Why Smart Executives Make Bad Calls by 4pm: The Science of Decision Fatigue

The acquisition nearly died at a quarter to five on a Thursday. Not over price, not over strategy — over a warranty clause that both legal teams agreed, weeks later, was standard. The CEO who blew it up had been in back-to-back sessions since seven that morning: a board call, two rounds of management presentations, a media issue at lunch, then four hours of deal terms in a room with no natural light. By late afternoon, a man renowned for surgical judgement was refusing to distinguish between a material risk and a boilerplate irritation. His CFO, who told me the story, said the strangest part was the style of it — no analysis, just an escalating series of flat refusals. “It wasn’t that he decided badly. It’s that he’d stopped deciding at all. He was just saying no to things.”

She was more precise than she knew. Saying no to things — defaulting to the safe, the status quo, the option that requires no further thought — turns out to be exactly what tired minds do. And the story of how we know that, including the part where the science got messy, is worth an executive’s attention. Both halves of it.

The study everyone quotes

Begin with the famous result. In 2011, Shai Danziger, Jonathan Levav, and Liora Avnaim-Pesso published an analysis in PNAS of more than a thousand parole decisions by experienced Israeli judges. The pattern was arresting: at the start of a session, roughly 65 per cent of cases received favourable rulings. As the session wore on, favourable rulings fell — towards near zero — then snapped back to around 65 per cent immediately after the judges’ food breaks, before decaying again. The authors’ interpretation ran everywhere from boardrooms to bestsellers: judgement is a depletable resource, and as it drains, decision-makers slide towards the low-effort default — which, for a parole judge, is “deny”.

The finding slotted neatly into a larger theory. Roy Baumeister’s ego depletion research had argued since the 1990s that self-control and deliberate choice draw on a limited pool — in his studies, people who resisted fresh biscuits gave up faster on puzzles afterwards. Decision fatigue looked like the executive-suite corollary, complete with a mechanism and a snack-based cure.

Then science did what science should. Researchers Keren Weinshall-Margel and John Shapard pointed out a confound in the parole data: case ordering was not random — unrepresented prisoners and weaker cases tended to cluster later in sessions, meaning part of the dramatic curve reflected which cases judges saw when, not just how tired they were. And ego depletion itself hit the replication crisis head-on: a 2016 multi-lab registered replication led by Martin Hagger, spanning more than two thousand participants, found an effect indistinguishable from zero for the standard laboratory paradigm. The tidy story — willpower as a glucose-fuelled battery — did not survive in its strong form.

So should you dismiss the whole idea? No — and the reason matters. Strip away the overclaiming and a robust core remains, resting on things that replicate perfectly well: sustained cognitive load degrades performance; sleep loss and long uninterrupted duty measurably impair complex judgement (which is why aviation regulates crew duty hours as a safety matter); and mentally taxed people demonstrably shift towards defaults, deferrals, and simpler decision strategies. You do not need a contested battery metaphor to accept what the CEO’s Thursday demonstrated: the quality of a difficult judgement is not independent of when, and under what load, it is made. That claim is modest, well-supported, and almost universally ignored in how executives run their days.

The tell: not bad decisions, but avoided ones

Watch for how fatigue actually presents, because it rarely looks like error. It looks like conservatism. The depleted mind does not start choosing recklessly; it starts protecting itself from choosing. Options are deferred. Meetings end with “let’s revisit this next week”. Standard clauses become sticking points because evaluating them properly feels impossible and refusing them feels safe. The parole curve — whatever mix of fatigue and case-ordering produced it — points at the same signature: tired judges drifted towards the decision that closed no doors and required no justification.

In organisations this failure mode is nearly invisible, because it wears the costume of prudence. Nobody gets fired for deferring a decision. But an executive team that does its hardest thinking in the late-afternoon slots of overloaded days is systematically buying the status quo — not because the status quo keeps winning the argument, but because 4pm minds stop holding arguments. The cost never appears on any report. It appears as the slow accumulation of unmade calls.

There is a reason the people with the heaviest decision loads engineer around this. Barack Obama, profiled by Michael Lewis in Vanity Fair in 2012, explained why he wore only grey 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.” He was routinising the trivial to protect capacity for the consequential — decision hygiene, practised at the highest-stakes desk in the world.

The Decision Hygiene Protocol

What follows is the discipline I put in front of leadership teams. None of it requires believing the strong version of any depletion theory. All of it follows from the modest, well-evidenced claim: judgement quality varies with load and timing, so treat your best judgement as a scarce resource and schedule it like one.

Put the big calls where your best hours are. For most people that is the first half of the day, before the accumulated load of meetings and micro-choices; if your rhythm genuinely differs, use yours — the principle is placement, not the clock. What the principle forbids is what almost every organisation does by default: parking the merger, the restructure, or the pricing decision at the end of a full agenda, in the graveyard slot where attention goes to die. Put the hardest item first on the agenda, not last. It is a free upgrade in decision quality.

Automate the trivial ruthlessly. Every recurring small decision — formats, routines, standing approvals, what happens when X occurs — should be converted into a default, a rule, or a delegation. This is the Obama move, scaled to a team: policies for the predictable, so that live judgement is spent only where it earns its keep. A leadership team that debates the same operational trivia weekly is paying for that trivia with its strategy.

Separate deciding from discussing. Long deliberation immediately followed by the decision itself stacks the choice on top of maximum load. Where stakes allow, split them: deliberate today, decide tomorrow morning. The overnight gap costs almost nothing and buys a fresh mind’s second look — which is also, not incidentally, a defence against the room’s late-day drift towards whatever ends the meeting fastest.

Build breaks like an engineer, not a host. Whatever finally explains the parole judges’ post-break snap-back — rest, food, mood, or docket composition — the practical reading is unchanged: long unbroken decision sessions are a known hazard. In marathon negotiations and all-day boards, schedule genuine breaks before the quality visibly drops, and treat a colleague’s sudden rigidity late in the day as a physiological signal, not a bargaining position. Often the single best move in a stuck 5pm negotiation is the least clever one available: adjourn until nine.

Never make the irreversible call exhausted. The capstone rule, and the one the Thursday CEO now enforces personally: any decision that cannot be undone — signing, firing, walking away from the table — gets made before mid-afternoon or it waits for morning. His phrase for it is better than mine, and it has spread through his company: “Nothing signed after four. Tired me is not the one they hired.”

Designing the organisation’s decision diet

The protocol above protects an individual’s judgement. The larger prize is organisational, because companies impose decision load on their leaders as carelessly as individuals accept it — and the design levers sit in plain sight.

Start with the meeting audit no one runs: for each standing meeting, what decisions does it actually exist to make? Most calendars, examined honestly, reveal three pathologies. Meetings that make no decisions at all, only “alignment” — pure load, no output. Meetings that remake decisions already made, because the original wasn’t recorded or the losers are relitigating — which is why a simple decision log, one line per call with owner and date, pays for itself within a quarter. And meetings that stack a dozen consequential calls into a single session, guaranteeing the later items get the parole-board treatment. The fix for the last one is brutal and effective: cap the number of significant decisions per meeting, and let the agenda’s length be determined by the cap rather than the other way round.

Then look at escalation culture. In many organisations, decisions travel upwards not because senior judgement is required but because ambiguity about authority makes escalation the safe move — every unclear mandate converts into load on the floor above. Clarifying who owns which calls, and pushing genuine authority down with the clarity, is decision hygiene applied at scale: it does not just speed the organisation up, it reserves the executive team’s finite best hours for the decisions that genuinely have nowhere else to live.

The compounding effect is easy to underestimate. An executive team that makes fewer, better-placed, better-recorded decisions does not merely decide well on Thursdays. It teaches every layer beneath it what deliberate judgement looks like — and organisations, like judges, perform very differently depending on what their days are designed to protect.

Judgement has a maintenance schedule

The deeper shift here is one of self-image, and senior people resist it precisely because their judgement is what got them the job. We prefer to believe good judgement is a property of the person — stable, portable, always on tap. The honest evidence says it is closer to a property of the person in a state: the same mind, across one working day, can produce both the surgical morning verdict and the 4:45 refusal to think.

That is not a flattering truth, but it is a wonderfully actionable one — because unlike intelligence or experience, state is schedulable. The deal, in the end, survived: the CFO called the pause, everyone slept, and at 8:30 the next morning the CEO read the warranty clause once and approved it without comment. Same clause. Same man.

Different brain. Schedule for the one you want in the room.