In 1955, a British historian and naval administrator named C. Northcote Parkinson published a short satirical essay in The Economist. The essay began with a one-sentence observation that was obviously a joke, and that has become, over the seventy years since, one of the most-quoted propositions in productivity writing.
Work expands so as to fill the time available for its completion.
Parkinson was making a joke about the British civil service. He’d noticed that the number of Royal Navy officers had continued to grow even as the number of ships had declined — the bureaucracy was, in effect, inventing work to justify its own size. The essay extended this observation into a semi-serious principle: give anyone a week to complete a task that could be done in a day, and they’ll take a week. Give them a day, and they’ll find a way to do it in a day.
The observation was offered without evidence. It has since acquired quite a lot of evidence, along with some important qualifications.
The mechanism
The intuition behind Parkinson’s law is familiar to almost anyone who has ever worked to a deadline. When a task has a long horizon, you work on it loosely — starting and stopping, adding complications, doing more research than needed, producing more draft than you’ll use. When the deadline approaches, the work tightens — you make faster decisions, skip unnecessary steps, simplify the approach, and somehow produce a finished version in a fraction of the time the earlier stages took.
The psychological literature has several concepts that clarify what’s happening.
The first comes from the Israeli cognitive scientist Amos Tversky and his collaborator Eldar Shafir, who studied how time pressure affects decision-making. Their counterintuitive finding: for many kinds of decisions, moderate time pressure actually produces better choices than unlimited time. People with unlimited time tend to over-analyse, generate excessive alternatives, and delay committing to any of them. People under moderate pressure focus on the factors that actually matter and commit. The quality of decision often goes up, not down, as available time decreases — up to a point.
The second comes from the Israeli physicist and business theorist Eli Goldratt, whose Theory of Constraints examined how project deadlines actually play out in organisations. Goldratt’s observation: when people estimate how long a task will take, they include generous buffer time as a form of self-protection. They don’t actually need that buffer to complete the task, but they’d prefer not to commit to a tighter deadline they might miss. The buffer then gets consumed by procrastination, scope expansion, and the slower pace that a long deadline permits, rather than by the difficulties the buffer was supposedly protecting against. Work, in effect, fills the buffer — and the task completes right at the deadline regardless of whether that deadline was tight or generous.
A third comes from Edwin Locke and Gary Latham, whose half-century of research on goal-setting has consistently found that specific, moderately challenging deadlines produce better outcomes than vague or generous ones. Their finding, across hundreds of studies in organisational settings, is that people perform better with tight specific deadlines than with loose ones — not because they work harder, but because the constraint forces prioritisation decisions that wouldn’t otherwise be made.
These three traditions converge on the same practical observation. Most tasks have an achievable compressed version and a loose expanded version. Which version you produce depends heavily on the deadline you’re working to, and the compressed version is often as good as the expanded one while taking much less time.
The counter-thread worth hearing
Before accepting Parkinson’s law as an unconditional licence to shorten every deadline, an important qualification from the psychology of stress.
The early twentieth-century psychologists Robert Yerkes and John Dodson published research showing that the relationship between arousal (stress, pressure, time constraint) and performance follows an inverted-U curve. Too little arousal produces sluggish, unfocused work. Moderate arousal produces optimal performance. Too much arousal produces mistakes, tunnel vision, and degraded judgement — the opposite of what pressure was supposed to achieve.
The Yerkes-Dodson law applies directly to Parkinson’s territory. Shortening a deadline modestly often improves the work. Shortening it drastically — especially below the actual complexity of the task — produces worse work rather than faster work. The compressed version that was just as good as the loose version has a floor below which compression starts cutting into quality.
Where that floor sits depends on the task. Simple well-defined tasks have low floors — you can compress them aggressively. Complex, uncertain, creative tasks have higher floors — compress them too far and you get a botched version rather than a tight version. The skill of applying Parkinson’s law well is the judgement about how far compression can go before quality starts to suffer.
A related caveat: the types of deadlines that work well differ from the types that burn people out. Brief sprints with real recovery periods between them seem to produce sustainable high output. Continuous time pressure without recovery produces the opposite — a gradual erosion of quality, sleep, and eventually of the capacity to do the work at all. This matches what the athletic-recovery research (covered elsewhere in this series) has found at the physical level. The brain, like the muscle, responds well to high-intensity work followed by genuine rest. It responds poorly to sustained high-intensity work with no rest.
What this actually means in practice
For an individual trying to get more done in a finite life, the research points toward a few specific moves.
Give tasks tighter deadlines than feel comfortable, and watch what happens. Many tasks will complete at the tighter deadline with similar quality to what the generous deadline would have produced. The experiment is almost always worth running, because the default in most people’s lives is to overestimate how long things need to take.
Be honest about which tasks have real complexity floors. Writing a dissertation can’t be done in a weekend. A programme launch can’t be planned in an afternoon. A major decision can’t be made in ten minutes. For these tasks, tight deadlines produce bad results, not good ones. The skill is recognising which tasks are genuinely in this category, because most people default to treating too many of their tasks this way.
Use sprints with recovery rather than continuous grind. If you’re trying to get serious work done, a focused two-hour sprint with a real break afterward tends to produce more good output than four hours of continuous mid-intensity work. The compression during the sprint works. The recovery makes the next sprint possible.
When you notice yourself working on something and it’s expanding beyond what the situation seems to need, ask: what’s the deadline, and what would I do differently if the deadline were tomorrow? The answer is often revealing. The version you’d produce under tomorrow’s deadline is often the version the task actually needed.
Where Parkinson’s law doesn’t apply
For completeness, some domains where shortening deadlines doesn’t produce the classic Parkinson effect.
Creative insight, in many fields, doesn’t respond well to pure time pressure. The moment of understanding, the surprising connection, the genuinely novel idea — these often require something like idle time, where the mind can wander through possibilities without pressure to commit. Trying to schedule insight has generally poor results.
Learning new material, especially unfamiliar material, has its own pace that compression damages. You can’t compress learning French to a weekend. The material needs time to consolidate, which involves sleep (covered elsewhere in this series) and the spaced practice that lets memory settle.
Relationships and the building of trust can’t be compressed. Trying to rush intimacy, either personal or professional, generally produces the opposite of what was intended. The relational timeline is what it is.
These domains matter because they’re the ones where Parkinson’s law most cleanly fails. Work that’s bounded, well-defined, and primarily about execution tends to respond well to tight deadlines. Work that’s open-ended, exploratory, or relational tends not to.
The question that remains
The deep thing Parkinson’s law, properly understood, teaches is something about how time relates to attention. Time doesn’t do the work. Attention does the work. A block of time without focused attention on it produces almost nothing of value, however long. A smaller block with full attention on it often produces everything you needed.
This should change how you think about your own productivity. The question isn’t usually how much time do I have? It’s what deadline would force me to actually apply attention? Deadlines, in this framing, aren’t constraints that interfere with good work. They’re the mechanism that converts ambient time into focused work, which is the only kind of time that actually produces anything.
The question worth carrying, next time you find yourself with what feels like plenty of time for a task:
If this were due tomorrow, what would you do — and is the version you’d produce meaningfully worse than the version you’ll produce with the longer deadline you actually have?
Key research referenced: C. Northcote Parkinson’s 1955 essay in The Economist, later expanded in Parkinson’s Law (1957); Amos Tversky and Eldar Shafir on decision-making under time pressure; Eli Goldratt’s Theory of Constraints (particularly Critical Chain, 1997); Edwin Locke and Gary Latham’s goal-setting research; the Yerkes-Dodson law from 1908.