Four economists went looking for what meetings actually cost, which nobody had measured properly before. They surveyed more than nine thousand workers and linked every answer to Norwegian administrative records holding those same people's earnings, occupations and employers.1 The paper opens with a line from John Galbraith, quoted approvingly: meetings are indispensable when you don't want to do anything.1

Then they put a price on it. The average Norwegian worker spends 4.7 hours a week in meetings, which is 12 percent of a 37.5-hour week. Around 60 percent of workers attend meetings regularly, and those people average about 80 minutes of meeting per working day.1 Converted into money at observed wages, firms spend roughly 14 percent of their wage bill on meetings. That comes to about nine thousand US dollars per employee per year, and around 24 billion US dollars a year across the Norwegian economy.1

The authors are careful to say that figure is a floor. It counts time in the room and nothing else, so preparation, travel between locations, the follow-up work and the cost of switching attention are all excluded.1

The finding that ruins the obvious conclusion

Everything above sets up a familiar executive move. Find the cost, cut the cost, book the saving.

What stops that move is where the meetings are. Meeting intensity is higher at higher-paying firms and at higher-revenue firms, and the relationship gets steeper toward the top of both distributions.1 These are the organizations where an hour of employee time costs the most, so if meetings were mostly waste they are the ones with the strongest reason to cut. They buy more instead. The authors put it plainly, saying the pattern is hard to reconcile with a view that meetings primarily reflect organizational inefficiency.1

The firms paying the most for an hour of attention spend the most of it in rooms together. Whatever is happening in there, they are not doing it by accident.

What is actually being bought

The paper's answer is that a meeting does more work than coordination. Workers in meeting-heavy environments report learning more on the job, and the association holds with controls for occupation, sector and the specific firm, across about six thousand observations.1 Contact with more senior colleagues in particular tracks with stronger wage growth.1 The mechanism they land on is knowledge transmission: meetings are one of the places where what a senior person knows gets into a junior person's head.

Read the purpose data with that in mind and it stops looking like theatre. The most frequently reported reasons are planning and strategy, collaboration and problem solving, project updates, and information sharing. Administration, networking and informal chat come well behind, and the pattern barely moves across occupations, industries and working arrangements.1

What that means for the cull

I have watched enough calendar purges to know the shape of the announcement. A number gets presented, usually a large annual cost or a headcount-hours figure. Recurring meetings get cancelled in a block. Everyone is told they have been given their time back, and for a few weeks the individual work does move faster, which makes the decision look correct.

The part that does not show up for a year or two is the apprenticeship. If the room was where a junior analyst heard a senior one reason through a problem, and the room is gone, the analyst still has a job and still hits their targets. They just develop more slowly, and nothing in the reporting will attribute that to a decision made in a cost review.

I wrote on Monday about how things survive inside organizations because continuing requires no decision while stopping requires an owner, and meetings are the example everyone reaches for. This paper is the other half of that thought. Making the stop decision is still right. Assuming the answer is stop is not, and the cost of the room turns out to be the easiest thing about it to measure.

Three honest limits

This is a working paper and has not been peer reviewed. It covers one country with unusually good administrative data and a 37.5-hour standard week, so the levels will not transfer cleanly to a US firm. Most importantly, the learning and wage-growth results are associations rather than proven causes, and the authors present them that way. It remains possible that better-run firms both meet more and develop people more, with the meetings along for the ride.

None of that rescues the cull, though. The case for cutting meetings has always rested on the claim that the time is wasted, and the evidence for that claim was that meetings are annoying. There is now a large dataset pointing the other way, and the people with the most expensive hours in the economy are voting with their calendars.

So the question worth putting to your own organization is narrower than how many meetings you have. It is which of your recurring meetings is the only place a less experienced person hears an experienced one think out loud. If you cannot name it, that is worth knowing before the next cost review, not after.