I was told there's a study that reveals how judges can be swayed by their hunger. Sure enough, a quick search on 'parole lunch' reveals both the original paper and news reports summarizing the study (or other news reports). The graph below says it all. Each circled point represents the first case in one of the three sessions:

  • at the start of the day
  • after morning snack
  • after lunch
study results

The paper concludes:

Indeed, the caricature that justice is what the judge ate for breakfast might be an appropriate caricature for human decisionmaking in general.

Maybe Oxford Press should change their dictionary entry under parole [verb] from

he was paroled after serving nine months of a two-year sentence


he was paroled after the lunch break.

But surely the results are a bit too dramatic? After all, the likelihood of parole approaches zero just before the snack? Maybe the order of cases isn't really random. Of course most news articles weren't revised to take into account the rebuttal. So are factoids born.

(Still, it would be interesting to study whether there is a significant intra-day difference on a truly random sample.)