A young Pakistani can spend sixteen years or more in school, earn a university degree, and still enter a labour market where the first salary barely covers the cost of living. If education is supposed to raise a person’s earning power, why do so many graduates struggle to turn it into economic security?
The reflex answer is that there are too many graduates and too few jobs. The evidence is messier. The latest Labour Force Survey, which the Pakistan Bureau of Statistics ran from July 2024 to June 2025, so it describes last year’s market rather than this one’s, puts overall unemployment at 7.1 per cent. Among degree holders it was 10.9 per cent, down from 16.3 per cent in 2020-21. Graduates were only 8.9 per cent of the employed. That is not what a glut looks like. Enrolment has turned too: Gallup Pakistan’s reading of Economic Survey figures shows university enrolment falling from 2.23 million in 2020-21 to about 1.96 million in 2024-25. Families may already be doing the sums.
The sharpest exception is women. Among those with a degree or higher, 23.9 per cent were unemployed, in a country where women are 48.8 per cent of the working-age population but 25 per cent of the labour force. Here, education has plainly outrun opportunity.
Unemployment is the wrong place to stop, though. A graduate who accepts any job leaves the statistic, and what matters is what the job offers. Almost half of all employment, 46.8 per cent, is informal and non-agricultural; only 18.1 per cent is formal and non-agricultural. The average paid employee earned Rs39,042 a month in 2024-25, but Rs54,038 in the formal sector and Rs30,834 in the informal one. Nominal pay rose 62 per cent from 2020-21. By my own arithmetic from annual CPI averages (12.2, 29.2, 23.4 and 4.5 per cent for 2021-22 to 2024-25), prices rose roughly 87 per cent over the same span, implying a real fall of about 13 per cent. Survey definitions changed between rounds, so read that as a direction, not a measurement.
Starting salaries are harder to pin down. I found no official series and no credible national survey of graduate pay. Indeed’s compilation of advertised new-graduate pay, drawn from 825 postings over three years, puts the median near Rs40,000, with the bottom tenth below Rs23,100 and the top tenth around Rs70,000. Advertisements are not pay slips, but the spread is the point: the same degree can lead to very different first salaries depending on which door it opens. Whether Rs40,000 supports an independent adult in Lahore or Karachi depends on rent and transport costs I could not verify, so I will not invent a living wage.
Does the degree still pay? On average, probably yes, but the research is unflattering about how much and for whom. A 2023 study from the Pakistan Institute of Development Economics found that returns to higher education without a professional degree, relative to lower levels of schooling, have been falling at an increasing pace since 2010-11. A 2024 paper using 2019-20 household data found that, once selection and ability bias are corrected, schooling beyond what a job requires adds nothing to pay. Both rest on older data than the latest survey, and I could not obtain wage-by-education tables from it. The best-supported reading is that the premium has narrowed and increasingly depends on field, credential and sector.
On skills, employer complaints about communication, digital ability and experience are everywhere, but Pakistan’s evidence consists mostly of small studies, and I could not verify a national employer survey. Employers who demand experience for entry-level roles are also part of the problem, since nobody acquires five years of it in a first job. The supply-side story is plausible but thinly documented.
The demand side stands on firmer ground. The World Bank describes firm productivity growth as sluggish and notes that exports have fallen from about 16 per cent of GDP in the 1990s to around 10 per cent in 2024. Growth reached 3.7 per cent in 2025-26, the highest in four years, yet short of the government’s 4.2 per cent target. Professionals are 6 percent of employment, and managers 2.2 per cent. An economy of small, informal, low-productivity firms cannot pay graduates much, however well trained. That is a hypothesis the data lean toward, not one they prove.
Migration is where some graduates vote with their feet, though less than headlines suggest. Bureau of Emigration figures reported for the period to May 2026 show 278,536 workers going abroad, of whom 172,006 were labourers; 1,284 were doctors, 2,341 engineers and 487 nurses. Remittances sustain households and the balance of payments, and workers abroad often earn far more. Still, a country that pays to train a doctor and watches her practise elsewhere has subsidised another economy.
Artificial intelligence adds a question without a Pakistani answer. US payroll data show employment of 22-to-25-year-olds in the most AI-exposed occupations falling 13 per cent relative to less exposed ones, according to Stanford researchers in 2025. I found no local equivalent. What we do know is that ICT-led services exports grew 16.5 per cent in the first half of 2025-26, and software development is among the occupations that study flags as exposed.
None of this argues for fewer universities. It argues for attention to what lets firms pay well: export-orientated sectors, credit for firms able to scale, formalisation that gives graduates contracts and pensions, apprenticeships that supply the experience employers demand, and labour-market data good enough for a student to see, before enrolling, what a degree has actually earned. A curriculum chasing every trend would only repeat the mismatch with a delay.
A degree should not guarantee a high salary. But an economy in which educated young people are repeatedly underemployed, underpaid or pushed abroad is telling us something about itself. Perhaps Pakistan’s graduate problem is not that too many young people hold degrees, but that too few parts of the economy have learnt what to do with them. If Pakistan is educating more of its young, why is the economy not paying more of them for what they know?
