PLFS CY2025 · the fresh-graduate window · 63,865 people surveyed

The qualification arrives.
The job does not.

The education ladder read at the moment it is sold: the first three years after each course normally finishes. That is the window a placement brochure is describing, and the only one comparable to a college's published placement figure. 60.0 million young people, every rung from below-middle-school to a medical master's.

41.6%
of fresh engineering graduates in the labour force could not find work
Ages 22–24 · 1.16m people
1.7%
for people the same age who never finished middle school
Same measure, bottom rung
34.6%
of fresh engineering graduates hold a formal job — against 52.6% across the whole 21–34 stock
Median entry wage ₹30,000
7.2%
for the 13.1m fresh graduates holding a general degree — four in five of all graduates
Unemployment there: 34.4%

The ladder at the moment of graduation

Each rung measured in its own window. The teal band is a formal job — regular salaried work with a written contract or social security, the thing a degree is sold as buying. It never passes half on any undergraduate rung. The crimson band is people actively looking and not finding, and it grows with every step up.

Faded rows are school rungs held at 21–23 for comparison, not fresh-graduate outcomes

Most of the degree's advantage has not arrived yet

The same rung read two ways: the outline is formal employment across the whole 21–34 stock, the solid bar is the fresh window inside it. Engineering ends up at 52.6% and starts at 34.6%. Whatever a placement figure is measuring, it is measuring the solid bar.

Formal work, fresh window Formal work, whole 21–34 stock

The ladder runs backwards

Unemployment as it is normally reported — people seeking work as a share of the labour force they are in, which excludes students and those at home. Someone who left school before middle school is almost never unemployed, because they take whatever work exists. The queue is a graduate phenomenon.

Unemployed as a share of the labour force · fresh-graduate window · CY2025

What the formal job turns out to be

Job function among those in formal work, from this analysis's committed NCO map. An engineering degree is in practice a software qualification; medicine is the only field whose degree names its job; a general degree has no characteristic destination at all — its largest function covers under a fifth of its holders. Wages are the median for that rung's formal workers.

Whole 21–34 stock · the job-function cut is not published by window

For women, the missing jobs show up as absence, not as unemployment

On the general degree — India's graduate labour market, four in five of all degree holders — 50.2% of the women are recorded as doing domestic duties against 0.2% of the men, and 9.5% hold formal work against 22.8%. The gap is not in qualification. It is in what happens after one.

Whole 21–34 stock · the gender cut is not published by window
The numbers behind every chart
Method, and what these numbers can't tell you

Source: Periodic Labour Force Survey (MoSPI) person-level microdata, calendar-year 2025 release, first visit only, ages 21–34 narrowed to each rung's own window. 63,865 people surveyed. Activity is usual principal status — a person's main activity over the previous 365 days, not one week. Every number here is generated from this analysis's own extracts; nothing is typed into the page by hand.

  • The window. Three years from each course's normal completion age: engineering 22–24, MBBS 24–26, three-year degrees and diplomas 21–23. Postgraduate rungs add two years to the underlying degree — that part extends the house rule rather than quoting it. School rungs have no graduation in this age range at all; they are held at 21–23 as a same-age floor and are faded on the chart for that reason.
  • "Formal job" is strict, and deliberately so. Regular salaried work with positive earnings and either a written contract or social security. A regular salaried job with neither counts as informal wage work, alongside casual labour. Loosen this and every rung's headline roughly doubles, which is why the definition is stated rather than assumed.
  • Bars close to 100 using an explicit residual. Unpaid family workers, self-employed reporting no positive earnings, rentiers and those unable to work sit outside the six named categories; they are carried inside the fourth band rather than normalised away. On the general-degree rung that residual is large and is mostly unpaid family work.
  • Two cuts are whole-stock, not fresh. Job function and the gender split come from the rung-detail extract, which is published across 21–34 only. They are labelled as such wherever they appear and should not be read as fresh-graduate figures.
  • The sample is clustered by village and urban block; the table gives the cluster count beside the sample size for every row. Treating respondents as independent understates the standard error by roughly 3.5 to 4.7 times, so read adjacent rungs as indistinguishable unless the gap is large. Rows under 250 respondents are flagged thin — postgraduate engineering and postgraduate medical are directional only.
  • Headcounts are lower bounds. PLFS weights sum to its own projected population, anchored to Census 2011 — roughly 15–25% below India's true current population. Shares are unaffected.
  • Wages are sample medians, not weighted ones. BigQuery's APPROX_QUANTILES ignores the survey weight, so the median sits beside weighted shares and is not on the same footing as them.