DU struggles to fill undergrad seats again: V-C backs CUET admissions, colleges to share data

“Now every stakeholder is aware of the status of every seat. In DU’s CSAS [Centralised Seat Allocation System] admission system, every allocation is made public, for which a very scientific method is adopted.” He added that allocations are carried out solely on the basis of preferences submitted by candidates and the CUET scores at which seats are offered.

On criticism that CUET has led to widespread vacancies, the V-C said the phenomenon pre-dates the entrance test and was present even when admissions were conducted through Class 12 cut-offs.

Citing data from the university’s admission branch, Singh said in 2019 — a pre-CUET, pre-Covid year — only 68,213 undergraduate seats were filled against a total sanctioned strength of 70,735, leaving 3.56% of seats vacant. In contrast, in 2025, under the CSAS, 72,229 admissions were made against 71,642 available seats, amounting to admissions that were 0.65% above the sanctioned strength.

Singh also pointed to instances where colleges admitted more than 200 students against a sanctioned strength of just 11 seats — an over-admission of nearly 1,745% — arguing that such discrepancies made it difficult to track or regulate actual enrolments.

Addressing concerns that prolonged vacancies could lead to closure of courses, Singh said the university had advised all colleges to revisit their seat matrices and propose ways to fill seats that remain vacant despite repeated rounds of allocation.

While colleges may explore reshuffling combinations under BA Programme, Singh said, “The University is clear that no course will be closed.”

Why vacancies persist

The Vice-Chancellor’s statement comes months after DU officials themselves acknowledged that thousands of seats remained vacant as the admissions cycle entered its final stages.

In August, with more than 9,000 undergraduate seats still unfilled, Haneet Gandhi, Dean of Admissions at DU, had told The Indian Express that the vacancies were largely the result of students not filling their preferences strategically in the initial rounds.

By September, DU data showed that seats were unfilled not only in evening and off-campus colleges, but also in some of the university’s most sought-after institutions — St. Stephen’s, Hansraj, Ramjas, Miranda House, Daulat Ram and Indraprastha College for Women on the North Campus, as well as women’s colleges such as Lady Shri Ram, Gargi and Jesus & Mary.

Many vacancies were concentrated in Language courses that have seen declining student interest in recent years, though a few science programmes were also affected.

In the online mop-up round held between September 4 and 7, only around 1,600 seats were filled. When colleges subsequently moved to on-the-spot admissions, more than 7,000 seats across categories remained vacant, including 1,439 unreserved seats, 2,136 OBC seats, and 1,092, 1,528 and 1,263 seats in the SC, ST and PwBD categories respectively.

In women’s colleges, a significant proportion of vacancies were in reserved categories, while in Christian minority institutions such as St Stephen’s and Jesus & Mary, several seats remained vacant under the minority quota.

Speaking to The Indian Express on condition of anonymity, an official said the university has asked all colleges to submit detailed data on course-wise vacancies. “We are analysing how many seats are vacant in courses across colleges. If seats in programmes with consistently low demand are being left unfilled, those seats may be reallocated to more popular programmes to accommodate a larger number of students,” the official said.

Singh, meanwhile, said under the current centralised system, both over- and under-admissions are “controllable and manageable”. Colleges, he said, are allowed to decide how many extra allocations they wish to offer in each course to minimise the number of allocation rounds, with this data being fed into the algorithm that drives the CSAS process.

The centralisation, he added, also allows the university to make predictive analyses about the popularity of programmes, something that was not possible under the earlier cut-off driven model.