Notes on the Idea of, and Power to, Amend the Indian Constitution

Constitutional amendments are, in any event, a terrific subject for a blog post. I also happen to have an exam on the subject tomorrow, which may perhaps explain the timing. This post is therefore written principally as (1) a revision aid and (2) to put some of it down in a more permanent form here on the blog. The questions involved are complex, and most of the cases that today form the basis of the law are considerably longer than one would ideally want them to be. A certain amount of condensation was therefore unavoidable, and some of the points may have lost their nuance in the process. For all such condensation that went too far, I apologise in advance. 

A good place to begin is with the text of Article 368 itself. It gives us three routes by which the Constitution may be amended:

• Simple majority: Certain provisions can be amended through the ordinary legislative process, and technically fall outside the provision.

• Special majority: Most provisions require a majority of the total membership of each House and a ⅔ majority of members present and voting.

• Special majority + State ratification: Amendments affecting specified federal provisions require the special majority in Parliament and ratification by at least half of the State Legislatures.

One can therefore see that the Constitution builds in different degrees of difficulty for different kinds of change. At the easier end, matters such as the creation or alteration of States are left to ordinary legislation -- and quite sensibly so -- because a constitutional order cannot, for instance, treat adjustments to its territorial map as a threat. The special majority, which comes next, is the ordinary rule for amendments under Article 368. And at the highest level, finally, are amendments concerning federal provisions, where the special majority is accompanied by ratification by at least half of the States. This additional requirement is because where the amendment affects the federal distribution of power, the States themselves have a stake and must therefore participate through ratification. 

The text of Article 368, however, only tells us how an amendment is to be made; it tells us very little about what Parliament may amend. Could Parliament, tomorrow, amend the Constitution so as to remove the protection for free speech? If the procedure were all that mattered, the answer would appear to be yes. Google Blogger LLC could shut up shop, and this blog would have no space, provided the prescribed procedure had been followed. 

This is where Kesavananda Bharati v. State of Kerala enters the picture. The 13 judge bench answered the question largely in the negative. Parliament’s power to amend the Constitution, it held, was not unlimited, and could not be used to damage or destroy its ‘basic structure’.

The majority’s conclusion relied on several lines of reasoning. Two of which are particularly worth noting:

• First, the ordinary meaning of “amendment” suggests alteration or improvement, rather than the destruction of the thing being amended. An amendment may change the Constitution, but it does not follow that it may abolish it altogether. The Constitution also elsewhere uses the expression “amendment by way of addition, variation or repeal”, while Article 368, in its operative provision, speaks simply of the power to “amend”. The distinction suggests that the constituent power under Article 368 did not extend to wholesale repeal. Furthermore, the language that “the Constitution shall stand amended”, together with the reference to a “change in” rather than a “change of” specified provisions clearly presupposes the continued existence and identity of the Constitution being amended. 

• Second, the Constitution was the product of a political movement directed against colonial rule, and many of those who led that movement went on to frame the Constitution that followed it. It is therefore difficult to suppose that the framers intended to leave entirely unprotected the principles of democratic government for which that movement had, over decades, claimed to stand. 

If this is correct, the constitutionality of the doctrine must also answer the counter-majoritarian difficulty that the doctrine appears to create (unelected judges overruling the will of elected representatives). Readers will recall that Article 245 confers upon Parliament the power to make laws, while Article 13 provides the basis for judicial review where those laws contravene Part III. This is now a fairly familiar feature of written constitutions -- originating in Marbury v. Madison (1803) (if you are reading this on a laptop, the image to your right is from Marbury; the connection with Article 13 should, by now, be apparent). 

The position is rather different, however, when one looks at Article 368, which confers upon Parliament what is described as the constituent power to amend the Constitution. Constituent power is, quite literally, the power to constitute i.e. to establish or bring something into being. And what does the Constitution constitute? In our case, quite literally, the Republic of India. The power to amend, therefore, is a power exercised within which all ordinary legislative power itself operates.

On this note, a detour into the early years of the Republic becomes apt. Much of the debate -- back when -- was, in essence, a debate over what constituted “law” for the purposes of Article 13. As it happens, this question went on to shape much of the discourse in Kesavananda as well.

1. Shankari Prasad v. Union of India (1951) 

In Shankari Prasad, the First Amendment was challenged on the straightforward argument that Article 13 prohibits the State from making any “law” which abridges a fundamental right. A constitutional amendment, the petitioners argued, was surely law; and if so, Parliament could not use the amending power to do what it was forbidden from doing through ordinary legislation. The Court, however, unanimously disagreed. It held that “law” in Article 13 meant ordinary legislative law, and did not include an amendment made in exercise of the constituent power under Article 368. Parliament could therefore amend Part III like any other part of the Constitution.

For the moment, therefore, the amending power remained substantively unlimited. The Constitution prescribed how Parliament was to amend it, but imposed no judicially enforceable restriction upon what Parliament could amend.

2. Sajjan Singh v. State of Rajasthan (1965) 

Fourteen years later, Sajjan Singh largely reaffirmed Shankari Prasad. The majority saw little reason to disturb the earlier position. More interesting, however, were the doubts expressed by Hidayatullah and Mudholkar JJ. 

Hidayatullah J. found it difficult to believe that the carefully framed guarantees of Part III were intended to become mere “playthings of a special majority”. Mudholkar J. went somewhat further and wondered whether the Constitution possessed certain “basic features” which the framers intended to preserve, and whether destroying those features could properly be described as an amendment at all (I have written in some detail about Mudholkar J.’s opinion here, for those interested).

(c) I.C. Golak Nath v. State of Punjab (1967)

The unease expressed by Hidayatullah and Mudholkar JJ. became a holding in Golak Nath. By the narrowest possible majority -- of 6:5 -- an 11 judge bench held that Parliament could not amend the fundamental rights in Part III. A constitutional amendment was “law” for the purposes of Article 13 and was consequently subject to its prohibition against abridging fundamental rights. The result was that fundamental rights were placed beyond Parliament’s future amending power, although the Court employed prospective overruling so that earlier amendments were not thereby invalidated. 

Yet Golak Nath created an obvious difficulty of its own. If Parliament could not amend Part III at all, the rule would apply not only to amendments restricting fundamental rights, but potentially even to those expanding them. Suppose, for instance, Parliament wanted to amend Article 22 to give an arrested person an additional procedural safeguard. That amendment would still alter a fundamental right and, on the Golak Nath approach, could therefore be impermissible. More fundamentally, the judgement located the restriction on the amending power in Article 13, rather than in Article 368 itself. . 

(d) Kesavananda Bharati v. State of Kerala (1973) 

The 24th amended Article 13 to undo Golak Nath by stating that nothing in Article 13 would apply to an amendment of the Constitution made under Article 368. At the same time, Article 368 was recast to describe Parliament’s power as a “constituent power”, expressly permitting amendment “by way of addition, variation or repeal”. So, in effect, if Golak Nath had derived a substantive limitation upon the amending power from Article 13; Parliament via the 24th amendment removed the very premise upon which that limitation rested. 

The Kesavanada bench accepted Parliament’s position to the extent that Golak Nath had rejected it. The amending power was plenary in the sense that no particular part of the constitutional text, including Part III, was placed altogether beyond Parliament’s reach. Fundamental rights could therefore be amended; and Golak Nath could no longer be sustained on the proposition that Article 13 itself rendered them wholly immune from constitutional amendment. But the Court was unwilling to take the next, and more consequential step that because Parliament could amend every provision, it could therefore do anything to the Constitution so long as it followed the procedure prescribed by Article 368. 

The power to amend, the majority reasoned, could not be understood as a power to destroy the Constitution in the process of amending it. It distinguished between the ‘scope’ of the amending power and the ‘limits’ of what may be done in its exercise. Parliament could alter the Constitution, including its most important provisions, but there were certain features of the constitutional order whose destruction would cease to be an exercise of amendment at all. 

The question had thus changed from “Is a constitutional amendment law?” to the question, “What does it mean to amend a Constitution?” And it was in answering the latter that the Court arrived at -- what we call today -- the ‘basic structure’ doctrine.

***Finally, here are some points I wasn't able to cover, but which are, very much in the syllabus -- so here goes:

With regard to procedure, the Constituent Assembly had two rough models before it. One was the 1922 Free State Irish Constitution which allowed amendment by ordinary legislative majority for an initial eight years, meant to fix drafting defects; except that same ordinary-legislation power was used to extend the window itself, pushing it to sixteen years in 1929, and no amendment of that Constitution was ever put to a referendum and the safeguard was itself simply amended away. The 1937 Constitution course corrected hard and Article 46 of the Irish Constitution now makes a referendum mandatory for every amendment, with only a narrow 3 year transitional exception.

On the other hand stood the American Constitution, whose Article V procedure is close to virtually impossible in practice -- ⅔ of both Houses of Congress, then ratification by ¾ of the state legislatures. The result is 27 amendments in over 200 years, 10 of which arrived as a single package (the Bill of Rights) right at the start. The path the framers were treading therefore sat somewhere between an Irish regime that had shown how easily a safeguard could dissolve itself and an American regime that had shown how a safeguard could freeze a Constitution; Article 368's three tier structure -- ordinary majority, special majority, special majority plus state ratification -- was the balance struck between the two.

On the necessity of amendment as such: Ambedkar himself described the Indian Constitution as, before anything else, a social document -- an observation true of most post-colonial constitutions, which could never content themselves with "liberté, égalité, fraternité" alone, but had to write in the socio-economic transformation the anti-colonial movement had actually promised. Hence the DPSPs, hence the whole "transformative constitutionalism" discourse. Which, finally, brings me back to why that power needs limits in the first place: Weimar Republic had an amending procedure too -- a fairly workable one -- and it got Hitler amended straight into unchecked power, generally regarded as a suboptimal outcome for constitutional democracy. 

Thus, if you think about it, necessity and danger are really two sides of the same coin. A Constitution needs to change or it dies slowly, and it needs limits on that power or it dies quickly. 

History has shown us both these endings.

Sakal Papers v. Union of India: Keeping speech free – and whatever that means

Sakal Papers v. Union of India (1962) is something of a classic in Indian free speech law. It has been widely discussed in the literature on Article 19(1)(a), and may even be found on occasions in books and discussions over Part III more generally. A lecture clip by Nariman J. recently sent me back to the case. And after a rather long spell away from the blog, it felt like as good a case as any, to return and do what I do here from time to time.

An understanding of the case, needless to say, is important. For now, however, three things in view suffice:

1. The Nehru Government, through the Newspaper (Price and Page) Act 1956 and the Daily Newspapers (Price and Page) Order 1960 – to prevent unfair competition and, ostensibly, in keeping with its Socialist commitments – set about regulating the number of pages a newspaper could publish by reference to its price. 

2. Sakal Papers – a Marathi daily based out of Poona – challenged the scheme. Interestingly, the Union of India defended it on counter intuitive grounds that the regulations would actually promote – rather than restrict – free speech. They argued that the regulations would leave smaller newspapers more room to survive and compete, and preserve a plural press (larger newspapers – after all – could produce at a lower cost, keep prices low, and still be in a position to indefinitely expand circulation).

3. Speaking for the Court, Mudholkar J. held that if a newspaper must raise its price to retain its existing volume, such a measure necessarily bears upon its circulation, and therefore upon its capacity to communicate. As for the Article 19(2) question: given it contained no general “public interest” exception – unlike Article 19(6) – the State’s justification found no home. 

Let us start with the Court’s syllogism: regulation reduces volume; volume is speech; therefore, regulation reduces speech. 

Nothing in this chain is particularly bad or troublesome, until one notices what it holds as a constant – the number of speakers. For then the question becomes: what the regulation does to the output of each individual newspaper? But that – if one thinks about it – is precisely the point that the State was asking the Court to consider: the market itself produced an asymmetry between speakers and regulating the larger papers might allow room for the smaller ones.

The Court's application limitedly focuses on the newspaper's individual burden – i.e. a standard frame for cases of direct censorship – to what was essentially a market design claim. The two are fundamentally different kinds of free speech problems, and therefore should have called for a different kind of enquiry. The latter enquiry concerns itself with the structure of the marketplace, and whether it produces more or less of ideas that the right to free speech is meant to protect. [A good example is the approach undertaken by the SCOTUS in Associated Press v. United States (1945 decison). It found that the Associated Press – a cooperative news gathering organisation – through its membership rules was effectively excluding competing newspapers from its service and thereby violating US antitrust regulations. What is worth noting is that the same was done by treating the problem as one of access to a channel of communication).]

For a market design claim such as the one we see in Sakal Papers should similarly have looked at whether the regulation preserved or diminished the plurality of voices available in the market? and whether limiting the advantages of larger newspapers actually created space for smaller ones, and increased the overall supply and diversity of information and viewpoints? The Court, however, found no merit in the argument and there was no engagement to this end at all.

The reason I push for this so much is ultimately a fairly simple one. And t0 be clear, I do agree that more volume is more speech. But if volume is a relevant factor to the value of free speech under the Constitution, then so too – I would submit – is entry into the marketplace of ideas. More newspapers, of course –– mean more speakers –– meaning more viewpoints –– and potentially a more plural public sphere (also if one were to choose between a speaker who gets to speak considerably more and a larger number of speakers saying somewhat less, one might well prefer the latter).

Admittedly, all of this is a personal choice. It is, however, also, I would submit: a reasonably good instinct for free speech and constitutional law. 

On the Question of Citizenship : A Critique of ADR v ECI

--Krishaan Doctor

This blog is an attempt to add to the ongoing discussion of the Supreme Court’s decision in Association for Democratic Reforms v Election Commission of India (“ADR v ECI”). The focus of this piece is the fourth issue framed by the Court, namely 

Whether, in the exercise of its constitutional mandate of preparation and maintenance of electoral rolls, and in furtherance of the statutory conditions governing such registration, the Election Commission of India is empowered to scrutinise the citizenship status of persons seeking inclusion or continuation in the electoral roll?

Put simply, the question before the Court was whether the Election Commission (‘ECI’)’s power to prepare the electoral rolls under Art 324 empowers it to specifically scrutinise the citizenship status of persons seeking inclusion on the rolls.The petitioners argued that the ECI possessed no such power, and that Parliament had intended that questions of citizenship fall exclusively within the remit of the Central Government under the Citizenship Act,1955 (‘CA’). The ECI, by contrast, contended that because citizenship is a condition precedent to registration on the electoral roll under Section 16(1)(a) of the Representation of People Act, 1951 (‘RP Act’), it must necessarily possess a corresponding power to satisfy itself that those included on the rolls are citizens.

The Court ruled in favour of the ECI, holding that it is empowered to undertake a “limited enquiry” into citizenship only for the purpose of satisfying itself as to an elector’s eligibility for inclusion in the electoral roll. The reasoning employed by the Court to arrive at this conclusion rested on two prongs. First, Articles 325 and 326 of the Constitution, read alongside Section 16(1)(a) of the RP Act, envisage a citizen-only electorate. Therefore, for the ECI to discharge its obligation under Article 324, it must satisfy itself that those entitled to registration are citizens. Moving on, the real question, according to the Court, was with respect to the nature and extent of this satisfaction. While agreeing with the petitioners that formal adjudication of citizenship status must take place exclusively under the Citizenship Act, the Court drew a distinction between such adjudication and what it termed "administrative satisfaction as to eligibility for enrolment". The former entails a conclusive determination of citizenship under the Citizenship Act; the latter, according to the Court, is merely a limited inquiry undertaken by the Commission for the purposes of “electoral registration” i.e determining whether an individual has satisfactorily established their citizenship for inclusion on the electoral roll.

A previous post  by Rajyavardhan argued that this distinction was, “on its own terms, not obviously indefensible”. However, it demonstrated that the distinction ultimately collapses under the weight of the judgment's own referral mechanism: if a finding of non-satisfaction triggers citizenship adjudication before the competent authority, citizenship has already been placed in issue. This post makes a more ambitious claim. It argues that the power to examine questions of citizenship (including for “limited purposes of electoral eligibility”) cannot be derived by implication from the Commission's obligation to maintain valid electoral rolls under Article 324, read alongside Section 16(1)(a) of the RP Act.

Does the ECI have the power to examine questions of citizenship?

 The Court's reasoning rests on the assumption that because Section 16(1)(a) disqualifies non-citizens from registration, the Commission must possess a corresponding power to determine whether an individual is a citizen. However, I submit that Court’s reasoning on the intersection of s16 of the RP Act and Art 324 is premised on a very fundamental, but obvious conflation, and such a conclusion does not follow. Instead, I argue that the ECI can only act upon a determination of citizenship made under the Citizenship Act by the competent statutory authority. 

Section 16 of the RP Act reads as follows:

1. A person shall be disqualified for registration in an electoral roll if he (a) is not a citizen of India; or (b) is of unsound mind and stands so declared by a competent court; or (c) is for the time being disqualified from voting under any law relating to corrupt practices and other election offences.

2. The name of any person who becomes so disqualified after registration shall forthwith be struck off the electoral roll in which it is included.
   

It is clear, at least for grounds (b) and (c), that the power to make a determination as to “unsoundness of mind” and conviction for certain “election offences” resides with an external adjudicatory body i.e. a Competent Court. The ECI may only proceed to strike   those names off the electoral roll under 16(2) once the Court has made the relevant determination. It would be inconceivable to suggest that the ECI has the power to conduct a limited inquiry into the mental capacity of a prospective voter under s16(1)(b), arrive at an “administrative satisfaction”,that an elector is of unsound mind, delete the elector from the rolls, and thereafter refer the matter to a competent court. Neither does the statute envisage the split status of “unsoundness of mind” that may follow, where an individual could be treated as of unsound mind for the limited purposes of voting while simultaneously remaining of sound mind for the purposes of, say, entering into contracts. In sum, the point I am trying to make is that ECI's responsibility under Article 324 to maintain a valid electoral roll does not entail the power to make a determination of unsoundness of mind simpliciter, but to act only after a Court - which is a specialised body with the expertise to take such a call - determines a particular voter is of unsound mind. 

I believe that by analogy and for several other compelling reasons, the ECI’s power to maintain a citizen only roll does not extend to determining citizenship for the limited purpose of “electoral eligibility". Parliament has enacted a complete code governing the determination of citizenship status by way of the Citizenship Act, 1955. The Act, read as a whole (see Sections 9(2), 13 and 14), contemplates citizenship questions being resolved by authorities operating under the supervision of the Central Government. The ECI is conspicuously absent from this scheme, and neither does any other statute lay down the procedures it may follow while determining citizenship for so-called “electoral purposes”. The "split status" of citizenship created by the Court as a result of this decision, whereby one is adjudicated to be a non-citizen for electoral purposes but a citizen for all other purposes, finds no basis in either the Citizenship Act or the Constitution.

One might expect that, had Parliament intended to confer upon the ECI a concurrent power to determine citizenship, even for the “limited purpose” of electoral registration, it would have done so expressly. In my view, this is  for two key reasons. First, the power to determine citizenship, even for “electoral purposes” is highly consequential. As the Court itself notes, citizenship is a “status that situates a person within the political community and enables participation in the democratic process”(para 168). The franchise is the principal means through which that membership in the political community is recognised and exercised. It would be reasonable to assume that Parliament did not intend confer such a significant power to the ECI with no statutory guidance as to how to go about the same. Second, unlike the other administrative determinations the ECI may undertake during an SIR—such as recording deaths or migration—the determination of citizenship entails a quasi-judicial application of mind. The competent authority must assess conflicting documents, determine their authenticity and decide what weight to attach to each. Moreover, such an exercise would require heightened procedural safeguards and mechanisms for notice, hearing and appeal. Therefore, not only is "determination of citizenship for electoral purposes" by the ECI for a highly consequential power that should probably find mention in some statute, but also would necessarily require some legislatively prescribed standards dictating the entire decision - making process. 

The strongest argument in favour of the Court's conclusion is that the ECI's powers under Article 324 are plenary and, in many respects, unfettered.  However, this overlooks the well-established principle that the ECI’s powers under Article 324 operate only in areas where the law is silent and cannot be exercised in a manner inconsistent with or contrary to parliamentary legislation. In Mohinder Singh Gill, the Court held that although the Commission's powers are plenary in character, they yield where Parliament has enacted a valid law "relating to or in connection with elections" (para 92). Parliament has done precisely that through the Citizenship Act, which constitutes a comprehensive code governing the acquisition and determination of citizenship. Therefore, Article 324 cannot be invoked to confer upon the ECI a parallel power to determine questions of citizenship. 

A Brief Note on the Distinction Between “Administrative Satisfaction as to Electoral Eligibility" and “Adjudication of Citizenship”

The Court tries to reconcile the proposition that the Citizenship Act 1955 is a complete code for matters pertaining to citizenship with the powers of the ECI by drawing a distinction between the formal adjudication of citizenship and something it terms “administrative satisfaction as to electoral eligibility". I submit that this distinction is entirely illusory, and that it has the effect of transforming the Competent Authority under the Citizenship Act into a body with a quasi - appellate jurisdiction over questions of citizenship. 

First, a close look at the functioning of the Commission finds that it resembles the key elements of an adjudicatory process According to Lon Fuller, the key elements of “adjudication” are as follows (a) a party whose rights are affected (b) an opportunity to present evidence and argument (c) a decision-maker bound to consider them (d) a conclusion that determines those rights. In the Commission's ad hoc contextual inquiry, an elector whose citizenship is doubted must furnish the “relevant documents”, which is then considered by the ECI. The Commission reaches a conclusion on the basis of said materials, which determines the elector's  right to be on the roll. The four elements of adjudication are therefore met. Second, at para 178, the Court holds that where the Commission is not satisfied as to an individual's citizenship, the matter may ultimately be referred to the competent authority under the Citizenship Act. The primary role of the Competent Authority therefore becomes to revisit or review the decisions of the Election Commission. This is difficult to reconcile with the statutory scheme of the Citizenship Act, which contemplates the competent authority determining questions of citizenship in the first instance. 

Conclusion

This blog has aimed to argue that the ECI’s power to ensure a valid electoral roll under Article 324 does not extend to determining questions of citizenship for itself, but to acting on determinations made by the Competent Authority under the Citizenship Act. Moreover, the classification between administrative satisfaction and formal adjudication as to citizenship is illusory, and also has no basis in either the Constitution or the Citizenship Act.

Joining the Fold: Krishaan Doctor

Hi all,

I could not be happier to welcome the wonderful Krishaan Doctor as a contributor to the blog..

Many of you will already know Krishaan's writing from the Constitutional Law and Philosophy blog. He has written some exceptionally thoughtful pieces, including contributions to the Facilitating Digital Censorship, a response to Gautam Bhatia on whether the Supreme Court can ban a book, and an analysis of the split verdict on the constitutionality of Section 17A.

I have no doubt that readers will enjoy whatever he writes here; I certainly will. 

And for those wondering, you can tell whether a post is by me or by Krishaan by looking at the footer at the end of each post.

Thanks, and I'm really excited about this.

ADR v. ECI (the SIR Case) – Looking Back

The Supreme Court's (“Court”) decision in Association for Democratic Reforms v Election Commission of India (“ADR v ECI”) is unlikely to be remembered as a particularly happy moment in Indian constitutional law. Delivered in the aftermath of one of the most controversial electoral disputes in recent memory, the judgement resolves virtually every significant question in favour of the Election Commission of India (“ECI” or “Commission”). In so doing, it accorded the Commission an exceptionally broad measure of constitutional and statutory latitude, while displaying a rather sanguine faith in the ability of administrative safeguards to prevent wrongful disenfranchisement. The consequence is a judgement that is markedly more attentive to the problem of keeping allegedly ineligible voters out of the electoral roll than to the equally serious problem of keeping eligible voters in it. 

That imbalance runs through the decision from beginning to end.

The SIR and Its Challenge

For context, the litigation arose from the ECI's decision on 24 June 2025 to undertake a Special Intensive Revision ("SIR") of the electoral rolls in Bihar ahead of the impending Assembly elections. The Commission justified the exercise by reference to demographic change, migration, duplication, and concerns regarding the overall accuracy of the rolls. Electors not appearing in the 2003 electoral roll were required to submit enumeration forms accompanied by prescribed documentary proof, failing which their names would not be included in the draft roll. 

Naturally, what might otherwise have been presented as a usual exercise in the housekeeping of electoral rolls soon came to be viewed as something considerably more consequential.

 The challenge proceeded on three grounds broadly.

1. Power

First, the petitioners questioned the Commission's authority to undertake the exercise at all. For simply Article 324 could not be treated as an inexhaustible source of constitutional power once Parliament had ‘occupied the field’ (see Mohinder Singh Gill, AC Jose, Kuldip Nayyar) through the Representation of the People Act, 1950 (“RP Act”) and the Registration of Electors Rules, 1960 (“Registration Rules”). Nor, it was argued, could s 21(3) of the RP Act be stretched into a charter for a statewide reverification exercise of this scale. 

One of  the first posts of this blog in fact was precisely about this, I had argued:

“...Section 21(3) is an exception to the general rule under Section 21(2), which contemplates annual revisions of the electoral roll. Seen in this light, the statutory scheme becomes relatively clear — for it indicates that routine revisions are to form the foundation of electoral roll maintenance, while special revisions are meant to step in only when routine mechanisms fall short… 

The Commission appears to be stretching the provision’s scope beyond what the Parliament had likely contemplated. When a statute authorises a “special revision,” it assumes that the normal revision has occurred or will occur in accordance with law, and that the special revision will operate as a limited corrective mechanism. The “reasons to be recorded” requirement perhaps even reinforces this logic. On such a reading, special revisions, in other words, would respond to identifiable contingencies, and not broad governance concerns… 

This becomes clearer through illustration. For instance, take “migration.” Internal migration in India, especially in states like Bihar, is a continuous phenomenon. Treating this as a trigger for a state-wide intensive revision has potential to turn what should be an exceptional measure into a routine one. Similarly, while underreporting of deaths is a long-standing challenge, it has never, until now, been seen as grounds for a statewide electoral overhaul.”

2. Procedure

The petitioners that under the statutory scheme, inclusion in the electoral roll carried with it a presumption of validity that could only be displaced through procedures specifically prescribed by law (see s 16 of the RP Act). The SIR, however, effectively required millions of already enrolled electors to re-establish their entitlement to remain on the roll. In practical terms, the burden shifted from the State demonstrating ineligibility to the elector demonstrating continued eligibility. This concern was amplified by the fact that non-submission of an enumeration form itself resulted in exclusion from the draft roll. The petitioners further argued that the exercise bypassed the safeguards embedded in r 21A of the Registration Rules which contemplates notice, publication, objections and hearing before deletion. They also criticised the absence of clear standards governing scrutiny by Booth Level Officers and Electoral Registration Officers, warning that broad administrative discretion, exercised under compressed timelines, created an appreciable risk of arbitrary exclusion. 

The objection, therefore, to put it simply: was (1) a burdensome procedure, that (2) fundamentally inverted the logic of the electoral registration process.

3. Citizenship

The petitioners contended that one of the evident purposes of the SIR was to facilitate a largescale scrutiny of citizenship status under the guise of electoral roll revision. Such an exercise, they argued, exceeded the Commission's constitutional mandate. They argued that while the ECI undoubtedly may possess the authority to prepare and revise electoral rolls, but the determination of citizenship is governed by an entirely distinct statutory framework principally administered vide Citizenship Act of 1955. The petitioners also demonstrated how an elector called upon to establish citizenship in order to remain on the roll faces consequences that closely resemble a citizenship determination, even if formally described otherwise. 

The SIR was therefore said to confer upon electoral officials a function which they were not competent to undertake (i.e. attacking competence), while at the same time exposing large numbers of voters to citizenship scrutiny through an administrative process lacking the safeguards ordinarily associated with such determinations (i.e. attacking design).

The Litigation’s Changed Shape 

The original challenge was directed at the legality of the SIR: i.e. whether the ECI possessed the authority to undertake the SIR, i.e. whether the procedure it adopted was consistent with the statutory scheme, i.e. whether the exercise impermissibly introduced citizenship scrutiny into the electoral registration process, and so on. 

Ordinarily, one might have expected the Court to engage these questions through the familiar lens of legality review. Which is to say by asking whether the impugned notification was lawful on the date it was issued. Instead, the Court adopted what commentators have described as a form of dialogic review. Rather than suspending the exercise or testing its legality in isolation, the Court entered into an ongoing institutional dialogue with the Commission, identifying perceived deficiencies and encouraging their subsequent correction. 

As the litigation progressed, therefore, the SIR scheme that ultimately came before the Court for adjudication was no longer the SIR scheme that had originally been challenged (see “The Story of Five Improvisations” below).

The Story of Five Improvisations

1. On 10 July, while issuing notice, the Court identified three principal concerns: (i) the Commission's authority to undertake the exercise, (ii) the procedure adopted for verifying eligibility and citizenship, and (iii) the timing of the revision immediately preceding the Bihar Assembly elections. Significantly, however, the Court declined to stay the exercise. Instead, it suggested that the Commission consider accepting Aadhaar Cards, Electors Photo Identity Cards (EPICs), and Ration Cards in addition to the eleven documents originally prescribed. 

2. The shape of the litigation altered further after publication of the draft electoral roll on 1 August 2025, which excluded approximately sixty-five lakh electors. On 14 August, the Court directed the ECI to publish the names of excluded electors together with reasons for exclusion and to ensure wide dissemination through newspapers, electronic media, and radio. Excluded electors were also permitted to submit claims supported by Aadhaar Cards. 

3. On 22 August, the Court went further still, and impleaded twelve major political parties to mobilise Booth Level Agents to assist electors with the submission of forms and supporting documents. 

4. Thereafter, on 1 September 2025, the Commission informed the Court that claims and objections would continue to be entertained beyond the original deadline and up to the last date of nominations. On the same occasion, the Court enlisted the Bihar State Legal Services Authority and district legal aid institutions to assist voters in filing claims, objections, and corrections. 

5. Finally, on 8 September 2025, the Court formally directed that Aadhaar Cards be accepted as an identity document for the purposes of the SIR, while clarifying that Aadhaar could not constitute proof of citizenship. 

(see the SC Observer’s day-wise coverage of SIR Hearings)

By this stage, the proceedings had assumed a distinctly dialogic character. The Court had become an active participant in reshaping the terms on which that action would proceed. This matters because many of the procedural safeguards that later featured prominently in the Court's defence of the SIR were not contained in the original notification. They emerged during the course of the litigation itself. 

The Judgement’s Priors – A Curious Place 

Before reaching any doctrinal question, the Court spends some time constructing a narrative. The judgement in ADR v. ECI opens with the republican traditions of ancient Bihar, the Vajji confederacy, colonial electoral administration, and the Constituent Assembly's adoption of universal adult suffrage (¶¶ 1–6). The idea is that electoral rolls are instruments through which a political community determines its own membership.

As the Court puts it: 

"[b]efore any representative government can count votes, it must first know whose votes may be counted" (¶ 1).

The judgement repeatedly emphasises the need for an electoral roll that reflects the "true composition of the political community" (¶ 7). That formulation in many ways foregrounds concerns of electoral integrity while pushing concerns of wrongful exclusion into the background. It appears to assume that the principal threat is an inaccurate roll. Much of what follows can be read as an elaboration of that assumption and the Court’s historical introduction is a good lens through which the remainder of the case can be viewed.

The Four Questions

After disposing of the preliminary objections and recounting the course of the litigation, the Court distilled the controversy into four questions (¶15). 

The first question was whether the ECI possessed the power to undertake the impugned SIR at all. On one level, this was the most basic issue before the Court. Yet beneath this seemingly “most basic” dispute lay a more fundamental disagreement about the nature of the Commission itself. Is the Commission primarily a statutory actor operating within a framework prescribed by Parliament? Or is it a constitutional institution endowed with an independent sphere of authority that survives legislative intervention? Because when Article 324 and parliamentary legislation appear to point in different directions, which one gives way? 

The second question concerned purpose and proportionality. Specifically, the Court asked whether the SIR was founded upon a legitimate objective and whether the measures adopted by the Commission were proportionate to that objective. Notice how the court did not ask whether the exercise risked disenfranchisement on a scale incompatible with constitutional democracy. Nor did it ask whether existing electors ought to be subjected to fresh documentary burdens in the first place. The premise underlying the inquiry was that electoral roll accuracy constituted a legitimate constitutional objective. The question therefore became whether the chosen means bore a sufficient relationship to that end.

There is, of course, nothing inherently objectionable about this formulation. But once the objective is accepted as legitimate, attention naturally turns towards the adequacy of safeguards rather than the propriety of the exercise itself. This is important because much of the criticism directed against the SIR that the exercise reversed settled assumptions concerning electoral registration and placed burdens upon electors that ought not to have existed at all.

The third question addressed whether the procedure adopted by the Commission was contrary to the statutory scheme in place. This issue occupied a rather funny position within the litigation. On one hand, it was arguably the petitioners' strongest argument. One need not have adopted a grand theory of constitutional law to understand it. The question was whether the ECI had followed the procedures laid down by law. On the other hand, the challenge was progressively diluted as the litigation evolved (see The Story of Five Improvisations above). 

The Court was therefore confronted with the legality of an exercise that had been substantially modified during the course of proceedings.

The fourth and final question concerned citizenship. Here, the Court framed the issue as whether the Commission, in carrying out its constitutional mandate of preparing and maintaining electoral rolls, could scrutinise the citizenship status of persons seeking inclusion or continuation on those rolls. 

I am reproducing the core argument. Readers are free – and encouraged, more so – to refer to the piece in full: 

“The Court presents its holding as striking a “constitutional balance” between two alternatives that it understands as mutually exclusive: (1) either the ECI enforces the citizenship requirement, or (2) it adjudicates citizenship and thereby exceeds its statutory mandate. If that is so, a genuine balance would have required each of these competing concerns to operate as a meaningful constraint upon the other. For if the ECI’s enquiry were truly limited in both purpose and consequence to the electoral roll (as the “principled distinction” suggests), it is difficult to explain what the competent authority is being asked to adjudicate when the mandatory referral arrives. Conversely, if the referral mechanism were coherent, it could only be because citizenship had already been placed in issue, in which case the claim that the enquiry remains limited simply falls away.

What is thus presented as a “balance” between two distinct enquiries comes to resemble – as noted above – successive stages within a single process concerned with the determination of citizenship. And a distinction which dissolves itself into a single enquiry upon its first application – in any meaningful sense – is not a balance at all.”

Concluding Remarks

Viewed individually, each of these four questions appear discrete. One concerns constitutional authority. Another concerns proportionality. A third concerns statutory compliance. The last concerns citizenship. Yet these questions are deeply interconnected and not remotely as self contained. Each answer is based upon a prior assumption about the nature of electoral democracy, the role of the ECI, and the relationship between electoral integrity and electoral inclusion. 

In that sense, the judgement in ADR v ECI  seems to be – in essence – a product of reasoning – the central imperative of which – is an “accurate roll.” The rest of the questions answer themselves on their own. Whether constitutional courts ought to assume that role, particularly where the complaint is that the exercise itself is unlawful, is likely to remain the more enduring question the SIR case leaves behind.