Depletion and parole
It turns out that being tired isn’t just detrimental to exam performance. One theory in decision making that we are beginning to understand better, Depletion Theory, holds that our ability to make any type of difficult decisions are also adversely affected by fatigue. In most of our day to day lives, being tired at the end of a long work day doesn’t lead to too many terrible decisions, maybe a candy bar here or there, or fast food when we should go for a healthy wrap. However, sometimes the effects are more significant.
Consider the dramatic results of a recent study by Shai Danziger, Jonathan Levav, and Liora Avnim-Pesso, investigating a large set of parole rulings of judges in Israeli courts. Their conclusion is striking. Judges systematically tend to grant parole when they are most refreshed: at the beginning of the day. After that their decisions change and the likelihood of granting parole drops as the number of decisions they make goes up. But the effect is not the same across the whole day — after their lunch break they get some extra energy that rejuvenates them and their decisions look much more like the ones they have made early in the day
It seems that the cognitive burden of making these difficult moral decisions builds such that over time it is easier and easier to simply accept the conservative, status quo decision not to grant parole.
My latest book, The Upside of Irrationality, explores some positive and some negative ways that irrationality plays out in our lives.
Dan,
Doesn’t the size of the effect look fishy? Are judges really 6-7 times more likely to grant parole when well rested compared to hungry and exhausted? Or have I misunderstood the chart?
thomas
I wonder if a smart lawyer can exploit this by placing their clients after food breaks.
Could you please cite the study?
What this doesn’t tell us is whIch decisions were the better ones. People might feel more generous after satisfying their hunger or just taking a break, and err in granting undeserved paroles.
Wouldn’t those judgment calls be made only on a case-by-case basis, whereas such dramatic trends indicate that timing is incredibly significant – - something no one intends to be a factor, no?
These results definitely warrant further study, but I think Sabine has a very valid point: this study does not directly compare the QUALITY of decisions vs. time. Maybe, the courts schedule the cases such that the easy ones come earlier in the day, and right after lunch.
The number of parolees is being Primed by feelings of hunger. As the number of parolees increases the empty stomach feeling also increases — resulting in a feeling that granting parole is increasingly sickening.
I don’t think it’s fair to assume (unless there is evidence to the contrary) that just because meal time is approaching that people are hungry. Some people eat enough (or the right kinds of food) that they aren’t hungry between meals. Some people have steady energy levels throughout the day. Some people feel better a few hours after a meal than directly after a meal because the feeling of being bloated passes. There are so many variables it’s mind boggling!
Making impared decisions when fatigued isn’t a new concept. We’ve all learned this inherently through our school years and reminded of it with much more fiscal consequences during our working years. What still baffles me to this day is that doctors, who makes life and death decisions daily, work long shifts with little to no sleep; the very same people that have read all the studies and should know better. When’s that going to change AMA? I want my physicians rested.
Looks like the morning meetings are the way to go! awesome post, Dan!
the graph is startling and fascinating, but
1) we don’t really know which is the correct decision. only after looking at criminal reoccurrence could that be judged and that would make a more interesting followup.
2) this might have nothing to do with being tired, in a physical sense, but with a numbing or commoditization of people. i’d image the first person i killed in battle would be a lot more shocking than the 10th, etc.
3) this sort of timeline could be very interesting in lots of other areas — are police officiers or soldiers more likely to shoot after being on the job more than halfway through their day?
Well that’s sobering information. I wonder what other high-stakes decisions are impacted by that effect, and how the short-term cumulative effects might play out in other settings.
Just did a paper late last year on “negotiation fatigue” for my b-school negotiations class – findings consistent, but wish this study had come out before then! Here’s a link to the citation of this paper: http://www.pnas.org/content/108/17/6889.short
Though we did reference this awesome study by Dr. Ariely and our professor, Dr. Schweitzer: Mead, N., Baumeister, R., Gino, F., Schweitzer, M., & Ariely, D. (2009). Too tired to tell the truth: Self-control resource depletion and dishonesty. Journal of Experimental Social Psychology. 45(3), 594-597
There is more detailed information here:
http://blogs.discovermagazine.com/notrocketscience/2011/04/11/justice-is-served-but-more-so-after-lunch-how-food-breaks-sway-the-decisions-of-judges/
Neato. Thank you.
The best remedy I have come across for fatigue at work: playing ping pong.
I have experienced two types of fatigue at work. When I get tired during the day, I go on “auto pilot” and just stop thinking. When I show up tired, I have a hard time adjusting when things go wrong; I get irritable; I fall apart. Maybe it’s the same behavior, just different events based on the time of day.
The peaks and valleys are interesting, but that last lunch break. Yikes. Straight down.
Any arbitrary decision, not Evidenced Based, is possibly influenced by outliers – “human” errors , unfortunately very influential to any rational decision ,correlated with arrogance and hyper selfconfidence.
Selfconfidence is irrational autoimpression and not the Value like knowledge,umility,ability to listen and mediate any important decision expecially regarding the destiny of other persons.
Have a good Meditation Decision’s Outcome!
I’ve read elsewhere, probably in your published books, that this phenomenon–outcomes having more to do with time of day than with relative merit–applies in other fields.
When it comes to parole, I agree with the other commenters who said, “we don’t really know whether “release” or “incarcerate” is the better outcome for any of these cases.” What we do know from the data is that the decisions are almost certainly not fair, nor based on the individual merits of any one prisoner’s case, rehabilitation, or likelihood of recidivism.
Has Charles Manson’s parole hearing ever come up at the top of the day’s schedule? Or does the bailiff make sure it appears at the end of the day?
I wonder if the nature of the convictions matters at all. I imagine that Israel doesn’t have as many “victimless” (use the word advisedly, because fewer US convicts are in for “victimless” crimes than they would like us to believe) drug convictions than the US system.
I can imagine also systems of “palate-cleansing-equivalent” activities, not unlike ping pong, or juggling, or watching kittens, that could clear the murk and ugliness and heaviness of a parole hearing out of a panel’s collective mind so that each case had a clean slate. But that is not going to fly once the tax payers hear about it.
The trouble is, the Department of Corrections is a heavy place, with the least empowered constituency of any state or federal agency. Nobody cares except for a few families, and few people care much about them. It’s not like tornado victims. We default to, “If they hadn’t been done the crime they wouldn’t be in this position, so who cares that the parole system isn’t fair? There are bigger problems in the world.”
Maybe this should be part of the training process for judges ( or retraining).
I would have thought that there must be quite some variability between the judges.
Another ( more humorous) suggestion would be to have randomly generated breaks during the day.
Thanks
Colin
This seems to be evidence in favor of laws that restrict judicial discretion generally, like mandatory sentencing and things like that. I’m generally opposed to such laws since they tend to come along with what I think of as exceptionally harsh mandatories, but this post has given me some serious pause to consider which is the greater evil – no judicial discretion to consider individual circumstances, or judicial discretion applied unevenly (and thus unfairly) even subconsciously caused.
Were the cases randomly distributed (as opposed to taking the hard cases first, and then filling in the schedule with not-a-chance decisions as lunch approaches)?
Outrageous! Very tempting to accept the result… But, the study should be conducted in different settings before concluding anything.
They really would need an accurate way of controlling for different cases being listed at different times. I am a judge’s associate and I generally decide the order of the matters that come on during the day. Like every other court administrator, I try to list brief matters first and longer matters last so that less people are left waiting around all day. The very first matters are always the consent matters. In the case of parole, a consent matter is inevitably one where both parties consent to parole being granted (the applicant is hardly going to consent to parole being refused!). Even if there is no consent, parties usually give a fairly honest estimate of the time they will require, and it always take a lot longer to argue an unarguable case!
Deliberate listing may have been controlled for, but I would be interested to know how robustly. Nonetheless, the results do make sense. Even the most wise and careful judge is still as human as the rest of us.
I have to wonder if being refreshed really has anything to do with it. Judges may instead tend not to grant parole if they have already granted someone else parole in that session, but if they haven’t granted anyone parole yet, they may be more lenient. Judges probably do not want to let out too many people (and be seen as too lenient) but also not be so tough that no one is granted parole. Meal times might only function as mental partitions for the day and eating and being refreshed may have little influence.
I noticed that the pattern of the judgments seems similar to the pattern of negative to positve mood measured using Twitter data: http://www.nytimes.com/2011/09/30/science/30twitter.html. Perhaps the depletion or decision fatigue pattern is very wide spread, affecting a lot of people in many ways.