I have updated the prior blog about judicial discretion to include clarification from one of the co-authors on the frequency of releases at the end of sessions. It is not quite as dramatic as in the quoted language — but still dramatic.
If forwarding, please use the version below:
The Guardian is running a potentially very disturbing study about judicial decisions at parole hearings in Israel. The study finds strong correlations between when in a session (relative to food breaks) a case is heard and the outcome. The paper has been peer-reviewed, and is based on a study of 1,000 cases. The correlations are very strong.
The research, which examined judicial rulings by Israeli judges who presided over parole hearings in criminal cases, found that judges gave more lenient decisions at the start of the day and immediately after a scheduled break in court proceedings such as lunch. Jonathan Levav, associate professor of business at Columbia University, who co-authored the paper, said: “You are anywhere between two and six times as likely to be released if you’re one of the first three prisoners considered versus the last three prisoners considered.”
The authors of the peer-reviewed paper looked at more than 1,000 rulings made in 2009 by eight judges. They found that the likelihood of a favourable ruling peaked at the beginning of the day, steadily declining over time from a probability of about 65% to nearly zero [actually, according to one of the authors, usually down to more like 20% to 30%, see note below — RZ], before spiking back up to about 65% after a break for a meal or snack.
Levav said the paper had implications for British judgments. He said: “What we’re finding here is a basic psychological effect, and there’s nothing different between the psychological effect on a British judge and an Israeli one.”
The only other variables that influenced a judge’s ruling were the number of times a prisoner had been to jail and the presence of a rehabilitation programme. Other factors, such as the severity of the prisoner’s crime, prison time, sex and ethnicity tended not to exert an effect on the rulings, according to the paper, which is published in the Proceedings of the National Academy of Science (Abstract free; text behind paywall.
Note: co-author Jonathan Levav notes that “One little thing: although sometimes the probability drops to zero release, this is only occasionally true. Most of the time it drops to about 20%-30% down from 65%, still enormous, but not zero.”
Now, parole hearings are very highly discretionary, perhaps more so than any other decision (at least in the US, where sentencing itself is constrained), and, of course, parole decisions here are made within parole systems and not by trial court judges. (The Guarddian story and the abstract uses the word judges to describe those making decisions about parole. I understand from Jonathan Levav, one of the authors of the paper, that these are trial judges who were presiding over parole boards and that the judge makes the decision on the release, with the benefit of avaiable advice from a criminologist and social worker. What’s really scary is that this is a system that many of us would consider almost ideal.)
But the study at a minimum suggests that judges need to find ways to make sure that they are not falling into similar patterns, particularly in areas with broad discretion. Most of the suggestions I have seen in this area have dealt more with safeguarding against unconscious racial and ethnic bias. It is probably easier to guard against dangers of which we are aware, than the less obvious ones, and this is surely a less obvious one.
It would also be interesting to try to find out what is the underlying dynamic — it may be, for example, that when judges are tired and hungry, they are less likely to take risks — and releasing someone on parole always feels like a riskier decision that denying parole — although from a long term point of view, a denial of parole may do more damage in the long term, it is just not damage to which attention will be drawn. (ABC News has a speculative piece on what could be causing the effect, focusing in part of the food part of the equation.)
Richard Zorza's Access to Justice Blog 
I cannot resist adding this which I just came across today in Posner’s volume, Law & Literature (3rd ed. 2009):
Meanwhile declining from the noon of day,
The sun obliquely shoots his burning ray;
The hungry judges soon the sentence sign,
And wretches hang that jurymen may dine.
From the Rape of the Lock (III.19-22).