For 10 months, eight judges on Israeli parole boards ruled on 1,112 requests from prisoners. Each working day ran in three sessions, split by a late-morning snack (the researchers noted it down as a sandwich and fruit) and then lunch. A case took about six minutes. Out of that routine came the most famous graph in the research on decision fatigue, and the quiet foundation of nearly every piece of decision fatigue relief advice written since.
At the start of a session, about 65% of requests were granted. By the end, the rate had fallen to nearly zero. After the sandwich, it jumped back to 65%. A sawtooth, three teeth a day. Published in PNAS in 2011, it became the standard proof that judgment runs down like a phone battery, and the reason so many articles tell you to eat something, take a break and make big calls in the morning.
The graph still turns up in talks and books. What happened to it afterward is the more useful story.
Is decision fatigue real? The hungry judges, reconsidered
Within months, two researchers who knew the Israeli parole system, Keren Weinshall-Margel and John Shapard, wrote to the same journal with a plainer explanation. The order of cases was not random. Boards tried to finish every case from one prison before breaking, and prisoners without a lawyer tended to be heard last in a session. Those prisoners also won far less often: in their data, 15% against 35% for prisoners with counsel.
A session that ends with the unrepresented cases will look, on a graph, exactly like a judge running out of patience.
The original authors answered that their result survived the objection. Then in 2016 Andreas Glöckner ran simulations showing that much of the pattern could be a statistical artifact. Favorable rulings take longer to write. A judge who simply manages the clock, and avoids opening a long case just before a break, would draw a similar curve without being tired at all.
The bluntest objection came from the psychologist Daniel Lakens in a 2017 post titled “Impossibly hungry judges.” He worked out that the drop was about as large as the height gap between young men and women. Psychological effects of that size are almost unheard of. If an empty stomach bent trained professionals that far, he argued, the whole working world would visibly come apart every morning shortly before lunch.
A session that ends with the unrepresented cases will look, on a graph, exactly like a judge running out of patience.
Ego depletion and the radishes
Behind the judges sat a bigger idea. In a 1998 study, Roy Baumeister and his colleagues put hungry students in front of chocolate chip cookies and told some of them to eat radishes instead. Afterward everyone worked on puzzles that, unknown to them, had no solution. The radish eaters gave up after about eight minutes. The cookie eaters lasted roughly 19. Resisting the cookies, the theory went, had used something up. They called it ego depletion, and it became one of the most cited ideas in psychology, with hundreds of studies stacked on top of it.
Then psychology started checking its own work. In 2016, 23 laboratories ran the same preregistered depletion test on 2,141 people. The pooled effect was close to zero. In 2021 an even larger effort, led by Kathleen Vohs, one of Baumeister’s longtime collaborators, used 36 labs and 3,531 participants. Its preregistered result was also close to zero.
So the tidy version, one fuel tank that every choice drains, is in real trouble. That leaves anyone tired of choosing in an odd spot. Almost everyone knows the feeling, and its best-known explanation has not held up.
Decision fatigue symptoms, and where they show up anyway
The more convincing evidence is less dramatic. In 2014 Jeffrey Linder and colleagues at Brigham and Women’s Hospital in Boston looked at 21,867 visits by adults with acute respiratory infections to 23 primary care doctors. The odds of an antibiotic prescription rose hour by hour through each four-hour clinic session, ending about 26% higher in the fourth hour than in the first. That held even for illnesses like acute bronchitis, where antibiotics do nothing. Writing the prescription is the easy path. It ends the conversation.
A modest effect, in a real clinic, and it matches what people describe when they say they are decision tired. The symptoms are rarely dramatic. You take whatever option is already in front of you. You say yes so the question goes away, or you put off a small reply for the fourth day running. By evening, what to cook for dinner feels like the hardest problem you have faced all week. Researchers still argue about the cause (boredom, a shift in what feels worth the effort, ordinary tiredness), but the drift toward the easy option late in a long run of choices is real enough to plan around.
One caution. If the flatness lasts for weeks and rest doesn’t touch it, take it to a doctor. Decision fatigue is not a diagnosis, and sometimes it is a polite name for burnout or depression.
How to reduce decision fatigue, whatever causes it
In 2012 Barack Obama told the writer Michael Lewis, for a Vanity Fair profile, that he wore only gray or blue suits. “I’m trying to pare down decisions,” he said, and he pointed to the depletion research as his reason. That research has since wobbled. His wardrobe still makes sense. Fewer decisions means fewer chances to make a lazy one, whichever theory of tiredness you end up believing.
The changes that survive the replication fights are mostly about subtraction:
- Turn repeat choices into defaults. Breakfast, gym days, what you wear on a Monday. Decide once, then stop reopening it.
- Write if-then plans. “If it is a weekday at 7 pm, dinner is one of these three.” A 2006 meta-analysis by Peter Gollwitzer and Paschal Sheeran, pooling 94 studies, found plans in this form had a medium-to-large effect on people actually following through.
- Make the hard call early, before 40 small replies have lowered your bar. The antibiotic data suggests the bar does slip over a long session, battery or no battery.
- Give small choices a deadline and a “good enough” line. Two or three options, a few minutes, take the first one that clears the bar.
Fewer options helps for the same reason. Sheena Iyengar’s jam table in a California supermarket is the classic case, told properly in our essay on the jam experiment and the paradox of choice. Fittingly, that study has its own replication argument, which is worth carrying alongside it.
For most people the biggest source of unnoticed decisions is the phone. Every notification asks keep or dismiss, every feed asks stop or scroll, many times an hour, which is the trap described in your phone runs on System 1. The practical answer we keep landing on, in how to digital detox without deleting everything, is fewer prompts arriving in the first place.
Breaks still earn their place. Take them for the reason that holds up: a long unbroken run of choices tends to end with the easy option, and stopping starts a new run.
Decision fatigue relief without a perfect theory
There is something almost comforting in how the hungry judges story unraveled. A clean graph met a messy room: a prison’s schedule, lawyers who never showed up, the simple fact that a yes takes longer to write than a no. Holding both at once, the familiar tiredness and the unreliable proof, is uncomfortable in the way we wrote about in why most people can’t sit with an unanswered question. It is also more honest than a battery.
You can still act on it. Shorten the list. Decide the boring things once. Put the hard choice where you are least likely to wave it through.
Somewhere in Israel, a sandwich and a piece of fruit got credit, for a decade, for the kinder rulings that followed it. The fruit probably deserves less of it than we thought. The few minutes around it, when nobody asked anyone to decide anything, are still worth keeping.