Ask the judge (but after lunch)


I’ve just written an entire book on irrational decision making, and still I find this new study startling. It comes from the consistently good Ed Yong, who writes Discover magazine’s “Not Exactly Rocket Science” blog. Read it yourself, but the gist is that judges’ parole decisions are more favorable to defendants when the judges are well-fed. A graphic illustrates the point dramatically, showing a high rate of favorable decisions in the morning, declining steadily to a very low rate just before lunch, then right back at the highly favorable rate after lunch, and so forth. As unsettling as this finding is, it makes sense in terms of what we know about the brain’s heuristic thinking: The deliberating brain gets tired with too many choices, and reaches “choice overload.” When that happens, it defaults to the least taxing choice (the default heuristic), which is simply to say no to parole. We know that much human judgment is not rational, even when we’d like to think so, but it’s nevertheless surprising to see such cognitive bias in those explicitly charged with being deliberate and rational.


  • Arrive2.net

    Exactly what happened in the research seems a little vague since there are thirteen data points on the horizontal (time) axis and I don’t see an explanation of what the points are. It does not appear to me that the data points correspond exactly with the peaks and troughs in the data line. The judge’s lunch and break-time were not exactly fixed, but their data was combined on a single data line with fixed break times. Nevertheless, assuming the conclusion is right it seems to me that this info could be useful in a lot of contexts.

    I’d like to know if the effect is primarily driven by the snack or the break. Should you arrange a snack for your PTA, dissertation committee, or school board, or do you just have to do your best to try to get in the schedule after a break..regardless?

    Bernard Schuster
    Arrive2.net
    Twitter.com/arrive2_net

blog comments powered by Disqus