EPW    Commentary April 14-20, 2001

Constitution 91st Amendment Bill: A Constitutional Fraud?

Alistair McMillan

The framework of India’s democracy was established in the Constitutional Convention, but one aspect of that framework, which lays down the means of maintaining a basic structure of one-person, one-vote, one-value, has been systematically undermined. This relates to delimitation, the process by which constituency boundaries are drawn up, and states are allocated seats in the national parliament. The subject is a somewhat esoteric one, but that does not mean it should be ignored. Delimitation affects the relative influence of each voter in India, and as such is at the heart of the democratic process.

Delimitation is a particularly sensitive political issue because the people it directly affects are sitting politicians. Changing constituency boundaries can have an immediate impact on an MP or MLA’s chances of re-election, and so the process of delimitation is always carefully monitored. Of course, politicians are not just affected by the process, but are in a position to actively intervene, and not just through influencing the operation of delimitation (or gerrymandering), but by changing the principles and methods laid down by law to regulate the process.

At the time of writing (March 2001), the Constitution (91st Amendment) Bill is before parliament. The bill is a curious one, for it is a constitutional amendment that seeks to prevent change, to block procedures that were established to enhance the representativeness of the Lok Sabha (House of the People). The amendment renews the effect of a similar Amendment (the 42nd) passed during the emergency at the instigation of Sanjay Gandhi, not a person associated with best democratic practice.1 Article 81(2) of the Constitution asserts that ‘there shall be allotted to each state a number of seats in the House of the People in such manner that the ratio between the number and the population of the state is, so far as practicable, the same for all states’. The amendment is designed to prevent adjustments being made which would have changed the allocation of seats to various states to reflect their growing population. The change, which under the Constitution is meant to occur automatically after each census, would have meant that constituency sizes throughout the country are roughly the same size. Preventing such redistribution means that certain states are becoming significantly underrepresented in the Lok Sabha, and the failure to put forward a coherent rationale for such a policy undermines the democratic foundation of the House of the People.

The 91st Amendment Bill seeks to cancel the redistributions due after the 2001, 2011, and 2021 censuses,2 although it does allow the boundaries of seats within states to be redrawn so the large differences in constituency populations within states can be addressed.3 The ostensible reason (and the same one put forward to justify the change in 1976) is that changing the number of seats according to population unfairly punishes those states which have managed to restrict the growth in population. The ‘Statement of Objects and Reasons’ released with the bill states that:

Keeping in view the progress of family planning programmes in different parts of the country, the government, as part of the National Population Policy strategy, recently decided to extend the current freeze on undertaking fresh delimitation up to the year 2026 as a motivational measure to enable the state governments to pursue the agenda for population stabilisation.

In a similar vein, Rami Chhabra, who helped draw up the NPP, has argued that had a delimitation taken place on the basis of the 2001 Census ‘The vast majority of the new seats would have accrued to the very states that had failed to curb population growth rates and improve their population’s well-being, thus further tilting the balance of power in favour of those who had been derelict in their duties’ [Chhabra 2000:34).4

States and the Balance of Power

The connection between the implementation of family planning strategy and the allocation of seats to the national parliament is far too tenuous to be taken other than as a smokescreen for more meaningful political considerations. To deny a person an equal voice in the democratic process because they happen to be living in a state that has had a high population growth rate is clearly undefendable. Such reasoning takes no account of migration between states, which may be just as significant a cause of population fluctuations than birth rate. There is no coherent rationale for fixing the base line for the ‘correct’ population balance at 1971 levels. If the idea were really to encourage best practice in the future the policy should be to peg the levels at the rate they are now, as given by the 2001 Census. As it is, voters in states like Rajasthan and Uttar Pradesh, are being punished for high population growth rates over the last 30 years which they can hardly be held responsible for, and which they can do nothing about now. The states that have actually had the highest population growth since 1971 are the small states of the north-east [McMillan 2000], and the issue does not affect them since they were over-represented in previous parliaments, and are only now catching up with their correct allocation of seats.

The real motivation behind the change in the Constitution is the balance of political power between the regions and in particular, between the north and the south. As K C Sivaramakrishnan notes, the ‘real fear is not about population control but political control’ [Sivaramakrishnan 2000:3093]. Simulations calculated using estimates of the population for 2001 suggest that the southern states (Andhra Pradesh, Kerala, Karnataka and Tamil Nadu) are currently over-represented by some 11 seats, whilst the Hindi-heartland states (Bihar, Jharkhand, Haryana, Madhya Pradesh, Chhattisgarh, Rajasthan, Uttar Pradesh and Uttaranchal) are under-represented by 14 seats.5 The fear seems somewhat over-played. At present the four large southern states fill 23.8 per cent of the Lok Sabha seats; such redistribution would reduce this by about 1.1 per cent. The Hindi-heartland states now return 39.4 per cent of the Lok Sabha seats, and an adjustment after 2001 would be likely to raise this by about 1.6 per cent. There does not actually have to be any reduction in the number of seats that are returned from any states. The effect of the 42nd Amendment was to fix the size of the house, when previous delimitations had seen the total number of MPs rising along with the population. In a bigger Lok Sabha new seats could be allocated to the states whose population has grown rapidly since 1971, without affecting the number of MPs returned from states like Tamil Nadu, where the population growth has been much slower.6

The extent of the distortion has been exacerbated by the creation of three new states in November 2000, Uttaranchal, Jharkhand and Chhattisgarh. This has been done with an ad hoc redistribution of seats based on the current geographical allocation, which, as has been noted above, is extremely out of date, and no consideration of the overall effect on statewise allocations of seats. The five seats which make up the core of the new state of Uttaranchal have an average electorate of 11,70,465, compared to the 80 seats left in Uttar Pradesh which have an average electorate of 12,16,018. The creation of Jharkhand is based on 14 constituencies with an average electorate of 10,65,554, whereas the 40 seats which are to remain in Bihar have an average electorate of 10,97,638. The new state of Chhattisgarh has 11 seats with an average electorate of 11,00,004 compared to 12,00,184 in the seats which remain in Madhya Pradesh. The consequence is that Bihar, Uttar Pradesh and Madhya Pradesh, which are already under-represented compared to other states, have had their proportion of seats reduced even further to facilitate the creation of the new states.7 Again, the danger is that ignoring the worsening under-representation of states like Bihar, Madhya Pradesh and Uttar Pradesh for another 30 years will make their grievance stronger.

Part of the problem following the delay in delimitation after the Third Delimitation Commission carried out its work after the 1971 Census has meant that the effect of any new delimitation is much greater – more seats should be transferred and boundary changes have to be more radical. However, the remedy is to carry out frequent readjustments of constituency boundaries, rather than keep putting off the problem. The 91st Amendment merely begs the question of what happens after another 30 years. The fact that population growth will have been smaller in the south will not have changed, and the danger is that putting off a difficult decision until the future not only delays the inevitable, but makes the change much bigger when it actually happens.

What explains the lack of an informed debate on the delimitation issue? The answer seems to largely rest with a political consensus embracing the main parties and the bureaucracy. At an all-party meeting in May 2000, there was reported to be ‘unanimity among the party representatives on the need to continue the freeze on fresh delimitation’.8 This seems to be partly driven by the regional ambitions of the BJP and the Congress parties, who are unwilling to alienate support, and are probably right in thinking that the losers are more likely to make an issue of any redistribution of seats than the winners. The issue has been sensitised by the key role played in the National Democratic Alliance by parties from Tamil Nadu, and the fact that assembly elections are due in Kerala and Tamil Nadu in May 2001.9 The silence of regional parties from states which would undoubtedly gain from any delimitation, such as the RJD in Bihar, the Shiv Sena in Maharashtra, or the Samajwadi Party and BSP in Uttar Pradesh, is harder to explain. Meanwhile, the Election Commission has repeatedly ducked the issue of an interstate redistribution of constituencies. The chief election commissioner, in an interview with Frontline magazine in March 2001, admitted that postponing delimitation had an ‘illogic’, but suggested that ‘in developing democracies, you have to do unique things which sometimes may look contradictory’.10 There is no bar on constitutional innovation in India, but constitutional reforms should have a transparent rationale and be based on evidence that any proposed change would have the desired effect. The 91st Amendment has no transparent rationale, and the proposed justification for the change is so clearly flawed that it is tantamount to a constitutional fraud.

Logic of Delimitation

The criticism of the 91st Amendment Bill and the tinkering with the delimitation process presented here does not mean that there is no justification in the concern that the Indian Constitution does not adequately address the balance of powers between the different states, and between the states and the centre. The point is that the proposed changes would undermine the democratic legitimacy of the Lok Sabha. The legitimacy of the Lok Sabha derives from the fact that it is a directly represented body that can be seen to reflect the will of the people of India. If that process of representation is seen to be distorted then the legitimacy of the parliament is weakened. Addressing the wider questions of the nature of democracy and the relationship between states requires a wider perspective on the functioning of the Constitution.

B R Ambedkar, architect of the Indian Constitution, argued, in his Thoughts on Linguistic States, that:

This disparity in the population and power between the states is sure to plague the country. To provide a remedy against it is most essential [Ambedkar 1970 [1955]:23].

Ambedkar was worried that the Hindi north would dominate the south, without giving appropriate consideration to the different culture or political requirements. His proposed remedies included dividing the large Hindi-speaking states of the north into smaller units, and providing a second political capital alongside Delhi to reflect the diversity of the Indian state – suggesting Hyderabad as the best choice. The recent creation of new states from Bihar, Madhya Pradesh and Uttar Pradesh goes a little way to reducing the block vote of the large northern states, and there appears to be an administrative logic to the further subdivision of, in particular, Uttar Pradesh. However, the creation of new states is a limited remedy, and fails to address the nature of the relationship between, in particular, the centre and the states.

A far more appropriate candidate for reform is the Rajya Sabha (Council of States). The current composition of the Rajya Sabha means that it is too often just a pale shadow of the Lok Sabha, but one way to protect the interests of less populated states is to reconstitute the Rajya Sabha in such a way as to give them an enhanced say in the parliament. Since the Rajya Sabha is not a directly elected body there is less of a threat to its democratic legitimacy, and since it is called the Council of States then there seems to be no real reason why it should not be used as a conduit for states’ rights. The allocation of seats to different states is laid down in the Fourth Schedule to the Constitution, and it would be simple to readjust this, or (perhaps more coherently) to devise a formula that gives each state a minimum amount of seats and then allocates seats to states according to some weighted measure of their population.11 This would alter the representative basis of the Rajya Sabha, but in such a way that was not open to partisan manipulation, and avoid undemocratic procedures such as nomination to the legislature.12

Other federal parliaments have used bicameralism as a means of entrenching the position of states within the structure of the national constitution. In the US, the Senate is made up of two members from each state, regardless of the population of the state; whereas the number of congressmen each state elects to the House of Representatives is determined on a strict population criterion. In this way the democratic element of the constitution is tempered by the need to consider states as a unit within the union. In Germany, the Bundestag is elected by proportional representation, whereas the Bundesrat is composed of delegates from the Lander. Each house is seen as representing a different aspect of the German federal polity, and the system adapted smoothly to unification in 1990. In both Germany and the US, each house has a veto over legislation and constitutional amendments, but the different houses are associated with different roles and competencies within the government structure.

The fears of the less populated states, or those on the periphery of the Indian state, might be further assuaged if the mechanisms which structure Indian federalism were made more transparent, and the balance of power shifted away from the centre to the states. The way in which powers are distributed between the centre and the states has strongly favoured the centre, and although in recent years the balance has slightly swung back towards the states this has been more a consequence of increased judicial activism and the development of an electoral federalism.13 Formalising a constitutional balance which gave the centre less power to intervene over matters of state governance; which made the flow of revenue from the centre to state less liable to partisan control; and which established clear parameters of state autonomy would go much further towards establishing a stable constitutional basis than trying to crudely manipulate the composition of the Lok Sabha.

The debate on the institutional structure of India, instigated by the setting up of the National Commission to Review the Working of the Constitution in 2000, has too often been couched in terms of a false dichotomy between stability with two-party system and an unstable multipartism. The danger is that methods which seek to suppress regional distinctiveness can actually lead to alienation from the political process, and which can in turn lead to more extreme protest. An alternative way of thinking about reform is to build institutions which are flexible, and which allow a certain level of autonomy to all states without the need to give special treatment to any, and with transparent procedures which govern relations with the centre which cannot be manipulated for short-term partisan gain.

The Constitution (91st Amendment) Bill is being used to twist the democratic framework of the Lok Sabha to reflect the diversity of the Indian states. In doing so it undermines what should be an essential and prized characteristic of the Lok Sabha, that each voter has an equal voice in the national parliament. However, within the scope of the Indian Constitution there exist channels which would be more appropriate for the purpose of entrenching states’ rights in relation to the centre and to other states. These include the composition of the Rajya Sabha, and the mechanisms specifying the federal structure of the union. When considering the bill, the parliament should consider whether a short-term expedient to prevent delimitation should be allowed to undermine the democratic core of the Constitution.


1 Granville Austin, in his recent book on the Indian Constitution, subtitles his chapter regarding the 42nd Amendment ‘Sacrificing Democracy to Power’ [Austin 1999]. The fact that the postponement of delimitation has happened at the same time as rather clumsy attempts to change the nature of the Constitution may not be entirely coincidental. The emergency saw speculation that Indira Gandhi would support a move to a presidential system, believed to be the reason for the setting up of the Swaran Singh Committee, which has some echo in the manoeuvring behind the establishment of the National Commission to Review the Working of the Constitution in February 2000. Ironically, one of the main opponents of presidentialism in the 1970s and early 1980s was L K Advani, one of the proponents of such a change in 1998 (see, for instance ‘The Case against Presidency’ in The Times of India, November 30, 1980, compared to ‘Why Advani Harps on Presidential System’, The Times of India, May 6, 1998; Austin 1999:378).

2 There have been reports that the delimitation process could be restarted after the 2011 Census. No justification has been given for the year 2026, since it does not relate to any census and so would by default actually mean 2031. My own suspicion is that 2026 was picked because it was roughly the same length as the last postponement, and that very few of the current parliamentarians would be around by then to have clear up the mess that had been left behind with the current fudge.

3 For instance in Jammu and Kashmir, the Ladakh constituency has just 64,706 voters, whilst Jammu has an electorate nearly 10 times larger at 6,01,007. In the capital, Chandni Chowk constituency has just 3,76,603 people on the electoral register, an eighth of the size of Outer Delhi with 31,01,838. 4 Why the National Population Policy felt the need to delve into matters of constitutional restructuring and electoral mechanics is unclear. The incongruity between the statements of the NPP on delimitation and the actual proposals for changing patterns of population growth, which move away from focus on incentives and disincentive towards a focus on health care, education, and female empowerment, is glaring. The NPP hardly appears as a neutral arbiter in the current debate, with Gita Sen noting that the National Population Policy 2000 ‘was drafted and discussed almost entirely within a closed circle of the government’ [Sen 2000].

5 Details of the methodology and results of these simulations are given in McMillan [2000, 1274: Table 4]. They are based on a redistribution of seats with the overall size of the Lok Sabha kept at 545, and assuming that for smaller states and union territories the allocation will remain constant. These figures were calculated prior to the creation of three new states in November 2000 (discussed below), but the overall pattern of regional change is not affected.

6 See McMillan 2000 Figure 1 for a depiction of the change in size of the Lok Sabha since 1952, with an estimation of how the size could have increased if there had been no freeze in 1976. Table 5 shows how seats could be distributed in an enlarged Lok Sabha so no state would lose representation.

7 Details of the new state boundaries are contained in the Uttar Pradesh Reorganisation Bill, the Madhya Pradesh Reorganisation Bill, and the Bihar Reorganisation Bill, all passed in the year 2000. In the Rajya Sabha debate on the Uttar Pradesh Bill, Akhilesh Das, MP, argued that Uttaranchal should receive four seats rather than five, leaving Uttar Pradesh with 81 seats (debate on August 10, 2000). Accurate allocations of seats are not possible until the population figures from the 2001 Census are made available. The current calculations are based on the electorates in 1999 of seats which approximately conform to the new state boundaries.

8 The quote comes, rather confusingly, from an article headlined ‘No Consensus on Delimitation of Constituencies’, The Hindu, May 14, 2000. The lack of consensus relates to the mechanics of any intra-state delimitation, in particular whether to ‘rotate’ reserved seats for SCs and STs, and whether any delimitation should be carried out by the Election Commission.

9 An example of the politicisation of delimitation in the run up to the Tamil Nadu assembly elections comes from Murasoli Maran, a DMK leader who was reported ‘taking a dig at the Congress(I)’ by suggesting ‘they had opposed a bill in parliament to freeze the existing number of Lok Sabha seats for another 25 years to ensure that states like Tamil Nadu that did well in family planning did not suffer in any delimitation exercise. “I hope Congressmen... will not pour mud on their own heads”, he said’ The Hindu, February 12, 2001.

10 Interview between M S Gill and V Venkatesan ‘Of a phase of positive reforms’ Frontline 18(5) March 3, 2001.

11 An example of weighted voting which reflects, but is not exactly proportionate to, population size, is the system of vote allocation in the council of ministers of the European Union.

12 At present, the president can nominate up to 12 members of the Rajya Sabha to represent literature, science, art and social service. The president can also (and does) nominate two members to the Lok Sabha to represent the Anglo-Indian community. Nominations to the Lok Sabha are clearly inappropriate, and the representation of Anglo-Indians should, if still deemed necessary, be switched to the Rajya Sabha.

13 Electoral federalism is used here to refer to more than just the increasing prominence of regional parties, but also of a regionalisation of voting patterns for all parties across the whole country, and within states according to different types of election.


Ambedkar, B R (1970): Thoughts on Linguistic States, Bheem Patrika Publications, Jalandhar.

Austin, Granville (1999): Working a Democratic Constitution: The Indian Experience, Oxford University Press, Oxford.

Chhabra, Rami (2000): ‘Baking a Bigger Pie: Linking Women’s Reservation with Delimitation’, Manushi, 121:32-35.

McMillan, Alistair (2000): ‘Delimitation, Democracy and End of Constitutional Freeze’, Economic and Political Weekly, April 8: 1271-76.

Sen, Gita (2000): ‘India’s National Population Policy 2000: A Comment’, Populi, 27(2) [http://www.unfpa.org/modules/populi/issues2000/sept00/viewpoints.htm].

Sivaramakrishnan, K C (2000): ‘North-South Divide and Delimitation Blues’, Economic and Political Weekly, August 26: 3093-97.

Current Issue    Previous Issue    Archives    Subscription
About the EPW    Sameeksha Trust    EPW Research Foundation
Advertising    Announcements    epw@vsnl.com    Home    The Site