There are methods of implementing women’s reservation fairly in the Lok Sabha and in state assemblies. This piece aims to recommend a fair outcome that should be least cumbersome to implement. Variant one: Women’s reservation if the Lok Sabha size was to remain at 543 seatsIf the total strength of the Lok Sabha remains fixed at 543, and there is a commitment to reserve one-third of seats for women (as envisioned in the Women’s Reservation Bill), then the reservation has to come from within the existing seats rather than by expanding the House. Alternative approaches could be as follows.First, rotational reservation of constituencies.This would imply that about 181 constituencies (a third of 543) would be reserved for women in each election. These reserved constituencies would rotate across elections. Male candidates cannot contest from reserved seats during that cycle. There are risks to this approach. One is that it interrupts the continuity for incumbents, whose commitments to their constituency is affected, and that could reduce accountability. It will also cause significant resentment against the women who replace the male incumbent. Please note that no more than 14-15% of all LS seats in recent houses have elected women.A second alternative is to propose dual-member constituencies, i.e. some constituencies could elect two MPs instead of one. The question here is: how will the dual-member constituencies be chosen? That itself will become a matter of great discomfort among the political parties and incumbents. Moreover, it will be complicated to implement as it needs a redesign of the electoral system. Besides, the strongest argument against such an idea is that it will lead to unequal constituency representation (some with two and others with only one seat representatives). However, it could be the most practical of the potential alternative methods.Also read: A New How-To for Delimitation That Balances Fairness and Works PoliticallyA third method could be to mandate by legislation party-level candidate quotas (as has happened in Pakistan for decades with some success, and better outcomes for women). This will involve political parties being required to field at least 33% women candidates across constituencies. This method may well encourage broader systemic change in party structures. However, it may not guarantee 33% women MPs as it depends on election results. Hence, we will revisit this subject later.However, there is a fourth method: proportional ‘top-up’ seats within the 543. Thus, after elections, seats could be adjusted internally. In other words, if the 14% representation of women in the Lok Sabha 2024 continues, then women would fall short of 33%. In that case, some seats could be reallocated via party lists or nomination. However, this would complicate India’s current first-past-the-post system, which may raise constitutional and legitimacy concerns.Thus the conclusion we are coming to, after analysing the possible methods, while leaving the Lok Sabha size at 543, is that all these methods have serious problems.Variant two: Increase the size of the Lok Sabha, then establish a proportional representation system to implement women’s reservation. If we establish a proportional representation system to supplement the first-past-the-post system, can one-third women’s representation be applied fairly, relying mainly on proportional representation alone?We find that that is feasible. If India were to add a proportional representation layer alongside the current first-past-the-post system, then achieving one-third women’s representation becomes much easier and more precise. But whether it is fair depends on how the hybrid system is designed.In a proportional representation system, parties submit ranked candidate lists, and seats are allocated based on vote share. This allows built-in gender balancing. What we would propose is a mandatory one-third women in party lists. This is unlike what is known as a ‘zipper system,’ in which the party lists alternate man–woman–man–woman names which guarantees close to 50% women, easily exceeding the 33% target. This is used in countries like Sweden and Norway. In this system, the law will require at least 33% of candidates on party lists must be women. Since proportional representation allocates seats from lists, this translates fairly directly into seats.In other words, what we suggest is reserved “top-up” proportional representation seats for women. It would mean that the 543 are left untouched by women’s reservation. Our proposal would keep first-past-the-post seats as they are. We then add proportional representation seats to ‘top up’ women’s representation to 33% if a shortfall exists. This is similar to mixed systems in Germany and New Zealand (though not specifically for gender).Why proportional representation works better than first-past-the-post for this goalUnder first-past-the-post (used in the Lok Sabha) you can reserve constituencies for women, but it disrupts voter choice; it requires rotation, and you cannot guarantee outcomes with just candidate quotas. Under proportional representation, by contrast, the seat allocation is controlled at the list level, and gender quotas can be enforced mathematically and transparently.However, relying mainly on proportional representation raises some trade-offs – this needs to be recognised. Its benefits are that we get to exact 33% or more, while ensuring there is no need to block male candidates from constituencies, and it encourages parties to promote women systematically. It also ensures better reflection of vote share into seat share, which is a huge improvement on the first-past-the-post system. How unfair the first-past-the-post can be is demonstrated by a few examples from national and state election outcomes. For instance in the 2014 general election, Bharatiya Janata Party’s vote share was ~31% but its seat share was 282/543 (~52%). With just about one-third of votes, BJP got a clear majority. Meanwhile, the Congress, with a vote share of ~19% , obtained only 44 seats (~8%). Thus, Congress got 1/5th votes but barely any seats.Similarly, one state election example will suffice. In Uttar Pradesh (2017 assembly election), BJP got vote share of ~39.7% , but seats were as many as 312 out of 403 (~77%). Thus BJP appeared to win a landslide, with just ~40% votes. Yet it ended up with three-fourth of the seats. On the other hand, the Samajwadi Party and Congress alliance’s combined vote share was ~28%, yet they obtained only 54 seats, i.e, significant votes, but very few seats, when they should have got 113 seats. Hence, the argument can be made for first-past-the-post to be supplemented by using proportional representation. Of course, a worry could be that proportional representation weakens constituency links – a core feature of first-past-the-post in India. Also, party leadership gains more power in deciding who gets elected. Indeed, it does mean that voters choose parties more than individuals. Its great value lies in the fact that the full disruption of the current system is avoided, while still ensuring guaranteed representation of women, as we will show. More importantly, it ensures fairness of vote share translating into seat share.How the first-past-the-post plus proportional representation system will work to ensure women’s reservation for womenProportional representation is seen as attractive primarily, though not only, because it guarantees outcome. India directly reaches ~33% women MPs, the goal of the Women’s Reservation Act of 2024. Besides, it has the added advantage of no disruption of constituencies, for the most part (as we propose) as it avoids reservations between first-past-the-post based constituencies. In any case, it is politically smoother, as most sitting MPs are not displaced.However, what is proposed is that the proportional representation is applied: only to ensure India reaches 33% seats for women, so it will be used only to elect women; and only to added seats, not for the most part to the existing first-past-the-post constituencies. How this will work is discussed below. First, an expansion of the Lok Sabha can be justified by population growth (India hasn’t increased seats since the 1971 freeze), and the principles for that have already been outlined in an earlier paper. Note also that the size of the Lok Sabha should not increase inordinately, as seems to have been suggested by the ruling party at the Union, to 816 seats, which amounts to an excessive and needless 50% increase in size. That would make the Lok Sabha into a political rally, not a parliament of the largest democracy of the world. Parliament should spend time debating and discussing issues of national importance, apart from legislating in accordance with India’s constitution.Hence, we suggest here that an increase of the Lok Sabha to 651 seats will ensure that the Lok Sabha gets 33% of 651 as women Lok Sabha members (or 216 total women). If half of this number of 216 (i.e. 108) was to be elected using first-past-the-post (with states receiving women seats in accordance with their share in total seats in the Lok Sabha of 543 members, then the remaining half can be elected using proportional representation in the expanded Lok Sabha. 651-543 = 108. Thus, we should only add 108 seats to India’s parliament of 651 members, with half of women elected from constituencies (i.e. 108 based on first-past-the-post, as now), and the remaining half (also 108) based on proportional representation-based party lists.Of course, in a federal country, we need to take a call on whether the party lists of women under the proportional representation system will be at the national or state-level. State-level is more federal and likely more acceptable. Hence, a constitutional change will be required, as an amendment alongside reforms similar to the Women’s Reservation Act.Hence, we now proceed to estimate how it works for a Lok Sabha of 651 seats.We reemphasise that it is very inappropriate to have a Lok Sabha larger than 651 seats. The latter size is itself very large for proper debate to occur unless the number of days that the Lok Sabha sessions occur in a year do not significantly increase, going back to levels prevailing in the 1950-1975 period, when it met for 135-140 days per year. By the early 1990s that had dropped to 95-100 days. The most spectacular drop has occurred since the mid 2010s to 55-65 days onlyConstitutional and electoral framework will need readjustmentTo implement this, India would need a carefully designed hybrid system. First, a constitutional Amendment will be required. We will need to amend Article 81 (Composition of Lok Sabha), and increase cap from 543 to 651 seats. Also, it will be required to modify Representation of Peoples Act provisions, and explicitly allow mixed electoral system (first-past-the-post and proportional representation), as well as state-level proportional allocation, and link with the Women’s Reservation Act. We will need to add a clause, “Not less than one-third of total seats shall be held by women,” and allow fulfillment via proportional representation top-up mechanism, not just constituency reservation. Whether we need to amend the Representation of the People Act depends on how the Delimitation Commission suggests that first-past-the-post seats for women will be elected. For example, there could be provision made for introducing Dual Voting System, in which each voter gets a constituency vote (first-past-the-post) and a party vote (for proportional representation seats). Moreover, each party submits a state-wise ranked list of women candidates for the state-level party vote.Further, no delimitation will be needed to redraw constituencies for half the seats for women (the first-past-the-post ones), if a dual member constituencies method is adopted. Then there is the question of Scheduled Caste and Scheduled Tribe reservation, and that intersects with women’s reservation. In the current Lok Sabha (total 543 seats) reservation is as follows: Scheduled Castes: 84 seats; and Scheduled Tribes: 47 seats. Total reserved seats for Scheduled Castes and Scheduled Tribes is 131. However, we have not suggested anything in this regard, and leave that judgment call to political parties and the Delimitation Commission. These Scheduled Caste/Scheduled Tribe reservations are provided under the Constitution (Articles 330–332) to ensure representation for historically disadvantaged communities, and the allocation is based on their population share across states.We make no proposal for how women’s reservation could be implemented while Scheduled Caste/Scheduled Tribe reservation is also in place. Scheduled Caste/Scheduled Tribe reservation will need to apply to the proportional representation portion of the Lok Sabha, otherwise, their share will drop in the House. There could be an intersection between Scheduled Caste/Scheduled Tribe and women’s reservation, and in principle there are two ways to integrate them. The first is “parallel quotas”, not overlapping between women’s reservation and Scheduled Caste/Scheduled Tribe. That way there are no sub-quotas required, and Scheduled Caste/Scheduled Tribe representation remains unchanged. The second is “nested quotas” which is a more socially balanced system. Here, you ensure representation of Scheduled Caste women and Scheduled Tribe women within Scheduled Caste/Scheduled Tribes seats. Thus, 84 Scheduled Caste seats means ~28 should be women. And 47 Scheduled Tribe seats means ~16 should be women. Thus, we will count Scheduled Caste/Scheduled Tribe women elected in the first-past-the-post system and use proportional representation seats to top up Scheduled Caste/Scheduled Tribe women specifically. Scheduled Caste/Scheduled Tribe quotas should extend into proportional representation seats. We may require parties to submit three lists: a General women’s list, a Scheduled Caste women’s list, and a Scheduled Tribe women’s list . The legal architecture will need to align with the constitution of India and the Election Commission of India. These issues should be discussed in the special session of parliament in April.Santosh Mehrotra is a former professor of Economics, JNU.