as an indication of a “risk” of future alcohol abuse, the Commission’s approach to the issue is lax, even naïve.
In one case, for example, the Commission dismissed the Police Department’s concerns about the applicant’s drinking with the observation that “there is no evidence in the record of any negative functioning due to alcohol.” In other words, the applicant’s drinking hasn’t caused any harm so far, so let’s put a gun in his hand and see how that works out.
Even where there is a “record of any negative functioning due to alcohol,” the Commission is inclined to overlook it. One applicant had been taken to the emergency room on one occasion due to excessive alcohol consumption, hospitalization being a pretty good record of negative functioning due to alcohol. The Commission dismissed the Police Department’s concerns about this candidate’s risk of future alcohol abuse because it only happened once – “without further evidence of negative history of functioning.” The requirement that the NYPD produce “further evidence of negative history of functioning” sounds an awful lot like shifting the burden of proof of psychological fitness from the candidate to the Police Department. The Police Department – commendably, I think – isn’t willing to take the chance that the “further evidence” the Commission requires will be, say, a shooting.
Furthermore, the Commission thought it important to note, the applicant “took responsibility” for the episode. In the Commission’s view, “taking responsibility” for something apparently consists entirely of admitting that it happened.
This expansive concept of “taking responsibility” pervades the Commission’s decisions. In one case, an applicant gave vague and contradictory answers to questions about his history of alcohol use. The Commission noted that the applicant “took responsibility,” meaning that, at the Commission’s hearing, he “admitted his behavior during the interview.” The applicant told the Commission that “he gave purposefully vague answers [to the NYPD] because he was offended” by his perception that the NYPD was accusing him of alcohol abuse. The Commission is fine, in other words, with a police officer who gives “purposefully vague answers,” as long as the officer has a motive of personal pique, and as long as he “takes responsibility” by admitting his conduct after the fact. I’m sure it will be a great reassurance to the defendant wrongfully convicted on the officer’s “purposefully vague answers” that the officer “took responsibility” by admitting later that he was ticked off at the defendant.
Notwithstanding the importance of police officers’ diligent honesty and attention to detail, the Commission goes to lengths to excuse inaccurate, incomplete, or misleading statements given by applicants to the Police Department. In two cases, the Commission found this excuse to be persuasive: the applicant perceived his interviewer to be “hostile.” One hopes that the defense attorneys’ cross-examination of those officers will be gentler. One hopes that these officers’ commanders will ask for reports only in the most kindly tones.
One applicant told the Commission that his PD interview was problematic because he knew that what he said could be used against him on his disqualification appeal. The Commission accepted that excuse. Let’s hope this officer never has to file a report or give testimony, because inaccuracies there most certainly can be used against an officer.
One applicant had a bad driving record: four speeding summonses and two summonses for driving while talking on a cell phone. But because he showed his “deep remorse” for his bad record, the Commission concluded that he should be allowed to get behind the wheel of a patrol car. And by the way, the applicant’s failure to reveal his driving record on his application was “unintentional”; no biggie.
Finally, the Commission shows a remarkably gullible streak. In one case, the applicant had been fired from a previous job for failure to comply with work rules. The applicant convinced the Commission to overlook the incident because his mother-in-law was dying at the time, and the applicant “freezes up when confronted by death.” Let’s just be grateful that police officers are never confronted with death; we wouldn’t want a cop “freezing up” on the job.
In a domestic dispute, one police officer candidate was accused of kicking down the front door and breaking the security chain after his wife locked him out. The candidate denied it, claiming that he gave the door a “slight push” and the chain just fell out because the door was old. The Commission bought it.
In another case, an applicant insisted that she didn’t know that the person on the phone was a detective, even though the evidence showed that it is standard practice for detectives to identify themselves by name and rank before questioning. Since she claimed she didn’t know she was talking to a detective, the Commission forgave her false statement on her application that she had never been questioned in relation to a criminal investigation.
In yet a third domestic dispute case, an applicant’s wife contended that the applicant had made “a laundry list of threats.” Not so, testified the applicant. In fact, the applicant insisted, he had told his wife that he was doing the laundry and had a list of things to do. Also, the dog ate his homework.
As disconcerting as are the specifics of the cases is the tone of animosity toward the Police Department that pervades the Commission’s appeals decisions. The Commission decisions are replete with pot-shots at the NYPD – by which I mean that the Commission goes out of its way to include criticisms of the Police Department in its decisions that have nothing to do with the Commission’s determination of the appeals.
There are other indications of hostility in the appeals decisions. I mentioned the two cases in which applicants failed to challenge the Police Department’s proposed disqualifications but were nonetheless allowed to appeal the final disqualification decisions. In several decisions, the Commission referred to evidence from other cases to support findings against the Police Department. Referring to evidence not in the record of the case under consideration is a serious judicial no-no.
More than once, the Commission insisted that bad applicant scores on MMPI scales had to be corroborated by actual behavior, because the MMPI is only to be used in a behavioral context. But the Commission did not hesitate to invoke a good MMPI score to excuse or overlook bad behavior. Apparently a good MMPI score is reliable but a bad MMPI score is not.
Of the 20 decisions I reviewed, the three sitting Commissioners were unanimous in 19 cases. In just one case was there a split decision, with Commissioner McFaul dissenting.
Given the quality of its work product, it was surprising to me to learn how long the Commission takes to decide cases. The 20 disqualification appeals cases I reviewed took an average of 637 days to resolve, measuring from the Police Department’s notice to the applicant that it proposed to disqualify the applicant to the date of the Civil Service Commission’s decision. Of the 637 days, an average of 516 days fell between the applicant’s appeal to the Commission and the Commission’s final decision.
For 17 months of work, I expect better from my government.
* * *
As I’ve mentioned, I’ve seen no indication that the Police Department’s psychological and character review of police officer candidates includes consideration of racial bias – but, to be fair, I don’t know for sure that racial bias is not considered, whether by the type of computerized test I discussed in yesterday’s post or otherwise.
I do know that, given the growing body of scientifically rigorous research literature showing that all people internalize racial biases, and given the life-and-death power of police officers over civilians, the Police Department ought to give high priority to screening out highly biased police officers, both during the application process and during each new police officer’s probationary period. And the Department ought to expend considerable effort training its police officers to be aware that they have biases, to recognize them and think about them, and to engage in reflection about them.
As part of a broader effort to reduce systemic racial bias in the criminal justice system, the Police Department should screen police officer applicants for unconscious racial biases, and should disqualify applicants who show especially strong biases. If the Police Department ever does so, it will be critically important that the Civil Service Commission not undermine that effort.
Lives are at stake, and lives matter.