Parkinson's Law: Why Work Expands to Fill the Time Available

What is Parkinson’s law?

Parkinson’s law says that work expands to fill the time available for its completion. Give yourself a week for a two-hour task and it will often take a week, not because the work grew harder but because it grew larger, slower and more polished than it needed to be.

As a mental model, it is a warning about default time limits. Deadlines that nobody chose on purpose (end of week, end of quarter, a one-hour meeting slot) quietly become the size of the work.

Three things usually fill the extra time:

• Scope creep: the task grows to include extras nobody asked for.
• Polish: a third and fourth pass that adds little.
• Delay: starting late because the deadline feels far away, then rushing anyway.

The model is not “work faster”. It is “decide how much time the work deserves, then hold to it”.

Where it comes from

C. Northcote Parkinson, a British naval historian, published an essay titled “Parkinson’s Law” in The Economist in November 1955. It opens with the famous sentence and the example of an elderly lady who spends an entire day sending one postcard to her niece: finding the card, the glasses, the address, writing, deciding whether to take an umbrella to the mailbox. A busy person would do it in minutes.

The real target of the essay was bureaucracy. Parkinson argued, with satirical numbers drawn from British administration, that the size of an office grows regardless of its workload, because officials want subordinates, not rivals, and officials make work for each other.

He expanded the essay into a 1957 book, which also introduced the “law of triviality”: a committee will approve a reactor in minutes and argue for an hour about a bicycle shed. The everyday use of Parkinson’s law, about personal deadlines, came later and stuck.

How to use Parkinson’s law

1. Estimate the honest effort first. Before setting a deadline, ask how long the task would take with full focus. Write that number down, then add a reasonable buffer, not an arbitrary one.
2. Define done before you start. Write one or two sentences describing the finished output. Scope written in advance resists creep.
3. Time box, then stop. Set a block (45 minutes, two days) and treat the end as real. Ship the version you have, then decide whether more time is worth it.
4. Shorten default containers. Book 25-minute meetings instead of an hour, two-week projects instead of a quarter. Watch whether quality actually drops.
5. Cut by importance, not by order. When time runs short, use the Pareto principle and via negativa: drop the parts that add least, not whatever happens to come last.
6. Make the deadline external. Tell someone, schedule the review, publish the date. Self-imposed deadlines are easy to move.

One habit makes this easier: track how long things actually take. For two weeks, write the start and end time of each task next to your original estimate. You will likely see both laws at work, some tasks expanding to their slot and others overrunning badly. That record is your own base rate, and it beats any rule of thumb when you set the next deadline.

For teams, publish the time box with the task: “two days, then we review what we have”. A shared limit is harder to stretch than a private one.

3 real-world examples

Business

Say a team is given a full quarter to write a short strategy memo. Drafting starts in week ten, and meetings in the first nine weeks keep widening the scope. A sharper approach: one week to draft a two-page version, one week for comments, done. If the memo truly needs more, the first draft will show where. Most of the original quarter was waiting and expanding, not thinking.

Personal life

Imagine you set aside all of Saturday to clean the apartment. It takes all of Saturday. The next week you give yourself 90 minutes before friends arrive, and the place is clean in 90 minutes, minus the drawer reorganizing you didn’t need. The lesson is not to rush every chore; it is that the time you allow becomes the standard you work to.

Tech

A developer estimates a feature at three days, gets two weeks on the plan, and spends the extra time refactoring nearby code and trying a new library. The feature ships on day ten. Breaking the work into small, separately shippable pieces, each with a short window, keeps the extra time from spreading. Goldratt’s theory of constraints describes the same pattern in projects: padding in each step tends to be used up, so the savings rarely pass downstream.

Common mistakes

• Treating it as permission to set impossible deadlines. Squeezing real work produces errors and burnout. The target is padding, not substance.
• Ignoring the planning fallacy. People already underestimate how long things take. Cut the slack, not the honest estimate.
• Shortening time without reducing scope. If the deadline shrinks, the definition of done must shrink too.
• Applying it to creative or exploratory work too early. Some tasks need open time to find the problem. Time box the exploration phase instead of the whole project.
• Forgetting the original point. Parkinson was describing how organizations grow. Check whether your team adds process faster than it adds output.
• Filling freed time with more of the same. If a shorter meeting simply leads to another meeting, nothing was saved. Decide in advance what the reclaimed time is for.

Parkinson’s law vs the planning fallacy

The Mentarium take

Parkinson’s law and the planning fallacy look like opposites, and that is exactly why you need both. One says you give work too much time; the other says you give it too little. The way out is to separate the two acts: estimate honestly from past data, then commit to a firm window with a written definition of done.

In The Mentarium Decision Loop (Frame → Map → Stress-test → Decide → Review), Parkinson’s law belongs in Frame. “By when” is part of what is being decided. Set the deadline on purpose instead of inheriting it, and the decision process itself stops expanding. The Decision Weight Score (DWS) helps here: a DWS of 1 to 3 gets under 10 minutes, a 4 to 6 gets about an hour. Without a limit, a small choice will happily eat an afternoon.

Our blunt advice: halve the time you would normally give a routine task once, as an experiment, and note what you cut. If nothing important was lost, the old time was Parkinson’s. Use the Eisenhower matrix to choose which tasks get that experiment first.