Marienthal was a small workers’ settlement southeast of Vienna that had grown up around one textile factory. In the last years of the 1920s the factory began to close, and in February 1930 its last workshops shut. The work the settlement had been built for was gone.
In November 1931 a team of young researchers from Vienna arrived to find out what that does to people. Marie Jahoda would write their book. They made a rule that nobody would come only to watch: each had to do something useful. One collected second-hand clothes in Vienna and handed them out, others ran a free class in cutting dress patterns, a gymnastics class for girls, and a doctor’s hour every Saturday. Meanwhile they kept records of everything they could count. They went through the workers’ library’s loan records. They listed what families ate at each meal and had children write essays. They had people write down how they spent each hour of the day. They even timed how fast people walked.
What they found ran against what the socialist parties of the day believed, that the unemployed would be angry and ready to rise up. Marienthal wasn’t angry. It was tired. The researchers called it “the weary community”. People had all the time in the world and did less with it: the library, the clubs and the park were used less than before. Days lost their shape. The team sorted the families into four kinds by how they were holding up: the unbroken, the resigned, the despairing and the apathetic.
Half a century later Jahoda put the lesson into words. A job gives a wage, but it also gives a day its shape, people to be with, a share in something bigger than yourself, a place in society and a reason to get up and do things. Lose the job and you lose all of them, not only the pay.
Ninety years later, from 2020 to 2024, the same town became the site of an experiment. Austria’s employment service offered a job to everyone there who had been out of work for nine months or more. Two economists studied it with a fair test, by drawing lots for who started first, and found what Jahoda would have predicted: people’s incomes rose, and so did their sense of purpose, their contact with others and the shape of their days. Long spells without work all but disappeared from the town.
So work matters. But before anyone can talk about having too little of it, they have to count two things: the people, and the work.
How many, how young
On census day, 25 November 2021, Nepal had 29,164,578 people. Ten years earlier it had 26,494,504. That is a rise of about 0.92% a year, down from 1.35% a year in the ten years before. Families have become smaller, from 4.9 people in a household in 2011 to 4.4, and more than two million Nepalis were away abroad on census day (Chapter 32 follows them).
The shape of the population tells the story better than its size.
For most of Nepal’s history the picture would have been a steep triangle, each younger group much bigger than the one above it, because many babies were born and many people died young. Now the bottom has pulled in. There are fewer children under five than children aged 10 to 14, and the largest groups are teenagers and people in their early twenties. In 2011, 35% of Nepalis were under 15; in 2021, 28%. The share aged 15 to 59 rose from 57% to 62%. Between 20 and 40 the bars on the men’s side are shorter than the women’s: many of those men were working abroad on census day, and the census counts them separately.
Where people live is shifting as well. Just over half, 54%, now live in the Tarai, up from 50% in 2011, and the Hills and Mountains hold a little less of the country than they did. Of the seven provinces Bagmati, with Kathmandu, holds the most people, 21%, and Karnali the fewest, under 6%. Some districts are growing fast and some are emptying: Bhaktapur grew by about 3.4% a year, the fastest in the country, while Ramechhap lost people. About 66% of Nepalis live in places the law calls urban municipalities, नगरपालिका, though that line is drawn by administration, and many of those municipalities still have plenty of fields in them.
A country whose people are mostly of working age has a particular chance. Children and old people need to be fed, schooled and cared for, and most of them earn little or nothing. Workers earn more than they spend, and the difference pays for everyone else. Worked out from the census’s age groups, Nepal had about 61 children and old people in 2021 for every 100 people aged 15 to 59. In 2011 it was about 76. With fewer mouths to feed for each pair of working hands, a family can save, a government can spend more on each child, and the whole country can grow faster. Economists call this the demographic dividend, जनसाङ्ख्यिक लाभांश.
The statistics office has measured how lopsided a life is. Taking the whole country together, people earn more from their work than they consume only between the ages of about 27 and 46. Before that they are children and students; after it, their earnings fall behind what they use. A dividend comes from having many people in those twenty years at once.
But a dividend is a chance, not a gift. It pays only if the people of working age have work, and good work. The planning commission has reckoned that Nepal’s window opened around 1992 and will begin to close around 2047, when the share of people of working age starts to fall as the large young groups in the picture grow older. Whether Nepal uses it depends on the rest of this chapter.
What people bring
A factory owner counts machines, land and money. An economy also has to count what is in people: what they know, what they can do and how healthy they are to do it. Nepal’s school books call this मानव संसाधन, human resources, and economists also speak of human capital, the skills and knowledge that make workers productive. Like a machine, it is built by spending now for a return later: years in school, training, a clinic that keeps a child well.
Nepal has built a great deal of it fast. In 2011, 66% of Nepalis aged five and over could read and write; in 2021, 76%. The gap between men and women is still wide, 84% against 69%. Yet the stock is still thin. Working-age Nepalis who live at home have spent about six years in school, on average.
It shows in who gets work. In the labor force survey of 2017/18, only 32% of working-age people without any secondary schooling had paid work, against 62% of those who had gone to college. School alone doesn’t make jobs, as the next pages show, but it decides who can take them.
Who counts as unemployed
Being unemployed sounds simple: you have no job. The international rule that Nepal’s surveys follow, set out in Chapter 5, is stricter: a person aged 15 or over counts only if they did no work at all for pay or profit in the week before they were asked, not even an hour, could start work now, and were actively looking for it. The unemployment rate is those people as a share of the labor force, श्रमशक्ति, the employed and the unemployed together.
By that rule, the survey of 2017/18 found 908,000 Nepalis unemployed, 11.4% of the labor force, and among people aged 15 to 24, 21.4%. The labor force itself was small: only 38.5% of people aged 15 and over were in it, 54% of men and 26% of women. Those numbers leave out a great deal, and the survey itself shows what.
First, the 663,000 people who had some work but fewer hours than they wanted. They count as employed. Then a much larger group: about 2.5 million who said they wanted work and could start, but weren’t looking. Since they weren’t looking, they weren’t counted as unemployed, or as part of the labor force at all. Add them in, as the international rules now suggest, and 39% of everyone who had work or wanted it was without work or had too little.
And the biggest group of all, about half of everyone aged 15 and over, is outside even that count. That doesn’t mean idle. Most of them were busy growing food for the family, fetching water, cooking and looking after animals and children. Of all Nepalis aged 15 and over, 57% did only that kind of work, for their own family’s use, and only 7% did no work of any kind. Since 2013 the international rules call this work, but not employment, because it is done for the household and not for pay or profit. The rate measures employment, and in Nepal most work isn’t employment.
That is why Nepal’s unemployment numbers disagree with each other. The census of 2021 asked about work in the whole past year, from age 10, and sorted people by how many months of it they had worked; it found about 7% usually unemployed. A household survey of 2022/23 found 12.6%. The Right to Employment Act, 2075, which carries out the right to work in Article 33 of the Constitution, has a definition of its own: a citizen aged 18 to 59 with fewer than a hundred days of work in the year. Each number answers a different question. A fourth Labour Force Survey, the first in nine years, has been in the field since March 2026 and runs until March 2027.
The kinds of unemployment
Even a busy economy always has some people out of work. Someone has left one job and not yet found the next; a graduate is still looking for her first. That is frictional unemployment, and it is a sign of people moving to work that suits them better. Then there are people whose skills nobody needs any more, because a machine or an import has replaced the work they knew. That is structural unemployment, and it lasts until the skills or the place change. When a whole economy slumps, as in Chapter 17, firms stop hiring everywhere at once: that is cyclical unemployment, and it goes away as the economy recovers. The rate an economy settles at when it is neither booming nor slumping is called the natural rate of unemployment. And in a farming country work comes in seasons: plenty at planting and harvest, little in between. That is seasonal unemployment, मौसमी बेरोजगारी.
Nepal has two more kinds that a count of jobseekers hardly sees. Imagine five members of a family working one small farm, where four could harvest exactly as much. Everyone looks busy, and each would say they have work, but the fifth adds nothing to the crop. That is disguised unemployment, अदृश्य बेरोजगारी. The test is simple: if one person left, would less be grown? Where land is small and families large, often it wouldn’t. The other is underemployment, अर्धबेरोजगारी: having work, but too few hours of it, or work far below what a person was trained for.
The survey’s numbers hold a surprise. The unemployment rate was 11% among people who never went to secondary school, 15% among those who had, and 8% among those who had been to college. A likely reason, though the survey doesn’t test it, is that people with secondary schooling look for better jobs than there are and can wait a while for one, while people with little schooling can’t afford a week without some kind of work. School leavers without the next step, waiting for jobs that aren’t there, are what Nepal’s books call educated unemployment, शिक्षित बेरोजगारी.
What makes jobs
Jobs aren’t handed out from a fixed stock. They exist when someone can earn more by hiring a person than the person costs: a shop busy enough to need a helper, a factory with orders, a hotel with guests. So most jobs come from firms that grow, and firms grow when they can sell more, borrow to expand, find the workers they need and count on the rules staying put.
Nepal’s difficulty is that work on farms has been shrinking faster than work anywhere else has been growing. Between 2000 and 2022 the share of working-age Nepalis in farm work fell by about a percentage point a year, about as fast as in similar countries. But jobs in industry and services grew by only about a quarter of a point a year, slower than in most countries the World Bank compares Nepal with, so the share of all working-age people with a job fell. Most new jobs were informal, in small family businesses, with no contract. The World Bank found that developing countries with more trade, firms that can borrow more easily and better schooling have more jobs off the farm, and Nepal’s exports and literacy are in the bottom quarter of those countries. For women the gap is widest: many women work, but their share in jobs off the farm is far below the average of developing countries.
Governments try more direct ways too: building roads that put people to work, training schemes, cheap loans for starting a business, and the hundred days of work, or an allowance, that the Right to Employment Act promises. How much of Nepal’s answer lies in factories, power, farming or visitors is the subject of Part 5.
One way a government changes the price of work directly is with a minimum wage, the lowest pay the law allows. Nepal’s was Rs. 19,550 a month from mid-July 2025 (Chapter 12). Will a higher minimum leave workers better off, or will employers simply hire fewer of them? The plain demand and supply picture of Chapter 7 says a price set above the market’s leaves some sellers unable to sell. Here the sellers are workers.
That was the textbook prediction. Then, on 1 April 1992, New Jersey in the United States raised its minimum wage from $4.25 an hour to $5.05, while neighboring Pennsylvania didn’t. Two economists, David Card and Alan Krueger, saw a natural experiment. They surveyed 410 fast-food restaurants on both sides of the border before and after the rise and compared them. Instead of falling, jobs in New Jersey grew compared with Pennsylvania, by about 13%. Their paper, published in 1994, started an argument that has never quite ended.
David Neumark and William Wascher went back with a different source, the payroll records of 230 of the same kind of restaurants, and found the opposite: jobs in New Jersey fell by about 5% compared with Pennsylvania. Card and Krueger replied with the government’s own employment records, and found employment in New Jersey growing about as fast as in Pennsylvania or a little faster; they put the difference down to a small group of restaurants in the other sample. A large study in 2019 looked at 138 rises in American states’ minimum wages between 1979 and 2016 and found that the number of low-paid jobs stayed about the same for five years afterwards, while pay at the bottom rose. A review in 2022 by Neumark and a colleague went through the published studies and found that most of their main estimates showed some job losses, especially for teenagers, young adults and people with little schooling.
So careful studies of the same kind of question still disagree, and the disagreement is mostly about method: which places make a fair comparison, and which records count the jobs best. What would settle it is more cases where a minimum rises in one place and not in a similar place next door, measured in ways both sides accept. Almost all of this evidence is from the United States, where records count nearly every job. In Nepal, where most work is informal and on farms, a legal minimum reaches far fewer people, and we found no study of what it does to jobs.
Back in Marienthal, nobody needed a study to see what a lost job cost. But the counting matters, because a country that only counts the jobless misses most of the people who want more work than they have. Nepal has many people of working age, more schooling than ever, and a window that won’t stay open forever. The plain lesson is that people are a country’s main resource only when there is work for them, and that a good count of work looks past one number to everyone who wants more of it.
Found a mistake? Please tell us.
