Girls in tech: Evaluating IIT’s Supernumerary Seats Scheme

Introduced in 2018, the Supernumerary Seats Scheme seeks to improve gender ratios among undergraduate engineering students at IITs, which have historically been male dominated. This article shows that the initiative has been successful in getting more females into these elite institutions. Further, on average, although girls start off with lower entry ranks, they are able to match up with their male counterparts academically over the duration of the programme.

The Supernumerary Seats Scheme (hereafter, SSS) was introduced in 2018 to address low and stagnant female representation in Bachelor of Technology (B.Tech.) programmes at the Indian Institutes of Technology (IITs) by earmarking 20% of the seats exclusively for female candidates in each IIT B.Tech. programme. The proportion of female enrolment in this elite engineering ecosystem was stagnant at about 8% till 2017, as opposed to the national average of 30% in B.Tech. programmes and about 50% in general science undergraduate programmes (Sengupta et al. 2023). Apart from nudging the IITs towards greater gender inclusion, two stylised facts provided additional rationale for the introduction of SSS. First, while around 20% females historically qualified for the Joint Entrance Examination (JEE-Advanced) each year, less than half of them enrolled in the programmes due to structural factors including location, inflexible course preference, and lack of family support. Second, on average, despite females entering IITs with lower ranks, they were found to consistently outperform their male counterparts in terms of their final CGPA (cumulative grade point average).

**Scheme design: Differences vis-à-vis traditional affirmative action**

While SSS is a major affirmative action scheme, its design is inherently different from the more traditional socioeconomic category-based reservation system in two ways.

One, unlike traditional reservations, SSS is not a ‘quota’, in that a proportion of total seats is not reserved for females. Rather, the scheme involves creation of additional female-only seats, with the number varying across branch-institute pairs as per the need for gender inclusion. The calculation of seats was a one-time effort based on the 2017 gender-wise enrolment numbers. Thus, a branch-institute pair with no female enrolment in 2017 (for example, Electrical Engineering at IIT Bombay) would imply that 100% of the female-only seats in the branch were created as a result of SSS. In contrast, a branch-institute pair with the mandated proportion of female enrolment in 2017 (for example, Biotechnology Engineering at IIT Madras) did not require any additional seats to be created. In other words, the female-only seats include the number of seats occupied by females prior to SSS.¹

Two, in contrast to traditional affirmative action initiatives, SSS functions by initially allocating female candidates into the female-only seats and thereafter filling the gender-neutral seats from the pool of all remaining candidates across genders. As a result, almost all gender-neutral seats are occupied by male candidates since higher-ranked females are already allocated a female-only seat in their preferred department. And since the number of gender-neutral seats is equivalent to the number of seats occupied by male students prior to SSS, there is no loss of seats available to male candidates. Only if there are more than 20% female candidates with higher ranks, then these high-ranking female candidates can also compete with male candidates for the gender-neutral seats. The algorithm thus has been carefully designed with the objective of ensuring that no male candidate allocated a seat in a programme can be ‘displaced’ by a female candidate of lower rank. This also implies that, unlike caste-based reservations, no separate merit list is prepared for female-only seats. Rather, the standard merit lists are used to fill up the seats – only the order in which seats are filled is reversed (that is, female-only seats first followed by gender-neutral).

Importantly, within the female-only category, places are allocated at the intersection of gender and caste to alleviate a ‘creamy layer’ concern, that is, caste and physical disability quotas are maintained within the female-only seats – as is the case with gender-neutral seats.

Analysing the scheme’s performance

The SSS was rolled out in three phases with a target of 14% female enrolment in 2018, 17% in 2019 and 20% for all years from 2020 onwards. We were interested in measuring the performance of SSS at both the extensive margin (that is, whether the scheme has been successful in bringing more females into IITs) and the intensive margin (that is, the performance of females entering IITs after the introduction of the scheme). We were also interested in analysing what the scheme revealed about the underlying branch preferences of females². For the analysis of the scheme’s performance at the extensive margin, we utilised publicly available enrolment data from annual JEE reports. For the analysis of the scheme’s performance at the intensive margin, we leveraged anonymised academic performance data at one of the participating institutes for the cohort that entered in the years prior to and just following the scheme’s introduction. To study female branch preferences, we compared the opening and closing ranks of the gender-neutral and female-only seats for a selected subset of branches. In particular we focused our attention on the ‘rank-gap’ between the closing rank of the gender-neutral seats and the opening rank of the female-only seats in particular programmes.

Extensive margin

At the extensive margin we find that the scheme has been remarkably successful in raising enrolment of female students to 20% across most IITs by 2022 as demonstrated in Figure 1, which presents historical trends in female allotment to the original or first-generation IITs³. The only exception to this is IIT Kharagpur, where enrolment has remained stuck at around 17% since 2020. Institute location plays a crucial role in influencing the decision of females to attend educational institutes and the remoteness of Kharagpur could be a deterrent factor, keeping female students away (Gupta 2012, Gautam 2015, Mukhopadhyay 1994). However, further research is needed to empirically verify the reasons behind the inability of this institute to meet the mandated target.

Figure 1. Percentage of females allotted to first-generation IITs

Intensive margin

We consider the intensive margin by analysing academic performance throughout a student’s B.Tech. journey. To do so we compare the performance of females and males entering IITs in the pre-SSS 2017 cohort with those entering in the post-SSS 2018 cohort (Sharma and Sengupta 2024).

Conditional on obtaining admission via affirmative action, multiple studies provide evidence of ‘mismatch’ and failure to ‘catch up’ among affirmative action candidates (Robles and Krishna 2012, Arcidiacono et al. 2016, de Silva et al. 2021). The ‘mismatch’ hypothesis states that affirmative action leads to admission of students to environments above their calibre, leading to worse outcomes in terms of academic performance (measured by ranks and GPAs⁴ and programme completion (measured by drop-out rates). The ‘catch up’ hypothesis says that students admitted under affirmative action often start off far behind those admitted without affirmative action, and the former may not be able to close this gap.

In our analysis we find clear evidence of catch-up by female students and no evidence of mismatch in terms of academic performance. While females coming to the IITs post SSS enter with significantly poorer average ranks, they exit with similar or better ranks, exhibiting the ability to catch up with students admitted to gender-neutral seats (Figure 2). Although these female students, on average, start with wider gaps in their first semester GPAs, their scores begin equalising soon and the differences are nullified by the end of the programme.

Females entering post SSS are as likely as others to graduate from their enrolled courses, indicating no mismatch. They also have a higher probability of graduating with a B.Tech. degree and that too within the stipulated time. Overall attrition/dropout rates are negligible in both periods. We also find that they complete an average of six additional credits to complete their degree, but they do so without delaying their graduation timelines. All of our findings are robust to the choice of elective courses.

Figure 2. Average CGPA trajectories of female and male students at an IIT in the cohort immediately pre- and post-SSS

Female branch preferences

When analysing what the data reveal about female branch preferences, we find that, across IITs, there are smaller rank gaps between female-only and gender-neutral seats in more competitive and previously more male-dominated branches (Figure 3). In fact, in the case of Computer Science and Electrical Engineering we find no rank-gap in some first-generation IITs, whereas rank-gaps are much larger in traditionally female-dominated branches. Note that the gender-neutral seats are almost completely occupied by male candidates, thus providing a proxy for the opening and closing rank distributions within each programme for male candidates.

This result suggests that branches like Biotech and Chemical Engineering may have traditionally had more females not solely because of inherent female preference for these fields, but also due to other strategic or psychological factors. We offer two potential explanations for this shift of high-ranking females to branches like Computer Science and Electrical Engineering post SSS. First, prior to SSS, females may have been averse to enrolling in male-dominated branches due to the absence of a sizeable peer group that they identified with. Second, gender differences in competitiveness could be correlated with the choice of higher education specialisation too, more so for quantitative fields like Engineering. With the introduction of SSS, it is possible that female candidates’ expectations of the gender mix as well as the level of competitiveness within a programme were tempered, leading them to apply to more selective and competitive branches in greater numbers.

Figure 3. Rank spread for female-only and gender-neutral seats across branches in first-generation IITs

Conclusion

Our analysis presents the first rigorous evaluation of the Supernumerary Seats Scheme introduced in the IITs in 2018 to improve gender ratios in these historically male-dominated elite educational institutions. We show that the scheme has been successful in getting more females into the IIT ecosystem. Further, once females enter, they are able to match up with their male counterparts academically, even if they start with lower entry ranks on average. Females demonstrate remarkable catch-up right from the first year, completely closing the gap with their male peers by the end of their B.Tech. programmes. Once females gain admission, they graduate successfully and within the stipulated time, possibly through additional grit and hard work as suggested by them completing a greater number of credits compared to male students. We also find that SSS is nudging high-ranking females to apply in greater numbers to highly competitive and historically male-dominated branches in IITs like Computer Science, Electrical and Mechanical Engineering.

Our analysis sheds light on the potential issues with the screening process of getting into the IITs in the first place. Research conducted globally suggests that negative marking-based exam measures may underestimate the true academic potential of females, perhaps by rewarding traits that females are disadvantaged at like willingness to guess or attempt (Baldiga 2014, Pekkarinen 2015, Sharma et al. 2025). Another key point is that ‘missing women’ in elite technical programmes is not a story of pure merit or choice. The data show that since the introduction of the scheme, more females are opting for selective programmes that were historically skewed towards males. Hence, in addition to increasing the number of available seats for females, SSS allows females to update their preferences and choices by addressing concerns like expected probability of getting admission and facing isolation or a hostile climate if admitted.

The long-term impacts of the scheme are yet to be studied and would need data from more cohorts and across a greater number of institutions. However, for now it is clear that SSS is a landmark affirmative action initiative that has been remarkably successful in producing a greater number of competent females with B.Tech. degrees across all undergraduate programmes from some of the most elite and sought-after institutions in the country.

Notes:

For example, if there were 15 seats occupied by females in 2017 within a particular branch-institute pair, and the number of female-only seats post SSS is 20, this implies that five additional seats were created.
Certain branches like Computer Science, Electrical Engineering, and Mathematics and Computing are considered to be more selective, prestigious, and known for lucrative placements. These programmes have very high opening and closing ranks as the top-ranking aspirants compete for admission here. These branches have been historically male-dominated unlike programmes such as Biochemical Engineering.
There are 23 IITs located across the country, with varying reputations. The seven original or ‘first-generation’ IITs (Delhi, Bombay, Madras, Kharagpur, Kanpur, Roorkee, and Guwahati) are the most prestigious. Another nine IITs were established during 2008-2012 (Ropar, Bhubaneshwar, Gandhinagar, Hyderabad, Jodhpur, Patna, Indore, Mandi, Varanasi), and seven during 2015-2016 (Palakkad, Tirupati, Dhanbad, Bhilai, Dharwad, Jammu, Goa).
Grade Point Averages, measured at a semester, annual, and programme scale.