Parkinson's Law and why every meeting fills its slot.

If Parkinson wrote his 1955 essay today, he would have started with the calendar.

C. Northcote Parkinson was not thinking about meetings. His November 1955 piece in The Economistwas about British civil servants and the observation that their workload kept expanding regardless of whether there was more work to do. The law he proposed, in the essay’s opening line, is one of those rare observations that has never been improved upon:

Book a one-hour meeting. It will take an hour. Book the same meeting for thirty minutes and it will take thirty minutes. There are exceptions — genuinely complex topics that need the whole hour — but the default is that the meeting fills its slot. Nobody wants to look like they’re rushing. The agenda was written to fit the time that was booked. And there is no structural cost, inside the meeting, to spending the remaining ten minutes on tangent.

Aronson and Gerard confirmed the effect empirically for individual task time in their 1966 paper in the Journal of Experimental Social Psychology. Steven Rogelberg at UNC Charlotte — whose The Surprising Science of Meetings(Oxford, 2019) is the best modern empirical synthesis on the topic — treats the shorten-the-slot advice as practitioner wisdom that is consistent with Parkinson rather than as a lab-established fact about meeting duration specifically. I haven’t seen a peer-reviewed study measuring whether a 30-minute meeting ends earlier than a 60-minute meeting. That’s a real gap.

But the advice is still right, and you do not need a paper to know it’s right. You have been to both meetings.

This is the cheapest, most widely-available productivity intervention on the internet, and almost nobody does it. Not because the mechanic is hard — Google Calendar has a “speedy meetings” setting that flips it by default — but because calendar software was built around round numbers, and working with round numbers is a social signal about being serious.

The five-minute gap between a 25-minute meeting and the next one is where people walk, breathe, drink water, write a two-line summary of what just happened, and prepare for what’s next. None of this is available in a calendar booked in 30-minute slabs stacked end to end. Rogelberg’s research on meeting fatigue is the closest empirical anchor on why this matters, though the causal link between slot length and recovery is still under-studied.

The Parkinson adjustment has a companion observation: Kahneman and Tversky’s planning fallacy(1979) tells us that people systematically under-estimate how long things will take. Meetings are one of the few contexts where people systematically over-estimate — they book a bigger slot than they need, because the cost of running long is social and the cost of ending early is cognitive (people feel owed the time).

The combined corrective is almost comical in how simple it is: take every recurring meeting on your calendar, subtract five or ten minutes from each, and see which ones complain. The ones that complain were probably using the time for tangent. The ones that don’t complain were already running short. You just made the calendar honest.

Parkinson wasn’t thinking about meetings. He didn’t need to be. The point was always about defaults, and your calendar is the most default-driven surface in your working life. The defaults are worth editing.