Decision Fatigue: What the Research Actually Shows and the Operator Protocol

Decision fatigue is real as an experience and shaky as a science. An operator has to hold both of those facts at once.

The popular story is simple: willpower is a battery, every decision drains it, and by late afternoon you are making worse calls. Parts of that story have not survived contact with large, preregistered studies. The tiredness you feel at 6 p.m. after forty small choices is not imaginary. But the mechanism you were sold is less certain than the books suggested.

Here is what decision fatigue means, what the research actually shows, and a protocol that works whether the science settles for or against the strong version.

What is decision fatigue?

Decision fatigue is the idea that making decisions is mentally demanding and that, after many of them, the quality of later decisions deteriorates. That definition comes almost word for word from a 2025 field study in Communications Psychology that set out to test it.

The concept grew out of the "strength model" of self-control, often called ego depletion. The model treats self-control as a limited resource that runs down with use.

The founding decision-fatigue paper is Vohs and colleagues, 2008, in the Journal of Personality and Social Psychology. Across four lab studies, participants who made choices among consumer goods or course options showed less self-control afterwards than participants who only thought about the same options. They showed less physical stamina, less persistence, more procrastination and worse arithmetic. The authors also reported that choosing was more depleting than deliberating, or than implementing choices someone else had made.

That is the claim. Now the evidence.

What the decision fatigue research actually shows

The hungry judges, and the dispute

The most cited example is a 2011 PNAS study of Israeli parole boards. Danziger, Levav and Avnaim-Pesso analysed 1,112 rulings by eight judges. The share of favourable rulings dropped "from ≈65% to nearly zero" across each decision session and returned to about 65% after a food break.

It became the go-to proof of decision fatigue. It is also contested.

In a letter in the same journal, Weinshall-Margel and Shapard argued that case order was not random. The board tried to finish all cases from one prison before a break. Unrepresented prisoners, about a third of cases, prevailed only 15% of the time against 35% for prisoners with counsel, and within each session they "usually go last."

The original authors replied that the pattern persisted when representation was included in their models. Then, in 2016, Andreas Glöckner published simulations showing the ordering effect could be produced by a statistical artifact: favourable rulings take longer, so a judge who avoids starting a long case right before a break would generate a similar curve without any fatigue at all. His conclusion was that the effect's magnitude is overestimated.

The honest summary: the hungry-judge study is a striking pattern with at least one credible non-fatigue explanation. It is not clean proof.

The ego-depletion replications

The bigger blow came from replication.

In 2016, a preregistered replication led by Martin Hagger and Nikos Chatzisarantis ran a standardized ego-depletion protocol across 23 laboratories with 2,141 participants. The pooled effect was d = 0.04, with a 95% confidence interval from −0.07 to 0.15. That interval includes zero.

In 2021, a second multisite preregistered test, this time led by Kathleen Vohs herself, ran 36 sites and 3,531 participants. Confirmatory tests were nonsignificant (d = 0.06), and the Bayesian analysis found the data four times more likely under the null than under the expected effect. The authors' own words: "preregistered analyses did not find evidence for a depletion effect."

One detail from that paper is worth keeping. Exploratory tests suggested the effect was larger for participants who reported more fatigue. Exploratory means it is a lead, not a finding.

The field test

The 2025 Communications Psychology study took the question out of the lab. Andersson and colleagues analysed 231,076 calls handled by 174 nurses at a national telephone triage service, a job that is hard, repetitive and decision-dense. They looked for the predicted signs of fatigue: drifting toward personal defaults and assigning higher urgency later in shifts.

They found none. The Bayes factors favoured no difference in every main test. Their title says it plainly: no evidence for decision fatigue.

The number you should stop repeating

You have probably read that adults make 35,000 decisions a day. Writing in Psychology Today, psychologist Eva Krockow asked how that estimate "was derived in the first place." No study is attached to it. Leave it out of your decks.

So is decision fatigue real?

Three things can be said with confidence.

• The strong version is not established. A simple, reliable "willpower battery" effect did not replicate in two large preregistered efforts.
• The landmark field example is disputed. The parole-board curve has a credible scheduling explanation.
• The subjective experience is not in question. People report feeling drained by long runs of choices. What is unsettled is whether that reliably degrades decision quality, and through what mechanism.

Here is the operator's point. You do not need the depletion model to justify reducing decisions. Every decision you make by hand costs time and attention, and it produces inconsistency, because the same question gets a different answer on Monday than on Friday. A system that answers a question once, in advance, beats one that answers it fifty times. That is true whether or not your willpower is a battery.

Barack Obama gave the practical version to Michael Lewis in a 2012 Vanity Fair profile: "You'll see I wear only gray or blue suits. I'm trying to pare down decisions. I don't want to make decisions about what I'm eating or wearing. Because I have too many other decisions to make."

The suits are a minor example. The principle is serious: spend judgment where judgment is the product.

The Stoic frame: what is actually yours to decide

The Stoics reached the same design nearly two thousand years earlier, from a different direction.

Epictetus opens the Enchiridion with the sorting rule: "There are things which are within our power, and there are things which are beyond our power. Within our power are opinion, aim, desire, aversion" (Enchiridion 1). A large share of the decisions that exhaust operators are attempts to control what sits in the second category: how a client will react, what the market will do, whether a hire will work out.

Marcus Aurelius adds the filter. "On every occasion a man should ask himself, Is this one of the unnecessary things?" He applies it to thoughts as well as acts, "for thus superfluous acts will not follow after" (Meditations 4.24, George Long translation, Project Gutenberg).

Sort first. Cut second. Decide what remains. The same logic applies to a team: see the dichotomy of control applied to team management.

The decision budget protocol

Seven steps. Run the first three this week. The rest compound over a month.

1. Log one normal day of decisions. Write down every choice you make that someone else, a rule or a default could have made. Do not estimate. Count. Then mark the repeat questions: which tasks first, whether to take a call, what to eat, how to reply to a standard request.
2. Sort each decision by reversibility. Jeff Bezos's 2015 shareholder letter calls irreversible decisions "one-way doors" that must be made "methodically, carefully, slowly," and reversible ones "two-way doors" that "can and should be made quickly." By his account, most decisions are two-way doors. Stop giving them one-way-door effort.
3. Convert repeat decisions into defaults. Same breakfast on workdays. Same training slot. A fixed response template for the five questions clients ask most. A standing rule for which meeting requests you accept. A default is a decision made once and executed many times.
4. Write if-then rules for predictable situations. "If a prospect asks for a discount, then I offer the annual plan, not a lower price." "If a task takes under ten minutes and arrives before noon, then I do it the same day." This is pre-commitment, and it has stronger evidence than depletion does. A meta-analysis by Gollwitzer and Sheeran of 94 independent tests found if-then plans, called implementation intentions, had a medium-to-large effect on goal attainment (d = .65).
5. Batch the remaining small decisions. Approvals, purchases, scheduling, inbox triage: one or two fixed blocks per day instead of a stream. The point is not to save willpower. It is to stop interrupting the work that needs your full attention.
6. Protect a window for one-way doors. Put hiring, pricing, partnerships and large spend into a scheduled block when you are rested, by your own record rather than a study's average. Run a Stoic pre-mortem on each before you commit.
7. Review the budget weekly. Which decisions came back to you that should have been a default? Which default broke and needs revising? Add the questions to your Stoic weekly review.

Common mistakes

• Using decision fatigue as an excuse. "I was depleted" is not a post-mortem. Given the replication record, it is often not even the right diagnosis. Look for the missing default or the badly scheduled decision instead.
• Treating the hungry-judge study as settled fact. Cite it, if you must, with the dispute attached.
• Defaulting the wrong things. Defaults belong on reversible, repeated, low-stakes choices. A one-way door deserves deliberation, not a template.
• Building systems you never revise. A default made in January can be wrong by June. Defaults without a review loop become blind spots.
• Trying to decide the uncontrollable. Re-deciding how a client will react is not decision-making. It is rumination. Sort it out with Enchiridion 1, then act on what is yours.

Frequently asked questions

Is decision fatigue real?

The feeling of being drained by many choices is common. Whether it reliably worsens decisions is unsettled: two large preregistered replications found no reliable ego-depletion effect, and a 2025 study of 231,076 nurse triage calls found no evidence of decision fatigue. Reducing unnecessary decisions is still sound practice for other reasons.

What are the signs of decision fatigue?

Commonly listed signs include putting off choices, defaulting to whatever is easiest, irritability, and feeling stuck weighing options. Treat these as signals to examine your workload and schedule, not as a diagnosis. If persistent exhaustion or low mood is affecting your work or life, talk to your doctor.

Did the parole judge study prove decision fatigue?

No. It found favourable rulings fell from about 65% to nearly zero within sessions and recovered after breaks. Critics showed case order was not random, with unrepresented prisoners usually heard last, and a 2016 simulation showed the pattern could arise from scheduling alone. The original authors maintain their effect held.

What is the difference between decision fatigue and ego depletion?

Ego depletion is the broader theory that self-control draws on a limited resource that runs down with use. Decision fatigue applies that idea specifically to making choices. Because decision fatigue was built on the depletion model, the failed depletion replications weaken its theoretical foundation too.

How do you reduce decision fatigue at work?

Count the decisions you make in a normal day, sort them into reversible and irreversible, and turn repeat reversible choices into defaults or if-then rules. Batch small approvals into fixed blocks. Reserve your best-rested window for the few irreversible decisions, and review the system weekly.

Why did Obama wear only gray or blue suits?

He told Michael Lewis in a 2012 Vanity Fair profile that he was "trying to pare down decisions," and did not want to decide what he was eating or wearing because he had too many other decisions to make. It is a clear example of setting defaults for low-stakes choices.

The fastest way to cut decisions is to install a few Stoic defaults and run them daily until they stop costing anything. The free 5-Day Stoic Operator Challenge walks you through one practice per day, built for people running a business.