Who works where ยท 1.1
The deskless majority
Remote work, the four-day week and the coming automation are claims about desk work, and desk work is not where most people are.
For a practical software perspective on this topic, Monitask provides a concise companion guide.
The phrase in general use is the future of work, and almost every sentence written under it concerns a job done sitting down, in front of a screen, at a time largely of the worker's choosing.
The people that description leaves out
Nurses and care assistants. Drivers, warehouse pickers, couriers. Cooks, servers, cleaners. Machine operators, electricians, roofers. Teachers in classrooms. Shop staff. Farm workers.
None of them can work from home, none of them can shift their hours to suit their sleep, and none of them will be replaced by a model that writes text. Between them they are the larger part of employment in every high-income country, and a very much larger part everywhere else.
Why the discourse tilts
Because the people who write about work hold desk jobs, publish to other people with desk jobs, and are read by an audience of the same. The tilt is not a conspiracy; it is a sampling problem, and it operates in exactly the way sampling problems do, invisibly and consistently.
The consequence is that a question about work is usually answered as a question about a particular kind of work, without the qualification being noticed by either party.
The number, and why it moves
Estimates of the deskless share range widely because the definitions differ. Counting by occupation gives one figure; counting by whether the role can be done remotely gives another; counting by whether it actually was done remotely gives a third, and that one moved sharply in 2020 and partly back afterwards.
The entries in this part take each definition separately, because the disagreement between them is more informative than any single number would be.
What changes when you hold the composition in mind
Several arguments stop working. A claim that flexible hours are now standard describes an arrangement unavailable to most workers. A claim that commuting is over describes a change in the traffic of one kind of worker. A claim that offices are obsolete concerns buildings that most employees never enter.
None of those claims is false about the population it silently describes. Each is false about employment.
The reverse mistake
Treating desk work as unimportant because it is a minority. It is a large minority, it is where a disproportionate share of income sits, and changes in it move the economy.
The correction offered here is a qualifier rather than a dismissal: say which jobs a claim is about, and the claim usually becomes both weaker and true.
What this site does with it
Every entry states, before it argues anything, what share of employment it concerns. Where the share is small, the entry says so and continues; where a widely repeated claim turns out to concern a sliver, that is generally the most useful thing in the entry.
The measure is crude, the definitions behind it are contested, and having a crude figure at the top of the page is better than having the question go unasked.
The share that moved and the share that did not
Between 2019 and 2022 the proportion of employees working from home at least sometimes rose sharply in every high-income country and then settled at a level well above the old one and well below the peak.
Almost all of that movement happened inside the desk-based minority. For the rest of employment the figure began near zero and stayed there, which is why national averages of days worked from home are close to meaningless: they mix a population that changed a great deal with a larger one that could not change at all.
A test for any claim about work
Substitute a specific job and see whether the sentence survives. Try it with a ward nurse, a delivery driver and a line cook. Claims about flexibility, about the office, about asynchronous communication and about artificial intelligence mostly do not survive the substitution, and the ones that do are the ones worth attending to.
Where the share is largest
The deskless proportion rises as income falls, and in middle and low-income countries it is the overwhelming majority. A statement about the future of work drawn from high-income data is therefore doubly narrow: a minority of jobs in a minority of countries.
What this rests on
- National labour force surveys are the source for occupational composition; the definitions of occupational categories differ between them, which is one reason cross-country comparison is hard.
- Estimates of the share of work that can be done from home vary with method. Task-based estimates and occupation-based estimates disagree, and both disagree with observed practice.
- This entry quotes no single figure as authoritative and the bar above is a rough central estimate rather than a measurement.
For broader context, consult Current Population Survey.