Justice on a Full Stomach

A new study of Israeli judges reported on recently in the Globe and Mail strongly suggests that if you’re a prisoner looking for leniency, you may want to wait until His Honour has finished his cream cheese and lox.

Researchers studied 1112 rulings by Israeli judges presiding over parole hearings. At the opening of the court session, 65% of rulings favoured the prisoner but the chance of a favourable ruling dropped nearly to zero by the end of the morning session. Amazingly, after the lunch break favourable prisoner rulings jumped back to the 65% level before beginning a steady decline again as the afternoon session wore on.

The researchers found that the rulings did not appear to be affected by the severity of the crime, amount of jail time served or the prisoner’s gender / ethnicity. Aggravating and mitigating factors such as attendance at prison rehabilitation programs and an offender’s criminal record did affect the rulings but the possibility of a favourable outcome remained highest at the start of any given court session.

While the results are humorous on their surface, the study raises a very serious question about the ability of jurists to routinely judge cases in a consistently fair manner. Criminal defence lawyers like myself have long known that shopping around for the “right judge” is an important part of our role as advocates. Now it appears we will have to add the “right time” to our judicial shopping lists as well.


  1. Could there be other explanations for this study?

    1,112 rulings and 8 judges is an extremely small sample size, especially if you want statistically significant results and proper controls for the other variables they looked at.

    Unfortunately I couldn’t find the study in either the April 12 or April 5 edition of PNAS to check their data.

    In the Canadian civil context judges and masters often arrange their docket in order of increasing complexity or length of submissions, an administrative function that could just as easily explain these findings as well.

  2. Omar, the study is reprised in Discover. The PNAS link to the study is
    Unfortunately it seems to be behind a paywall. But the supporting data is available online at