India's Higher-Education Census Is Two Years Behind. Here's What the Numbers We Do Have Are Showing

~ Edcom Research Desk, Akshara J Menon, 8 min read.

India's most recent full undergraduate census is now two years old. A cross-check of that survey against the one live regulator dashboard available in the meantime shows a higher-education landscape moving faster, and more unevenly, than expected.

India's National Education Policy commits to putting half of the country's 18-to-23-year-olds from in higher education by 2035. The government plans to add 3.5 crore (35 million) new seats to colleges across the country. The single dataset built to track that goal-AISHE, the government's own annual survey of every college and university in the country-hasn't produced a full public report since 2023-24. Surveys for 2024-25 have reportedly already been collected but not yet released; RTI responses cited in national media (The Telegraph, among others) describe them as still "under compilation." This status has not been independently re-confirmed via a fresh RTI filing for this piece.

That means the enrolment figures currently in public circulation represent the state of India's higher education system two admission cycles ago. For a system enrolling 3.46 crore undergraduate students in 2023-24, such a lag is non-trivial. Analysis of the available evidence indicates that several major undergraduate programmes have experienced notable shifts during the intervening years, highlighting the limitations of relying solely on the latest published national census.

This analysis reads the 2023-24 AISHE data-the newest full census available-against itself, and against the live dashboard the regulator for engineering (AICTE) continues to publish in the meantime. Three things stood out.

1. Some of India's fastest-growing degrees aren't the ones getting the attention

Pharmacy enrolment more than doubled in five years-2.94 lakh students in 2019-20 to 5.91 lakh by 2023-24, a trend so consistent across all five surveyed years that its statistical reliability score (R²=0.98, on a scale where 1.0 is a perfect straight line) is among the highest of anyprogramin the dataset. Nursing shows almost the same shape-2.89 lakh to 5.45 lakh, R²=0.995-plausibly tracking sustained overseas demand for Indian-trained nurses in the UK, the Gulf and Australia.

Business education (BBA) posted the single cleanest growth line in the entire survey: 5.29 lakh to 7.86 lakh students, growing at 10.8% a year with an R² of 0.988-a trend so linear it barely needs the statistics to make the point. B.Ed (up to 15.52 lakh) and law (up to 5.12 lakh) both grew at a reliable 4.7–7.0% a year. None of these four programmes gets anywhere near the media attention that engineering or medicine do, despite outgrowing both of them on the government's own numbers.

2. Engineering: what the AISHE figure measures, and what AICTE measures instead

AISHE's 2023-24 window shows engineering enrolment recovering strongly-36.44 lakh students in 2019-20, up to 44.16 lakh by 2023-24. The previous five-year window (2017-18 to 2021-22) had shown the sameprogramessentially flat to declining, with a trend line so unreliable it carried an R² of just 0.14-statistically, close to noise. The newly extended series carries an R² of 0.90, a reliable and consistent trend, at a growth rate of 4.9% a year.

It is worth being precise about what that 44.16 lakh figure actually is: AISHE's Box 14 counts every student currently enrolled across all four years of the BTech/BE program in a given survey year-first-year through final-year students, added together. AICTE, the regulator that approves engineering seats annually, publishes a different kind of number on its dashboard: approved intake capacity and actual admissions for that specific academic year alone-new entrants, not the standing total. The two describe related but distinct things, and one cannot be substituted for the other without accounting for how a change in annual admissions works its way through to total headcount over the following three to four years.

With that distinction in mind, AICTE's own year-on-year figures for engineering are still informative on their own terms:

Source: AICTE Statewise/Programmewise Dashboard-Engineering & Technology, Level = UG, Institute Type = All (BBA/BCA institutes excluded, per the dashboard's own note). Figures read directly from the dashboard; 2026-27 enrolment is not yet available as that admission cycle is still underway.

New admissions into engineering have risen every year since 2020-21, and the vacancy rate has narrowed from 43% to under 16% by 2025-26. AISHE's newly released 2023-24 figures confirm that this admissions recovery has already worked its way into total enrolled headcount, which AISHE itself reports rose 21.8% over the 2019-20 to 2023-24 window.

3. The clearest trend in the data isn't about which degree students choose-it's who's showing up

Set the program-level churn aside for a moment, and the single most consistent pattern in the 2023-24 AISHE report is a gender one: female Gross Enrolment Ratio (GER) has now exceeded male GER for seven consecutive years-31.2 for women against a national average of 30.0, up from 22.9 for women in 2014-15. This is the seventh straight AISHE cycle in which that gap has run in the same direction.

The pattern holds, and in places widens, once the data is broken down further. Over the five years to 2023-24, enrolment among Scheduled Tribe women grew 39.2%, ahead of the 33.7% growth for Scheduled Tribe students overall. Scheduled Caste women's enrolment grew 25.2%, ahead of the 23.2% growth for Scheduled Caste students as a whole. Since 2014-15, the gap is starker still: ST female enrolment is up 102%, against 75.7% for the ST category overall.

A second, structural pattern sits alongside it: government institutions, though a minority by count, account for a disproportionate share of who they teach. Government-run universities account for 68.1% of total university-level enrolment; government colleges are 17.1% of all colleges but account for 28.6% of college-level enrolment. Private unaided colleges run the opposite way-70.0% of all colleges, but only 50.8% of enrolment-consistent with a landscape of many smaller private institutions rather than a few large ones.

Neither pattern gets much space in the degree-by-degree growth conversation that usually follows an AISHE release, but both describe the same underlying system every program-level number in this piece sits inside.

What the fuller picture looks like

L = lakh, Cr = crore. Growth rates via OLS log-linear regression on AISHE's full 2019-20 to 2023-24 series unless otherwise noted. For BTech/BE, AISHE's figure is a total-enrolled headcount across all four years of the program; AICTE's figure is annual new-admission intake-the two are related but not interchangeable. MBBS enrolment growth (AISHE) sits alongside NMC's own seat-matrix data, which shows approved MBBS seats growing from 92,065 in 2021-22 to over 1.28 lakh by 2025-26-capacity expanding roughly as fast as, or faster than, enrolment. The “Proj. 2025-26” column is the projection this desk estimated from the previous (2017-18 to 2021-22) AISHE series; measured against the actual 2023-24 figures now released, two years ahead of that horizon, BCA has already overtaken its projected 2025-26 mark (8.65 L actual against 7.10 L projected) and MBBS has nearly closed the gap (4.05 L against 4.21 L); most other programs remain below their projection and on pace to reach it as scheduled, while BCom (34.10 L against 42.46 L projected) is furthest off track given its reversal into decline.

Why this matters now

None of these findings required new data collection-they came from reading government sources against each other and against themselves, rather than in isolation. That distinction has real stakes for the roughly 2.5 crore students who sit a board exam and choose a degree every year: a growth rate built on an R² of 0.01 is not the same fact as one built on 0.98, even though both circulate as a single confident-sounding percentage. And with the next full national census (2024-25) still unpublished, the numbers currently shaping that choice are, for several fast-moving fields, already out of date.

This isn't a comment on any one institution's diligence-AISHE and AICTE are each publishing what they're mandated to, on their own schedules, for their own purposes, and measuring what they're respectively designed to measure. It's a case for reading official data sources together rather than in isolation, and for updating the public conversation as soon as better data allows.

Methodology & sources

●        AISHE 2023-24 Final Report-Box 9 (total UG enrolment & official CAGR), Box 14 (programme-wise five-year time series, regular mode), Section 2.2.5 (programme-wise enrolment, actual response).

●        AICTE Statewise/Programmewise Dashboard-Engineering & Technology, Level = UG, Institute Type = All (dashboard.aicte-india.org), figures captured directly from the dashboard, July 2026.

●        National Medical Commission-MBBS seat-matrix data, 2021-22 and 2025-26.

●        Growth rates for the ten non-engineering programmes computed via OLS log-linear regression on the full five-year 2019-20 to 2023-24 AISHE series; engineering figures cross-reference AICTE's more current data, given the two sources measure related but non-identical things (total enrolled headcount vs annual new-admission intake).

© 2026 Edcom. All rights reserved. For permissions and media enquiries, contact advisory@edcom.in