Economy & Society · March 2026

The Degree and the Dead End: Why India’s Most Educated Generation Is Also Its Most Unemployed

India has just produced the largest generation of educated young people in its history. Nearly 40 percent of them cannot find a job. A landmark new report from Azim Premji University maps the widening chasm between a classroom revolution and an economy that was never rebuilt to receive it.

Rajan is twenty-four years old and holds a Bachelor of Arts degree from a government college in Gaya, Bihar. He graduated two years ago. Since then, he has appeared for the Bihar Public Service Commission examination twice, the Railway Recruitment Board test once, and a state bank clerical exam. He is currently studying for the fourth attempt at the BPSC.

Rajan is not a failure story. He is, by any reasonable measure, a success of India’s education system — the first in his family to attend college, the first to hold a degree.

A new report from the Centre for Sustainable Employment at Azim Premji University — the State of Working India 2026, titled The Youth in the Labour Market: Pathways from Learning to Earning — has spent the better part of a year mapping what happens between the moment a young Indian graduates and the moment, if it comes, when they find stable work. What it found is both a story of genuine achievement and a crisis hiding inside it, and the two are so tightly wound together that separating them is itself an act of policy.

I. The Numbers That Should Be Good News

Start with the real achievements. India’s tertiary enrolment ratio — the share of young people attending university or college — stands at 28 percent, broadly comparable to countries at similar income levels. Over the past four decades, the number of colleges has increased from 29 per lakh young people in 2010 to 45 per lakh by 2021. Women, in particular, have made extraordinary gains: female educational attainment among the 15-to-29 cohort has risen faster than male attainment, and young women are increasingly finding employment in sectors — IT, automobile manufacturing, textiles, business services — that were largely closed to them a generation ago.

The caste-based occupational lock that once made certain jobs hereditary — and certain people invisible to other sectors — has weakened. In 1983, young Scheduled Caste and Scheduled Tribe workers made up 40 percent of their communities employed in the leather and footwear industry. By 2023, that share had fallen to 24 percent, as more SC and ST youth entered manufacturing, telecommunications, and paper industries. Young people are moving out of agriculture and into the formal economy faster than any previous generation.

These are not small things. They represent decades of investment in schools, in reservations, in public college expansion, in shifting social norms around who belongs in a classroom. Azim Premji University president Indu Prasad said, on releasing the report, that more young people today are educated, informed, and ambitious than ever before — and that these are real achievements worth acknowledging.

The problem is what happens next.

II. The Degree and the Dead End

Between 2004 and 2023, India added roughly 5 million graduates to its labour force every single year. In those same years, the economy created approximately 2.8 million graduate-level jobs annually. The gap — 2.2 million disappointed graduates per year, year after year — is not a blip. It is a structural feature of the economy. And its consequences are now visible in the unemployment statistics.

Read those numbers carefully. Two in five young Indians between 15 and 25 who are actively looking for work cannot find it. Among those aged 20 to 29 who are unemployed, two-thirds now hold a graduate degree — a share that has risen sharply from 46 percent in 2017 to 67 percent in 2023. Unemployment in India has not retreated from education. It has followed education upward. The more the country has expanded its colleges, the more it has produced graduates who cannot find work commensurate with what those colleges promised.

Of the young male graduates who do find some form of employment within a year, only 7 percent secure a permanent salaried job. The rest land in casual work, self-employment, or the vast, informal, unprotected grey zone of the Indian labour market — gig deliveries, construction labour, unregistered teaching — where the degree they spent years and money earning provides no premium at all.

III. The Geography of a Skewed System

India’s education crisis is not evenly distributed. Its best institutions — IITs, IIMs, central universities, premier medical colleges — are clustered predominantly in southern India and in a handful of large northern cities. The rest of the country makes do with what remains: general degree colleges in small towns, running three-subject BA programmes with lecture halls built for forty students and classes attended by one hundred, staffed by faculty hired on ad-hoc contracts who are themselves waiting for a permanent government appointment.

The student-to-teacher ratio at private colleges averages 28 students per teacher; at government colleges, it is 47. The All India Council for Technical Education recommends a maximum of 15 to 20. The gap between what exists and what is required is not marginal. It is the gap between a classroom where learning happens and a classroom where attendance is taken.

The subject composition of Indian graduates makes the regional problem worse. Over 30 percent of India’s graduates hold degrees in social sciences — a proportion vastly higher than in China and significantly higher than most comparable economies. Specifically in northern states, the dominant qualification is a general BA: a multi-subject degree that touches History, Political Science, Geography, and sometimes Economics across three years, going deep in none of them. These programmes were originally designed as preparation for public service examinations. Decades later, they still function primarily as that — waiting rooms for government jobs that may or may not materialise.

The consequence, as economist K.R. Shyam Sundar, who has studied the report’s findings, points out, is a vicious cycle in education itself. The shortage of qualified STEM teachers at secondary and higher secondary levels — itself partly a product of the dominance of general arts programmes — means that too few teenagers study science, technology, engineering, or mathematics. Which means that the next round of graduates is, once again, disproportionately arts-trained. Which means that the teacher shortage in science perpetuates itself across generations. The shortage creates the shortage.

IV. The Government Job That Became a Lottery

Ask almost any unemployed graduate in north India what they are waiting for and the answer arrives quickly: a government job. The Union Public Service Commission exam. The state PSC. The Railway Recruitment Board. The public sector banks. The defence forces. These are not just jobs. They are the destination that the entire architecture of general arts education in northern India has been orienting students toward for the better part of five decades.

The preference is not irrational. Government employment offers something that private employment in India, particularly for people without elite college degrees, rarely does: security. Pension. Protection from arbitrary dismissal. And, for those from marginalised communities, something else that the SWI report identifies carefully — protection from discrimination. In private markets and administrative systems, caste and community identity can close doors. Inside a government institution, a hire protected by reservation and seniority rules is harder to dislodge. The preference for government work is not only about the salary. It is also about whether the system will treat you fairly.

The problem is that the vacancies have not kept pace with the aspirants. Since 2016, evidence from government surveys shows a dramatic fall in non-farm private sector job creation — precisely the years when the number of graduates entering the labour market was growing fastest. The result, Shyam Sundar’s research finds, is that the number of young people competing for Union and state civil service, railway, and public sector bank posts has more than doubled from the early 2010s to today. Government jobs, always scarce, have become something close to a lottery.

V. The Inequality That Persists Inside the Progress

The SWI 2026 report notes, carefully and with data, that caste-based occupational segregation has declined. This is real. But it also documents, just as carefully, how uneven that progress has been — and how the baseline disadvantage of marginalised communities shapes not just what jobs they get, but whether they get any at all.

Among Scheduled Caste youth aged 15 to 29, 33.5 percent work as casual labourers — hired by the day, with no contract, no benefits, and no security. Another 41 percent are in self-employment. Compare this with young people from other social groups: 9.8 percent in casual labour, 51 percent self-employed. The self-employment numbers look similar on the surface, but the composition is very different: SC and ST self-employment is more likely to be subsistence — small repairs, marginal trade, family agriculture — while other communities’ self-employment includes more viable small businesses. The sectors tell the same story: 26 percent of SC youth work in construction, 31 percent in agriculture. Only 12 percent are in modern services. For other social groups, the modern services figure is 25.5 percent.

The educational gap compounds this. Among Scheduled Tribe members aged 15 to 29, just 7 percent hold a graduate degree. Among Scheduled Castes, 10 percent. Among other social groups, 18.3 percent. These gaps exist not because of individual choices but because of dropout rates that remain high in marginalised communities — pulled by economic necessity, pushed by the inadequacy of elementary schools in regions where these communities are concentrated. Low-quality primary education feeds low secondary retention, which feeds low higher education access, which feeds low employment prospects. The cycle is closed and self-reinforcing, and it runs along caste lines with a consistency that no amount of aggregate progress fully dissolves.

VI. AI and the Jobs That Aren’t Coming

The crisis of graduate unemployment was building long before artificial intelligence became a mainstream concern. But the SWI report, and the commentary around it, flag AI as a new variable that makes an already difficult equation harder still.

Technological leaps — digitisation in the 2000s, platform economies in the 2010s, AI in the 2020s — consistently favour the well-educated and the well-connected. They create high-value jobs at the top, eliminate mid-level jobs in the middle, and leave low-skill manual work at the bottom largely unchanged. For India’s vast middle tier of general-degree graduates — not elite enough for the high-value AI-adjacent roles, not manual enough to be unaffected by automation of clerical and administrative work — the squeeze is intensifying. Data-entry jobs, basic accounting, routine legal research, customer service — the sectors that have historically absorbed arts graduates in the organised economy — are the exact sectors most exposed to AI-driven displacement.

The irony is bitter. The graduates who were already struggling to find stable jobs are now competing in a market where the mid-level jobs they might have occupied are evaporating at the same moment they arrive to fill them.

VII. What Could Actually Change Things

The SWI report, and economist Shyam Sundar’s commentary on it, converge on a policy argument that is simple, specific, and somewhat counterintuitive: stop paying unemployment allowances to graduates, and use that money instead to fund mandatory apprenticeships in organised industry.

The logic runs as follows. A graduate receiving a modest unemployment allowance from a state government — a policy that several states have either implemented or discussed — has a rational reason to remain unemployed, or at least to defer the decision of what to do next. The allowance sustains the waiting game. Redirect that money toward paid apprenticeships in manufacturing companies, logistics firms, healthcare providers, and technology businesses — as the Apprentices Act of 1961 already requires, though rarely enforces — and you create a different incentive. Employers get subsidised, trained workers. Graduates get work experience and a paycheck. And the economy builds a pipeline between the education system and the productive sector that currently exists more on paper than in practice.

India’s demographic dividend — the economic advantage of having a large working-age population — will begin to shrink after 2030. That is four years away. The window is not closed, but it is closing. The generation that will determine whether the dividend becomes a dividend or becomes a burden is already in college. A large share of them are, right now, staring at a timetable of competitive exam dates and wondering when their degree will finally count for something.

Rajan, in Gaya, is preparing for his next BPSC attempt. He will be among hundreds of thousands sitting the same exam for the same few seats. He is educated, ambitious, and completely clear-eyed about his situation. “If there were jobs,” he said, when asked why he keeps trying, “I would take a job. There are no jobs.” That sentence contains, in compressed form, everything the State of Working India 2026 takes four hundred pages to demonstrate. The report is thorough and important. But Rajan already knew.