Your Decisions Are Better After You've Eaten

If you have read any pop-psychology book about willpower in the last decade, you have probably seen this claim: Israeli judges grant parole 65 percent of the time when they have just eaten a meal, and almost 0 percent of the time right before a meal break. The implication is dramatic — your fate, in some courtrooms, depends on whether the judge had lunch.

It is a great story. It also turned out to be a much more complicated story than the original headline suggested, and the way that came out is itself a useful lesson in how to think about behavioral research.

The original study, and the headline that escaped it

The paper was published in 2011 in PNAS by Shai Danziger, Jonathan Levav, and Liora Avnaim-Pesso. They analyzed 1,112 parole decisions from eight Israeli judges over ten months. Right after each meal break, the proportion of favorable rulings was about 65 percent. By the end of each session — just before the next break — it had dropped to nearly zero. The authors interpreted this as decision fatigue: as judges made more decisions in a row, they got mentally tired, and started defaulting to the safer "deny parole" option.

The finding spread fast. It is now cited in business books, productivity podcasts, and judges' own training materials. "Don't make important decisions when you're hungry" became conventional wisdom largely on the strength of this one chart — a reminder that confident framing earns trust independent of evidence quality.

The reanalyses that almost no one read

Within months of publication, two other research teams pointed out something the original paper had not addressed: the order of cases in each session was not random. In Israeli parole hearings, cases that involve a lawyer present go first; cases without a lawyer go last. Cases that the prison's rehabilitation board recommends go earlier. And rejected cases take less time, so deny-decisions naturally cluster at the end of sessions.

A 2011 commentary by Keren Weinshall-Margel and John Shapard, also in PNAS, showed that once you controlled for these scheduling factors, most of the apparent meal-break effect disappeared. A 2016 reanalysis by Andreas Glöckner went further — simulating realistic case-ordering effects, he showed that you could reproduce the original pattern with no decision fatigue at all, purely from how cases were sequenced.

This does not prove decision fatigue is fake. It does mean the famous parole study is much weaker evidence for it than the popular version suggests.

What we do (and do not) know about decision fatigue

Step back from one study and the picture gets harder to summarize cleanly:

Decision fatigue probably exists, but the effect is smaller than the early hype. Roy Baumeister's 1998 "ego depletion" experiments — the foundational work for the whole idea — showed that resisting temptation in one task makes people less likely to persist on another. But a massive 2016 pre-registered replication across 23 labs (Hagger et al., Perspectives on Psychological Science) found the effect to be near zero. Whatever decision fatigue is, it is not the robust, easy-to-trigger phenomenon the early studies implied.

Glucose is not the missing piece. The "your brain runs on sugar so eating restores willpower" story sounds neuroscientific but does not hold up well. Your brain uses about 20 percent of your daily energy and that usage is remarkably stable — strenuous mental work increases it only a few percent. Drinks of glucose sometimes restore performance in lab tasks, but the effect appears to be partly psychological (knowing you had sugar) rather than purely metabolic.

Choice overload is real-er. A more replicable finding is that when people face many similar options in quick succession, they tend to default to the status quo or to a heuristic shortcut — the same pattern as working memory reaching its limit. This shows up in supermarket aisles, online dating, and yes, parole boards. But the mechanism is not necessarily "depleted willpower" — it might just be that comparing N options takes effort, and humans cap that effort at some point.

What the practical advice actually is

So: should you avoid making big decisions when hungry, tired, or near the end of a long meeting?

Yes, but the reason is not quite what the original story said. It is not that your brain has run out of glucose. It is more like this:

• Hunger and tiredness make you more risk-averse and more likely to use shortcuts. The decision you make at 5 pm after eight meetings is not exactly worse — it is just more likely to be the easy default. Sometimes the easy default is right; often it is not.
• Sequence matters more than meals. If you have to make ten difficult decisions in a row, the tenth gets less attention than the first, regardless of whether you eat between them. Spacing decisions out is more useful than timing them around food.
• The trick is to recognize the state, not to game the timing. When you notice you are tired or hungry, the move is to pause important decisions, not to assume your judgment is fine because you ate recently.

A more honest version of the rule

The cleaner statement is: humans are not equally good at deciding all day. We get worse when fatigued, when rushed, and when we have just made a series of similar judgments. Build a process that:

• Pushes routine decisions onto defaults (uniforms, weekly menus, automatic transfers) so the decisions you do consciously make are the ones that matter.
• Spaces out hard decisions instead of clustering them.
• Treats your own confidence as suspect when you are tired — feeling sure is not the same as being right.

And take the parole study with the salt it deserves. The behavioral science it pointed at — that decision quality drifts under load — is probably real. The specific claim that judges deny parole when they are hungry is much shakier. Both things can be true at once. Knowing which is which is the actual skill.