From The Washington Post:
A friend of mine was suffering such severe back pain that it was difficult for him to walk or stand. He consulted three doctors about the best course of treatment. The first was adamant that he needed surgery right away. The second advised my friend that he didn’t need surgery and that if he continued physical therapy, his condition would improve gradually over the coming months. The third prescribed strong steroids and recommended that, if his condition didn’t improve in a month, then he should have surgery. My friend followed the third doctor’s guidance, and it seems to be working. But he was mighty upset and confused by all those clashing perspectives. And he is still unsure whether that third doctor’s approach is the right one.
This undesirable variability in professional judgment is an example of noise, the ubiquitous and often-ignored human failing that is the focus of this well-researched, convincing and practical book. “Noise: A Flaw in Human Judgment” was written by the all-star team of psychologist and Nobel Prize winner Daniel Kahneman, former McKinsey partner and management professor Olivier Sibony, and productive legal scholar and behavioral economist Cass Sunstein. Kahneman won the Nobel Memorial Prize in Economic Sciences for his pathbreaking work with Amos Tversky on systematic biases in judgment. It prompted armies of psychologists and behavioral economists (including Sibony and Sunstein) to study the causes and remedies for many such faults, including overconfidence, stereotyping and confirmation bias — or seeking, remembering and placing excessive weight on information that supports our beliefs.
The authors kick things off by distinguishing between bias (systematic deviations) and noise (random scatter). The book then sustains a relentless focus on explaining and documenting the wallop packed by the simple and omnipresent error of noise — and what decision-makers can do about it. It blends stories, studies and statistics to make a compelling case that noise does at least as much damage as bias: undermining fairness and justice, wasting time and money, and damaging physical and mental health.
Kahneman and his colleagues show how unwanted variation in judgments (evaluations) and decisions (choices) creates “noisy systems” — which plague professionals including criminal judges, insurance underwriters, forensic scientists, futurists and physicians, who routinely make wildly varied judgments and decisions about similar cases. Systems are noisy, in part, because different professionals apply different standards. There is disturbing evidence, for example, that when multiple physicians evaluated identical cases for evidence of heart disease, tuberculosis, endometriosis, skin cancer and breast cancer, they agreed on diagnoses only about two-thirds of the time. In such noisy systems, errors add up rather than cancel each other out. As the authors put it, “If two felons who both should be sentenced to five years in prison receive sentences of three and seven years, justice has not, on average, been done.”
Systems are also noisy because, over time, the same professionals apply inconsistent standards. To illustrate, a study of 22 physicians who each examined the same 13 angiograms two times, several months apart, found that they disagreed with themselves between 63 percent and 92 percent of the time. To explain such swings, the authors use research on “occasion noise”: Fluctuations in a person’s mood, fatigue, physical environment and prior performance that are (objectively) irrelevant, yet shape judgments. Like the study titled “Clouds Make Nerds Look Good,” which examined 682 actual decisions by college admissions officers: They weighted applicants’ academic strengths more heavily on cloudier days and applicants’ nonacademic strengths more heavily on sunnier days.
Link to the rest at The Washington Post
PG will certainly second the notion of judges hearing criminal cases being subject to “noise” in the OP.
Search for “inconsistent sentencing” on Google. You will find that a great deal of research and scholarship devoted to studying why criminal defendants that are convicted of or plead guilty to seemingly similar crimes receive widely divergent sentences from judges who are acting under similar laws.
More severe sentences imposed on racial minorities than on white defendants are the hot button in this legal field. On average, racial minorities in a given jurisdiction receive harsher punishment than non-minorities for the same crimes.
However, female criminals of all races are treated much more leniently than male criminals of the same race who are convicted of the same crimes.
“In 2012 Sonja B. Starr from University of Michigan Law School found that, controlling for the crime, “men receive 63% longer sentences on average than women do,” and “[w]omen are…twice as likely to avoid incarceration if convicted”, also based on data from US federal court cases.” See SSRN
Having represented a judge in one case and a handful of attorneys in other cases in a past professional life, PG can assure you that going to law school or being elevated to the bench does not change a human into a machine.
Determining a suitable sentence for a criminal is anything but a precise process. There are zillions of factors that can come into play – far too many and far too subtle for any legislature to craft written legislation that addresses all of them.
Is the defendant remorseful or not? How remorseful or how not-remorseful? PG can attest that reading a defendant’s words as captured in a court transcript and watching and listening to the defendant as they speak is two different things.
Just as some people are better at fooling others when they are lying than other people, some people are better at conveying truthfulness than others, even when both tell the truth.
There are also different purposes for punishment. Simply punishing an individual for committing an evil deed is the most obvious. This is the simple application of the rule evil deeds must be penalized for fundamental justice in a society.
However, assuming that the defendant will be released from confinement at some future time, deterring that defendant from committing another crime is another purpose. Deterring other individuals from committing similar crimes is another purpose for punishment. Enforcing broad societal standards of behavior is another.
To take an extreme example, if everyone who exceeded the posted speed limit by five miles per hour or more were sent to prison for a minimum of 30 years, that would be quite a powerful deterrent. However, there is an inherent societal value that the punishment should fit the crime.
If our speeding driver received the same sentence as someone who killed another person with a knife, we would not feel that such a system was just and fair.
PG suggests that the “noise” described in the OP and what sounds like an interesting book is not the only reason why human judgment is complex to understand and assess fully.
I seem to remember reading that prisoners had a significantly higher rate of being granted parole if they appeared before the board in the morning vs. the afternoon.
It seemed to make sense – by the end of the day the board members are tires of b.s. – but I’d hate to have an appearance, know this, and to find myself scheduled in the last time slot of the day.
A – Government efficiency and careful consideration of each petition in action. This is the type of information a savvy criminal lawyer would love to know so she/he could take steps to get their client a hearing first thing in the morning.
PG is no longer a savvy criminal lawyer (and barely justified that description in olden days when he occasionally was assigned to defend an indigent defendant), but if he had this knowledge and received a schedule that had his client appearing before the parole board, he would have quickly found a conflicting court obligation to move the hearing to the next date the parole board was hearing petitions to see what the schedule looked like then.
Who knows, if a hearing originally scheduled in the afternoon in July had to be postponed, maybe it would show up at the top of the morning hearing list next time around in August because it had been carried over from the previously scheduled hearing date.
Of course, the down side of “move to the next day the board is hearing petitions” sometimes runs into parole boards that only meet once or twice each year… and I only wish that was a hypothetical, it was the default in both Illinois and Missouri in the late 1990s outside the major metropolitan areas. Just coincidentally, that meant that inmates with colorable claims of innocence who had been convicted primarily on eyewitness testimony were often clustered in the same locations and hearing schedules; and the less said about the other relevant demographics, the better.
Or you could just have the system for military prisoners. There isn’t any parole; any modification of a previously confirmed sentence requires either a full hearing before the respective service’s convening authority in DC, which doesn’t provide a right to live hearing (it’s all on paper), or a presidential pardon or commutation (ditto). (Even a judgment by the Court of Appeals for the Armed Forces that reduces a sentence without voiding the conviction is not self-executing.) The gates of the facility at Fort Leavenworth should read “Abandon all hope ye who enter here” on them.
You obviously know more than I do about parole board schedules than I do, C.
I had forgotten how much easier it was to practice law in places where you could actually get a hearing/trial to resolve a dispute within a reasonable time. “Well, we’ll see what the judge says,” was a useful negotiation strategy.
Running a retail practice where there were a lot of poor people who definitely couldn’t afford a five-year lawsuit with all the trappings also made legal life easier and more rewarding. I don’t think the ultimate justice reached for most disputes was impaired in any significant way either.
Once upon a time, I was prosecutor, judge, jury, and executioner — that is, a commanding officer (squadron — equivalent to a battalion — level). The fundamental problem with imposing punishments is that no two crimes are really identical; no two defendants are really identical (and the less said about “equating victims,” the less time it will take me to pull my foot out of really delicate places on the people who try to treat victims as widgets). The example PG gave is from 1960s and 1970s criminology, and it was known to be wrong by the 1970s… so, naturally, the war-on-drugs/law-and-order movement enshrined it, at the federal level, in the grievously flawed Sentencing Guidelines.
The entire concept of “uniform sentences” makes sense if, and only if, the only purpose of criminal punishment is retribution. Any 1L taking the first-year required class in Criminal Law knows otherwise; and that still can’t account for the differences in circumstances or victims. What I’m amazed at is that judges can come so close to agreement!
Aside concerning Professor Sunstein: He’s one of the few law professors I know who pays attention to the facts and data first before engaging in grand theorizing. (I’ve edited some of his work for a law journal, so I’ve seen his process.) I’ve also seen how his own views have evolved over time. None of this is to say that I agree with him all that often; it’s just that I can respect our differences. I’m quite disappointed by a preceding commenter’s ideology-based ad hominem attack that appears to have been made with no real knowledge of Professor Sunstein’s body of work. I therefore propose a suspended sentence of revocation of literary license for 90 to 120 days as an appropriate Guidelines-based punishment, to be expunged upon reading at least three of Professor Sunstein’s books, at least one from each of the past two decades.
I must admit that I have only read two – “Conformity” and “The Cost-Benefit Revolution.”
That was sufficient, for me. I note, once again, that what facts and data you pay attention to is crucial – there are indeed facts and data in those two books, and I assume also in this latest – but they do not by any means paint a complete picture, but instead the picture that the proponent of “behavioral economics” wants to see.
But PGs front parlor is not the place to argue such – so this is my last comment on the subject.
I would expect this from anything that has Cass Sunstein attached to it. These are the same social “science” methodologies used to “prove” that, for instance, women only make 80% of what men do in the workplace. Find two measures that give you the result that you want; once you have them, don’t look at, or report, anything else that might disprove your conclusion.
Yes, the “victimized” minorities do tend to receive, on average, harsher sentences for the same criminal conviction. But they tend to be convicted for the same charge at the same rate as the “non-victimized” population. Well, I do recall a study from several years ago that showed a slightly higher conviction rate for violent crime charges of blacks – where the jury was composed entirely of blacks from their neighborhood. Juries are supposed to be unbiased, but I can understand those from high crime neighborhoods being predisposed to getting the criminals put away.
Sentencing is also supposed to be unbiased – as in not affected by race, ethnicity, gender, etc. But, unlike the jury, the prosecutor and the judge are allowed to consider the “whole picture,” not just the instance of the crime on which the person was convicted. They, unlike the jury, can consider whether there have been previous charges and/or convictions, whether the criminal is likely to re-offend without a major wrench to their situation, etc. This is why the same crime, committed by a repeatedly offending black male (say third degree burglary), will often receive a much harsher sentence than a first-time white female. Controlling for these factors in their study, however, would have resulted in very little difference in sentencing – but that result was unacceptable.
Not to deny that sentences are frequently skewed by factors that should not have been considered – but that is why we have an appeals system.
So we should have a 100% agreement between “experts”? That sounds like a recipe for fragility – who enforces that perfect agreement? What happens if the perfect solution, well, isn’t so perfect?
The details that go into sentencing are one example. But even in health, well, yes, while it might be nice if doctors always had perfect judgement, but remember that the details of each injury or disease are different, people have different risks for surgery, respond differently to treatments, and so on.
100% agreement? Two rationales for that:
1- Risk avoidance. If everybody agrees on a “solution” then nobody can be blamed when, inevitably, the “consensus” proves wrong/inadequate.
2- Suppression of dissent. No need to actually be right or to defend the “consensus”. Safer for those taking that position. See above.
Ideologues don’t like challenges. It makes people think. Bad for the “leaders”.
100% agreement is acceptable if (a) the subject is one where expertise really exists and is able to provide precise results and (b) the “experts” actually do have the necessary expertise. If the second criterion is met, 100% agreement is normal in mathematics and many parts of physics and chemistry but suspect in most other cases. I have no doubt that all three doctors in the example have considerable expertise in the treatment of back pain, but it seems – as is common in medicine – that a precise answer as to the best treatment does not exist and that all three doctors failed, probably due to overconfidence in their own answer, the desire to appear infallible or not trusting the patient enough to present them with the alternatives.
In my experience, when I read that “an expert says” I find that the subject is often one where no expert knowledge exists – or if it does only allows broad generalisations and qualified results – whilst the “expert” often lacks the necessary sub-field knowledge. Plus what is being described as “noise” frequently seems to just be a combination of the lack of a definite answer with people being afraid to admit this.
As for PG’s point about a “speeding driver received the same sentence as someone who killed another person with a knife” I cannot but recall Larry Niven’s stories where the third speeding offence would have resulted in the driver becoming an organ donor, to the general applause of the population awaiting transplants.
“Plus what is being described as “noise” frequently seems to just be a combination of the lack of a definite answer with people being afraid to admit this. ”
Or it conflicts with ideology/bias/preconceived notions. A definitive answer may exist but be unpalatable. Me, I’m reminded about the line (Asimov? Clarke?) that an expert is a person who knows more and more about less and less until eventually they know everything about nothing. Demanding 100% is oversimplifying. The proper answer to most everything outside academia is: “It depends.”
“I cannot but recall Larry Niven’s stories where the third speeding offence would have resulted in the driver becoming an organ donor, to the general applause of the population awaiting transplants.”
He nailed that one. It is already happening in China.
Just a matter of time elsewhere. They’ll start with “deniers”.
Mike, you unintentionally hit a trigger issue for me. Think about this for a moment:
9mm Glock G17 Gen 4, 9×19 standard round
muzzle velocity 300 m/s
projectile mass 115gr = .075 kg
kinetic energy = .075(300)(300) = 6750 Nm
2016 Prius driven by an impaired Deadpool
velocity 42mph = 18.8m/s
projectile mass 1460 kg (curb weight plus est. driver weight)
KE = 1460(18.8)(18.8) = 516022 Nm
2016 Cadillac Escalade driven by an impaired celebrity
velocity 42mph = 18.8m/s
projectile mass 2650 kg (curb weight plus est. driver weight)
KE = 2650(18.8)(18.8) = 936616 Nm
Which is to say that Deadpool is only applying about eight times the kinetic energy to the victim as is that round from a Glock; the nameless celebrity is worse.
Is this picking perhaps too much on problems with the specific comparison? Probably. And that’s my point (says the former premed who spent waaaaaaaay too much time dealing with victims of drunk drivers).
Well, actually, math CAN be hard:
516022 / 6750 = 76.4
Lack of caffeine (after going water-only in preparation for medical testing) obviously impairs my math more than I wish…