Parkinson’s law: why work expands to fill the time you give it
A deadline decides how big a job is allowed to become.
An elderly woman with nothing else to do sets out to send a postcard to her niece. She spends an hour finding the postcard and another hour hunting for her spectacles. An hour and a quarter goes on writing a few lines, and twenty minutes on deciding whether to take an umbrella to the postbox.
By evening she is exhausted. A busy person would have done the whole thing in three minutes.
This little scene opened an essay in The Economist on 19 November 1955. Its author, the British naval historian C. Northcote Parkinson, called the idea Parkinson’s law. His first sentence has been quoted ever since: “Work expands so as to fill the time available for its completion.”
Parkinson’s law is the observation that a task tends to take up all the time allowed for it, whether or not the task needs that much. Parkinson stated it in 1955 as a joke about government offices, which he said kept hiring staff regardless of how much work there was to do.
Where did Parkinson’s law come from
Parkinson was then teaching history at the University of Malaya in Singapore. His essay appeared, unsigned in The Economist’s usual style, soon after a Royal Commission had reported on Britain’s civil service.
The postcard was only the warm-up. The real target was bureaucracy. If work can stretch to fill any amount of time, Parkinson argued, then there need be no link between the work a department does and the number of people it employs.
His evidence came from the Royal Navy. Between 1914 and 1928, the navy shrank by about a third in men and two-thirds in ships. Over the same years, the number of Admiralty officials in London grew from 2,000 to 3,569, a rise of 78%. He added that the Colonial Office kept growing as Britain’s empire shrank.
Parkinson offered two explanations, both stated with a straight face. First, an official wants to multiply subordinates, not rivals. An overworked manager does not ask for a colleague of equal rank, who might compete for the next promotion. He asks for two juniors. Second, officials make work for each other. Parkinson walked his readers through seven officials passing one file between them, drafting and redrafting, until the senior man finally sent out the same reply he would have written alone.
He even supplied a mock formula, and concluded that staff would grow by between 5.17% and 6.56% a year “irrespective of any variation in the amount of work (if any) to be done”. The essay grew into a bestselling book, Parkinson’s Law, in 1957.
The popular version has drifted from the original. Today, Parkinson’s law is mostly a tip about personal productivity: give yourself less time and you will finish sooner. Parkinson was writing about organisations. The postcard line was his opening joke. The law he named was about how offices grow.
Parkinson’s law examples in experiments and at work
Psychologists took the joke seriously enough to test it. In 1966, Elliot Aronson and E. Gerard found that people who had been given too much time for one task went on to work more slowly on the next one, even when its deadline was gone. The researchers called it the excess-time effect.
In 1967, Judith Bryan and Edwin Locke ran two experiments with simple addition problems. One group was given twice as much time as the task needed; the other was given just enough. For the first three rounds, the group with spare time took longer to finish. The reason was the goals people set. With more time, they aimed lower and worked at an easier pace, even when they said they were trying to work quickly.
A 1999 study by Laura Brannon, Paul Hershberger and Timothy Brock produced a sharper version. Students expected to judge four sets of photographs. Some were told, before starting the third set, that the fourth had been cancelled. Those students slowed down on the third set and stretched it into the freed time. The effect held in three further runs of the experiment.
You meet the same pattern at work. A meeting booked for an hour lasts an hour, even when the decision is made in twenty minutes. A report due on Friday gets finished on Thursday night, whether it was given a week or a month.
Parkinson’s bureaucratic version is harder to measure, but the incentive he described is easy to recognise. Wherever a manager’s standing rises with the size of their team, managers have a reason to ask for more people, whether or not the work has grown.
Why does work expand? Bryan and Locke’s answer is the most useful. Time limits shape the goals we set. A loose deadline quietly lowers the target, and the work rises or sinks to meet it. There is also a simpler reason. Most tasks can always be improved a little more, so a task with spare time attached tends to absorb it.
Is Parkinson’s law true
Partly, and less neatly than the slogan suggests.
Parkinson’s essay was satire. His formula was a parody of social science, and his figures were chosen to make a point. He did raise the obvious objection himself: the navy of the 1920s was adopting aircraft and submarines, which might need more administrators. His answer was that dockyard officials grew by only 40% while Whitehall’s grew by 78%. Two numbers make a good joke. They do not make a proof.
The experimental support is genuine but limited. In Bryan and Locke’s study, the effect faded in the later rounds. In 1973, C. Orpen and E. Riese reported in the Journal of Social Psychology that they had failed to replicate the excess-time effect. Brannon and her colleagues noted in 1999 that the law had survived for decades without a clean demonstration at the level of the individual.
The experiments also used simple tasks, such as adding numbers and judging photographs. They say little about complex or creative work, where cutting time can cut quality. The useful finding is narrower: people tend to pace themselves to the time they are given, so the time you give a task should match the task.
The fair reading: Parkinson’s law is a real but modest tendency, not a law of nature. It shows up most when a task has no clear finish line and nobody checks how long it should take.
How to beat Parkinson’s law
Estimate the task, not the calendar. Before you set a deadline, ask how long the work would take if you started now and worked on nothing else. Set the deadline from that, with a sensible margin.
Break long deadlines into short ones. In 2002, Dan Ariely and Klaus Wertenbroch found that people given evenly spaced deadlines for three proofreading tasks performed better than those given a single deadline at the end. Interim checkpoints stop a long deadline from filling with slack.
Define what finished means. Write down what a complete version contains before you begin. A task with a clear finish line has less room to expand.
Book meetings for the time the agenda needs. A 25-minute slot sends a different signal from an hour, and the agenda has to fit it.
For teams, ask what the work requires before you ask for headcount. If a new hire mainly adds reviewers to the same file, Parkinson’s seven officials are back.
Time is not just the space work happens in. It is one of the instructions you give.
Daily Reflection
Pick one task on today’s list. Decide how long it should really take, set a timer for that, and stop when it rings.
