The Grey T-Shirt Won’t Save You: The Shaky Science of Decision Fatigue

Decision fatigue — the idea that making choices depletes a finite mental resource, degrading every subsequent decision — is one of the most cited concepts in founder productivity culture. The grey T-shirt, the pre-set morning routine, the ruthless calendar block: all of it rests on a theory that, under rigorous scientific examination, has largely failed to hold. Founders deserve to know this, because the strategies that follow from a wrong model are, at best, harmless theatre and, at worst, a distraction from what actually governs decision quality under pressure.

Why the grey T-shirt became a management philosophy

The story is familiar. ¹ Mark Zuckerberg told a public Q&A audience in 2014: “I really want to clear my life to make it so that I have to make as few decisions as possible about anything except how to best serve the community.” Steve Jobs, Barack Obama, and a cohort of high-profile operators followed similar logic: eliminate trivial choices to preserve cognitive bandwidth for consequential ones. The wardrobe uniform became a symbol of executive discipline, and the underlying theory — that decisions draw from a shared, exhaustible pool — became received wisdom in every leadership podcast and productivity book published in the decade that followed.

The theory has a name: ego depletion. ² In 1998, social psychologist Roy Baumeister and colleagues published a landmark paper introducing a powerful new paradigm: willpower operates like a muscle. Use it on one task — resisting cookies — and you have less of it for the next. ³ The radish group, who had been told to resist freshly baked cookies and eat radishes instead, gave up on a subsequent unsolvable puzzle 60% faster than the cookie group. Baumeister called the phenomenon ego depletion, and it launched a research empire.

The problem is not that the insight was wrong in spirit. The problem is that the specific model — willpower as a depletable biological fuel — did not survive the rigours of scientific self-correction.

What happened when the science was put to a proper test?

⁴ A 2010 meta-analysis by Martin Hagger and colleagues surveyed 198 tests of the ego depletion effect and found a medium effect size of d = 0.62 — a robust, meaningful number that gave the theory enormous credibility. Then researchers began to look more carefully. ⁵ Carter and McCullough (2014) argued that the evidence was strongly influenced by small-study effects and publication bias toward significant results, suggesting the true effect size could in fact be zero.

The decisive test came in 2016. ⁶ A pre-registered Registered Replication Report involving 23 laboratories across multiple continents and 2,141 participants attempted to replicate a standard ego depletion protocol. ⁷ The result: the ego depletion effect was statistically indistinguishable from zero (d = 0.04, 95% CI [−0.07, 0.15]). ⁸ Only two of the 24 research groups found a significant positive effect — a rate consistent with what would be expected by random chance alone.

Baumeister disputed the methodology, arguing that the replication tasks lacked the psychological nuance of the originals. ⁹ Michael Inzlicht, a prominent ego depletion researcher who had won a theoretical innovation prize for his work in the field, was more direct: he attempted to replicate ego depletion in his own lab, over and over, and failed. “Ego depletion,” he concluded, “at least as typically studied in the lab, was a mirage.”

¹⁰ The replication crisis forced the field to confront methodological weaknesses and publication bias, paving the way for more nuanced theories that incorporate motivation, attention, and resource allocation — rather than a simple fuel-tank metaphor. ¹¹ Researchers now note that there is no single, universally agreed-upon operational definition of “self-control,” and the tasks used to manipulate and measure it are extraordinarily diverse, often lacking independent validation. Elegance is not evidence, and quantity is not quality.

What about the hungry judges?

No piece of evidence has done more to embed decision fatigue in business culture than the Israeli parole board study. ¹² Danziger, Levav, and Avnaim-Pesso (2011), published in the Proceedings of the National Academy of Sciences, analysed 1,112 parole hearings by eight judges over ten months. ¹³ Within each session, a prisoner’s probability of receiving a favourable ruling began at roughly 65% and dropped to near zero by the end of the session, before resetting after a food break. The interpretation: mental fatigue was causing judges to default to the safe, status-quo choice of denial.

The study has been cited close to 2,500 times and was featured prominently in Daniel Kahneman’s Thinking, Fast and Slow. It is also contested. ¹⁴ A subsequent analysis published in the same journal found that case ordering is not random: the board tries to complete all cases from one prison before a break and start with another prison after it. Within each session, unrepresented prisoners — who are less likely to be granted parole regardless — tend to go last. ¹⁵ Psychologist Daniël Lakens has argued that the size of the effect in the original study is implausibly large. More recent studies show that certain legal decisions can become more lenient with increasing case order, which may reflect a direction-of-comparison mechanism rather than decision-maker fatigue.

This does not mean judicial decision-making is unaffected by extraneous factors — it clearly is. It means the specific causal story (depleted willpower → default to denial) is far less certain than a decade of management writing has implied. A study that has been cited 2,500 times is not automatically a study that has been replicated 2,500 times.

The better model: motivation and attention, not fuel

If the resource-depletion model is wrong, what replaces it? The most credible alternative comes from Inzlicht and Schmeichel’s 2012 process model, published in Perspectives on Psychological Science. ¹⁶ Rather than framing depletion as a failure of capacity, the process model proposes that exerting self-control at Time 1 causes temporary shifts in both motivation and attention that undermine self-control at Time 2. The brain is not running out of fuel; it is recalibrating its priorities. ¹⁷ Current concepts of ego depletion therefore emphasise a process of motivational and attentional reallocation rather than the depletion of resources.

This reframing has a critical implication: stakes and motivation can override apparent depletion entirely. ¹⁸ Research on intrinsic motivation and mental fatigue shows that with high intrinsic motivation, people can maintain performance over time as they appear willing to invest more effort as time progresses — even when subjective fatigue is reported. ¹⁹ The process model accounts for the finding that high motivation or strong incentives can completely erase the ego depletion effect: if a task is sufficiently engaging or rewarding, the motivational “want-to” aligns with the “have-to,” and apparent fatigue does not arise.

For founders, this is not a trivial distinction. A founder deciding whether to terminate a key hire at 6 p.m. is not operating in the same motivational context as a laboratory participant crossing letters off a page. The decision is high-stakes, identity-relevant, and intrinsically charged. The conditions under which the ego depletion effect — even if real — would apply to that decision are far narrower than the popular narrative suggests.

The belief effect: your theory of willpower shapes your willpower

Perhaps the most operationally significant finding in this literature comes from Carol Dweck, Veronika Job, and Gregory Walton at Stanford. ²⁰ Their research proposes that whether depletion takes place or not depends substantially on a person’s belief about whether willpower is a limited resource. ²¹ People who viewed the capacity for self-control as not limited did not show diminished self-control after a depleting experience. A longitudinal field study found that a non-limited theory of willpower predicted better self-regulation — better time management, less procrastination, less impulsive spending — for students facing high self-regulatory demands.

²² Stanford Professor Carol Dweck’s work found that people get fatigued or depleted after a taxing task primarily when they believe that willpower is a limited resource, but not when they believe it is not so limited. ²³ A separate set of experiments showed that even the performance-enhancing effect of glucose on self-control — long cited as biological proof of the fuel model — only appeared in people who already believed willpower was limited and easily depleted.

The productivity industry may have inadvertently taught founders to believe their willpower is limited, thereby creating the very effect it sought to describe. If the grey T-shirt ritual encodes a belief that every small choice costs something irreplaceable, the ritual may be generating the fatigue it claims to prevent.

What founders should actually do

None of this means that decision architecture is useless. Reducing low-value decisions is still sensible — not because it conserves a biological fuel, but because it frees attentional bandwidth and reduces the cognitive load of context-switching. The rationale matters, because the rationale shapes the belief, and the belief shapes the outcome.

Three principles follow from the evidence:

1. Prioritise by stakes and novelty, not by sequence. The process model suggests that motivation and perceived importance are the primary governors of decision quality. Structure your day so that genuinely novel, high-stakes decisions receive your most engaged attention — not simply your earliest hours.
2. Audit your implicit theory of willpower. If you have internalised the belief that every decision costs you, you are likely experiencing a self-fulfilling prophecy. The evidence from Job, Walton, and Dweck suggests that adopting a non-limited theory of willpower — treating engagement as generative rather than depleting — produces measurably better self-regulation outcomes.
3. Design for motivation, not just for minimisation. The most robust finding across this literature is that high intrinsic motivation modulates fatigue effects. Founders who are genuinely energised by the decisions in front of them are operating in a different cognitive regime than the laboratory participants on whom the decision fatigue narrative was built.

What this means

The grey T-shirt is not the problem. The problem is building a decision architecture on a theory that the science has largely abandoned — and then mistaking the ritual for the result. Founder agency is not about conserving a finite resource. It is about directing attention, sustaining motivation, and making the decisions that matter with the full weight of your engagement behind them. The science, imperfect as it is, points in that direction. The wardrobe is optional.

At Business Growth Accelerator (a FounderWise brand), the frameworks we use with founders are built on the best available evidence — which means updating them when the evidence changes. If your operating system is still running on the 1998 cookie study, it may be time for an upgrade.