On 15 February 2024, a Constitution Bench of the Supreme Court of India held that a voter cannot be asked to choose a government in ignorance of who pays for the parties asking for that vote. The electoral bonds judgment, delivered in Association for Democratic Reforms v Union of India and reported at (2024) 5 SCC 1, struck down the Electoral Bond Scheme of 2018 and the statutory amendments that had been built around it, and it did so on two distinct constitutional grounds that are routinely merged in summary accounts. The disclosure provisions fell under the freedom of speech and expression, because the Court held that information about the funding of political parties is essential to the effective exercise of the franchise. The removal of the ceiling on corporate contributions fell under the guarantee of equality, because the Court held that treating a company’s political contribution as the equivalent of an individual’s is manifestly arbitrary. The distinction matters beyond this scheme. A legislature that designs a replacement funding mechanism with complete transparency could still fail the second holding, because the equality objection is not about secrecy at all. It is about the legal character of corporate money in an election.

The interpretive question the Court had to answer was not written anywhere in the constitutional text. Article 19(1)(a) guarantees the freedom of speech and expression. It says nothing about political finance, nothing about donations, and nothing about a right to information. The petitioners’ case required the Court to travel from that spare guarantee to a conclusion that a scheme of anonymous donations, enacted through amendments to three statutes and notified by the executive, was unconstitutional. The Union of India’s case required the Court to stop that journey early: the right to information, it submitted, extended to the antecedents of candidates, which earlier decisions had established, and no further. Between those two positions lay the whole case, and the Court’s resolution of it is the reason this judgment will be read long after the bonds themselves are forgotten. This article states the holdings separately, as the Court delivered them, reconstructs the reasoning of the majority opinion and of the concurrence, narrates the disclosure sequence that followed the judgment as part of its outcome, and ends with the questions the Court expressly did not decide, the largest of which concerns the legislative vehicle by which the entire arrangement had been enacted.

When was the electoral bonds judgment delivered, and who decided it?

The judgment was delivered on 15 February 2024 by a Constitution Bench of five: Chief Justice D. Y. Chandrachud and Justices Sanjiv Khanna, B. R. Gavai, J. B. Pardiwala and Manoj Misra. The decision was unanimous. The Chief Justice wrote for himself and three colleagues; Justice Khanna wrote separately, agreeing in the result on partly different reasoning.

The unanimity is worth pausing on, because it shaped the reception of the case. A five-judge bench speaking without dissent on a question that divided political opinion left no minority text around which a later bench might reorganise the law. There were two opinions, and the difference between them is a difference of route, not of destination. The lead opinion of Chandrachud CJ struck down the scheme and three sets of amendments and issued the operative directions. The concurrence of Khanna J reached the same conclusions by applying the doctrine of proportionality in a form that, on one point, went further than the lead opinion, as the discussion of the Reserve Bank amendment below records. Any account that treats the judgment as a single undifferentiated text misses the one place where the two opinions diverge in what they strike down, and it misses the doctrinal interest of the case, which lies in how the standard of review was constructed.

The scheme in outline

The electoral bond scheme itself is the subject of its own article in this series, the account of the electoral bonds law, and only its outline is needed here. The scheme was notified on 2 January 2018 under the authority of amendments made by the Finance Act, 2017. A bond was a bearer instrument in the nature of a promissory note, purchasable in fixed denominations from designated branches of the State Bank of India by a citizen of India or a body incorporated in India, and payable to an eligible political party, meaning a registered party that had secured at least one per cent of the votes polled in the preceding general election to the Lok Sabha or a state legislative assembly. The bond carried no name. Its validity ran for fifteen days from issue, within which the receiving party had to deposit it in its designated account. The purchaser’s identity was recorded by the bank for its know-your-customer requirements, but the scheme’s confidentiality clause, Clause 7(4), barred the bank from disclosing the purchaser’s particulars to anyone except on the direction of a competent court or on the registration of a criminal case by a law enforcement agency.

Three statutory amendments made the instrument work. The Finance Act, 2017 amended Section 29C of the Representation of the People Act, 1951 to exempt contributions received through electoral bonds from the contribution report that parties must file for donations above the statutory threshold. It amended Section 13A of the Income Tax Act, 1961 to relieve parties of the record-keeping condition, the maintenance of names and addresses of contributors, in respect of bond receipts, as a condition of their income tax exemption. And it amended Section 182 of the Companies Act, 2013 in two ways: it deleted the first proviso that had capped a company’s political contributions at seven and a half per cent of its average net profits during the three immediately preceding financial years, and it altered sub-section (3) so that a company was required to disclose in its profit and loss account only the total amount contributed, and no longer the name of the party to which each contribution was made. A fourth amendment, to Section 31 of the Reserve Bank of India Act, 1934, equipped the issuing mechanism by permitting the bonds to be issued as bearer instruments. The challenge attacked the scheme and each of these amendments, and the Court’s holdings track that structure: four things fell, and they fell on two different grounds.

How the scheme entered the statute book, and why the vehicle mattered

The legislative history of the scheme is brief, and its brevity became part of the case. The electoral bond was announced in the Union Budget speech of 1 February 2017, as the centrepiece of a package presented as a reform of political funding: donations would move from cash to banking channels, the donor’s identity would be shielded to protect the donor from retribution, and the scheme to operate the bonds would be notified under the Reserve Bank of India Act. The Finance Bill that followed carried the necessary amendments across the four statutes described earlier in this article. Because the Bill was certified a Money Bill, the legislative process available for those amendments was the truncated one the Constitution prescribes for financial legislation: the Rajya Sabha could debate and recommend, but it could not amend or reject, and when the upper House returned the Bill with recommendations, the Lok Sabha was free to disregard them, as it did. The Finance Act received the President’s assent on 31 March 2017 and the amendments took effect. The scheme itself, a creature of executive notification rather than primary legislation, followed on 2 January 2018, and the first bonds were sold in March 2018.

Every feature of that history generated a strand of the challenge. The substance generated the rights claims examined in this article. The vehicle generated a distinct claim, that electoral reform of this magnitude, amending the Representation of the People Act and the law governing companies and the central bank, was not the kind of matter Article 110 reserves to the Money Bill route, and that the certification had deprived the Rajya Sabha of its constitutional role. The claim drew its strength from the text of Article 110, which confines a Money Bill to provisions dealing with the matters the Article enumerates, taxation, borrowing, the Consolidated Fund and matters incidental to them, and from the Aadhaar litigation, in which a majority of a Constitution Bench had upheld the certification of the Aadhaar Act as a Money Bill against a powerful dissent. It drew its weakness, as events proved, from the state of the docket. In Rojer Mathew v South Indian Bank, a Constitution Bench considering tribunal reforms carried in a later Finance Act had doubted the Aadhaar majority’s reading of Article 110 and referred the scope of the Article to a Bench of seven judges. That reference was pending when the bonds case was heard, and it converted the vehicle claim into a question the bonds Bench was institutionally unable to answer without pre-empting a larger Bench seized of the identical issue.

The consequence is a feature of the judgment that commentary often understates. The bonds Bench’s silence on the Money Bill question is not an endorsement of the vehicle, and it is not a rejection of the challenge to it. It is a postponement compelled by the Court’s own discipline of Bench strength, under which a question referred to seven judges is not decided by five. The practical effect is a two-stage constitutional history. The rights stage is complete: the amendments are void for violating Articles 19(1)(a) and 14, and no future bond scheme in this form can issue. The vehicle stage is outstanding: whether the amendments could have been enacted by that route at all remains open, and the answer, when the seven-judge Bench gives it, will govern a family of Finance Act legislation far beyond electoral bonds, from tribunal appointments to the funding architecture of regulators. For the reader of this article, the important discipline is to keep the two stages separate. The judgment is authority for the rights propositions it decided and for nothing on the vehicle, and any account that treats the striking down as an implicit ruling on Article 110 misreads the record the opinions themselves set out, in which both the lead opinion and the concurrence record the reference and the consequent restraint in terms.

There is a final feature of the entry history that the judgment turned to analytical account. Because the scheme was a notified scheme under an amended statute rather than a chapter of an Act, its operative secrecy lived in a subordinate instrument, clause 7(4), whose validity depended entirely on the provisions above it. The architecture gave the Court its points of attack in ascending order: the clause fell with the scheme, the scheme fell with the amendments that exempted its transactions from the disclosure code, and the amendments fell on the rights analysis. Nothing in the structure required the Court to lay a glove on the Finance Act as a whole, which is one reason the judgment, for all its consequence, is surgical in form: it invalidates what it identifies, provision by provision, and leaves the rest of the 2017 Act standing.

The road to the Constitution Bench

The lead petition, Writ Petition (Civil) No. 880 of 2017, was filed by the Association for Democratic Reforms before the scheme had even been notified, attacking the Finance Act amendments in anticipation of the instrument they would enable. Connected petitions followed, including those of Common Cause and the Communist Party of India (Marxist). The Election Commission of India, a respondent, did not defend the anonymity. In its submissions during the proceedings the Commission supported disclosure, treating anonymous funding of parties as a retrograde step for the transparency of the electoral process, a position that left the Union of India effectively alone in defending the design it had created.

What did the interim order of 12 April 2019 require while the challenge was pending?

By its interim order of 12 April 2019, the Court declined to stay the scheme but refused to let the money move unrecorded. It directed political parties to submit to the Election Commission, in sealed cover, particulars of contributions received through electoral bonds, including the identity of donors and the amounts, for bonds received up to that point and thereafter.

That interim arrangement governed the field for nearly five years, and it explains a date that runs through everything that followed. When the final judgment ordered disclosure, it ordered disclosure from 12 April 2019, the date of the interim order, because that was the date from which parties and the bank had been on express notice that particulars were being collected under the Court’s authority and might be called for. Donations made under the scheme in its first year, from the first issue in March 2018 until the interim order, sat outside the disclosure window the judgment drew. The choice was deliberate, and when the Court was later invited to widen the window backwards it declined, treating the cut-off as a considered feature of the judgment rather than an oversight to be corrected on a miscellaneous application.

The petitions were eventually placed before a Constitution Bench because they raised substantial questions of constitutional interpretation. The Bench heard argument over three days, from 31 October to 2 November 2023, and reserved judgment. The arguments it heard frame the analysis that follows. The petitioners submitted that the scheme legalised opacity in the single most consequential flow of money in a democracy, that it privileged the ruling party of the day because the issuing bank was state-owned and the government could in principle access what the voter could not, and that the deletion of the corporate ceiling destroyed the distinction between a citizen’s political gift and a company’s commercial outlay. The Union submitted that the scheme was an improvement on a prior regime of unaccounted cash, that its safeguards of know-your-customer compliance, short validity and party eligibility were real, that donor confidentiality protected contributors against retribution and victimisation by parties they had not supported, and that curbing black money in political finance was a legitimate and weighty objective. The Court’s task was to convert those submissions into doctrine, and it began by deciding what kind of law it was reviewing.

The record before the Bench

A constitutional judgment is only as intelligible as the record it answers, and the record in this case had features that shaped the reasoning at every stage. The first was documentary. The Union’s defence of the scheme rested on the purposes it attributed to the design, and those purposes were not merely asserted in argument; they appeared in the correspondence between the Ministry of Finance and the Reserve Bank of India that preceded the scheme’s notification, in which the bonds were described as an instrument of electoral reform directed at curbing black money in the electoral process and at protecting the informational privacy of contributors. The petitioners used that correspondence forensically. If the measure’s own architects described it as electoral reform, the Union could hardly ask the Court to review it with the deference owed to fiscal policy, and the correspondence fixed the purposes against which proportionality would later be measured: the Court would not test the scheme against purposes invented for the litigation, but against the two purposes its designers had recorded for it.

The second feature was the position of the Election Commission. The Commission is the constitutional body charged with the superintendence of elections, and a court testing an electoral measure ordinarily gives weight to the regulator’s assessment of how the measure works in practice. Here the regulator declined to support the measure. In the course of the proceedings the Commission placed its position on record in favour of disclosure of political funding and against the anonymity the scheme created, describing anonymous funding as a retrograde step for transparency in the electoral process. That alignment mattered to the atmosphere of the case and to one strand of its logic. The Union’s claim that anonymity was a considered electoral reform had to be maintained against the stated view of the very institution that administers elections, which regarded the reform as a regression from the disclosure regime it had operated for decades.

The third feature was the five-year life of the interim regime. Between the interim order of April 2019 and the final hearing in the autumn of 2023, the scheme operated at full scale. Bonds were sold in successive windows, purchased, transferred and encashed, while the underlying challenge awaited a Constitution Bench. The data generated in those years accumulated in two places: in the sealed covers that political parties filed with the Election Commission under the interim order, and in the records of the State Bank of India, which alone knew both ends of every transaction and was organised, as its later application revealed, to keep the two ends apart. The passage of time thus changed the remedial question. In 2017 the Court could have been asked simply to stop a scheme before it began. By 2024 it was being asked to pronounce on a completed financial history running into thousands of crores of rupees, and any answer it gave would have to decide what to do about the record of that history. The judgment’s disclosure directions, analysed later in this article, are unintelligible except against this feature of the record: the Court was not only invalidating a law, it was deciding the fate of an archive.

The arguments, when the Bench finally heard them, were structured by these features. The petitioners’ case, in its strongest form, had three limbs. The informational limb submitted that the right to know recognised in the candidate cases extends in logic and in principle to party funding, because the party is the effective object of the vote. The equality limb submitted that unlimited corporate contributions, freed of the profit-linked ceiling, destroyed the premise of political equality on which the franchise rests and created a lawful channel for quid pro quo government. The structural limb submitted that the anonymity was in any event fraudulent in design, because it concealed the donor from the voter and the opposition while leaving the donor visible to the State that owned the issuing bank. The Union’s case, as Khanna J summarised it, also had three limbs: that donors apprehend retribution and victimisation from parties they do not fund, so that confidentiality protects a genuine privacy interest and indeed the donor’s own freedom of expression; that the scheme, by incentivising banking channels, checks the use of black or unaccounted money in political contributions; and that the scheme was an improvement on the prior framework, with inbuilt safeguards in know-your-customer compliance, a fifteen-day validity that prevented the bonds circulating as currency, and an eligibility threshold confining receipt to established parties. Every element of the judgment that follows answers one or more of these six limbs, and the reader who holds the six in view can see the opinion’s architecture whole: the information holding answers the first and third Union limbs, the proportionality analysis answers the privacy limb, the black money discussion answers the second, and the equality holding answers the petitioners’ second limb in their favour on a ground the Union’s improvement argument never engaged.

The standard of review: electoral reform, not economic policy

Before reaching the rights, the Court settled the posture of the review, and the settlement mattered to everything else. The Union’s submissions invited a deferential standard: legislation touching finance and the economy, it was said, attracts latitude, and courts do not sit in appeal over the wisdom of fiscal design. The lead opinion rejected the characterisation. The amendments under challenge did not deal with economic policy in the sense of trade, business and commerce. They dealt with the electoral process. The Union had itself described the package, in the correspondence between the Ministry of Finance and the Reserve Bank that the record contained, as electoral reform introduced to curb black money in the electoral process and to protect the informational privacy of contributors. A measure whose stated subject is the conduct of elections is to be tested as electoral law, and the importance of the nexus between money and electoral democracy required an in-depth review, albeit under the settled powers of judicial review. Khanna J agreed, adding that to give such legislation the latitude of economic policy would dilute the principle of free and fair elections.

The lead opinion then addressed the presumption of constitutionality, on which the Union had also leaned. The presumption, the Court explained, rests on democratic accountability: legislators are elected representatives aware of the needs of citizens and privy to information a court does not have. But the presumption does not immunise a measure against Part III. Once the petitioners established a prima facie case of violation of a fundamental right, the onus shifted to the State to justify the restriction. That shifting of the burden structured the rest of the judgment. For each impugned measure, the question became not whether the petitioners could prove harm in the abstract, but whether the Union could demonstrate that the measure pursued a legitimate purpose by means that were suitable, necessary and proportionate. On the central question of anonymity, the Union failed that demonstration, and the manner of its failure is the doctrinal core of the case.

The first holding: information as a condition of the vote

The lead opinion began from authority it treated as settled. In Union of India v Association for Democratic Reforms, decided in 2002, the Court had held that a voter has a right to know the antecedents of a candidate, including criminal history, assets and liabilities, and educational qualifications, because the casting of a vote is itself a form of expression and an uninformed vote is a diminished one. In People’s Union for Civil Liberties v Union of India, decided in 2003, the Court struck down a statutory attempt to dilute that disclosure regime and confirmed that the right to information about candidates flows from Article 19(1)(a). The Union’s submission in the bonds case was that this line of authority stops at the candidate. The funding of the party, it argued, is a different subject, and no decision had extended the voter’s informational right to it.

The Court refused to stop the line there, and its reason was institutional rather than rhetorical. Political parties are not incidental to the electoral process; they are at its helm. Parties select the candidates, allot the symbols on which candidates contest, finance their campaigns and, through the anti-defection discipline of the Tenth Schedule, effectively own the legislative behaviour of the members elected under their banner. A voter does not, in the ordinary case, vote for an individual in isolation. The voter votes for the candidate of a party, in the expectation that the party will come to power and act on its programme. To hold that the voter may know everything about the candidate’s antecedents but nothing about the sources of the money that sustains the candidate’s party would produce a dichotomy the Court declined to accept: the funding of political parties cannot be treated differently from the funding of the candidates who contest in their name. Information about the funding of a political party, the lead opinion held, is essential for a voter to exercise the freedom to vote in an effective manner.

The opinion gave the holding a second, harder edge. Information about who funds a party is not merely useful background for an interested citizen. It is the material from which a voter, a journalist or an opposing party can identify the possibility of quid pro quo arrangements between a contributor and a government formed by the recipient party: a policy changed, a contract awarded, a prosecution softened. A donation made in expectation of such a return is not an act of political expression by the donor alone; it is a transaction with the democratic process as its subject matter, and the process has a right to be visible to those whose votes constitute it. That is why the Court treated the deletion of party-wise disclosure in Section 182(3) of the Companies Act with particular seriousness. Even where the aggregate flow of corporate money might be known, the voter who cannot learn which party received a company’s contribution cannot connect the contribution to the conduct of the government that party forms, and the informational right is defeated at exactly the point where it bites.

The holding, stated precisely, is this. The Electoral Bond Scheme, and the proviso to Section 29C(1) of the Representation of the People Act, Section 182(3) of the Companies Act and the proviso to Section 13A of the Income Tax Act to the extent that they anonymise contributions through electoral bonds, violate the right to information of the voter under Article 19(1)(a) and are unconstitutional. The Section 29C amendment and the Section 182(3) amendment had to be read together, the Court held, because the purpose of amending the Companies Act provision was to bring it into line with the election statute’s exemption: once the scheme and the anonymising exemptions fell, the corporate disclosure amendment became otiose and fell with them. The conclusion was unanimous. Its reasoning, however, required the Court to answer the Union’s strongest point, which was not about information at all. It was about privacy.

Two rights, one standard: the double proportionality analysis

The Union’s defence rested in substantial part on a competing fundamental right. A contributor to a political party, it submitted, has a right to informational privacy: privacy over the fact and object of the contribution, protected by the right to privacy recognised in Justice K. S. Puttaswamy (Retd.) v Union of India and by the freedom of expression itself, since a donation is a mode of expressing political support. Disclosure exposes the donor to retribution and victimisation from parties and actors the donor chose not to fund. On that footing, the case was not a simple contest between a right and a policy. It was a contest between two rights, the voter’s right to information and the donor’s right to informational privacy, and each side could point to constitutional text and precedent.

The Court accepted that framing and then constructed the standard to govern it. Where a single fundamental right is restricted by the State for a purpose external to the right, the ordinary proportionality standard applies: the measure must pursue a legitimate aim, bear a rational connection to it, be necessary in the sense that no equally effective less restrictive alternative exists, and strike a proper balance. Where two fundamental rights collide, the lead opinion held, the proportionality standard must be applied to both rights. This is the double proportionality standard, drawn from the English decision in Campbell v MGN Ltd and adopted in Indian law in the concurring opinion in Subhash Chandra Agarwal, where Chandrachud J, as he then was, had first employed it. The court must identify the precise interests weighing in favour of each right, apply the proportionality enquiry in relation to each of the competing rights the measure is said to balance, and ask whether the means chosen restrict each right no more than the balance requires. The standard is more demanding than a loose balancing of interests precisely because it refuses to let one right be traded away wholesale for the other.

What is the double proportionality test in simple terms?

When a law restricts one right to serve a policy, the court asks whether the restriction is proportionate to the aim. When a law is said to balance two rights against each other, the double proportionality test applies the same structured enquiry to both rights: the law must be a proportionate way of protecting each, not a compromise between them.

Applied to the bonds, the double standard required the Union to show two things, not one. It had to show that the confidentiality the scheme conferred was proportionate as a protection of the donor’s informational privacy, and that the resulting deprivation of the voter’s information was proportionate as a restriction of the voter’s right. The scheme failed principally at the necessity stage of that enquiry. The Union, the lead opinion held, was unable to establish that Clause 7(4) of the scheme, the confidentiality clause, was the least restrictive means of balancing the two rights. Less restrictive designs were available within the law as it already stood, and the existence of a working alternative that protected the legitimate core of donor privacy while surrendering far less voter information was fatal to the claim that total anonymity was necessary.

What was clause 7(4) of the scheme, and why did it carry the case?

Clause 7(4) was the confidentiality provision of the Electoral Bond Scheme. It barred the issuing bank from disclosing the particulars of a bond purchaser to any authority or person, with two exceptions: a direction of a court, or the registration of a criminal case by a law enforcement agency. It converted a banking product into an anonymity regime.

The analytical weight Clause 7(4) bore in the judgment is easily understated. The bonds themselves were only paper. What made the scheme constitutionally distinctive was that the identity of every purchaser was locked inside the records of a single state-owned bank, releasable only through criminal process or court order. The Court’s necessity analysis therefore fastened on the clause: if the purpose was to move political money into banking channels and out of cash, the banking channel did that work by itself, through know-your-customer records and the audit trail of an account-based transfer. The confidentiality clause added nothing to the anti-cash purpose. What it added was the concealment, and concealment was the very thing the voter’s right forbade. The treatment of informational privacy elsewhere in this series, including the judgment that recognised it as a fundamental right, is traced in the account of the Puttaswamy privacy and Aadhaar judgment, whose proportionality framework the bonds Bench both used and extended.

The Court was careful, however, to separate two things the Union’s argument had run together. The privacy of political affiliation is real and constitutionally protected. Which party a citizen supports, like the vote the citizen casts, lies close to the core of political freedom, and the State may not compel its disclosure as a general matter. A financial contribution stands differently. A contribution is a transfer of resources into the electoral contest, with consequences for other participants and for the equality of the contest itself, and the pre-existing law had never treated it as private: contributions above the threshold in Section 29C of the election statute were reportable by the receiving party long before the bonds were conceived. The donor who contributed above that threshold under the ordinary law did so on terms of disclosure. What the scheme changed was not the boundary between public and private giving; it abolished the boundary for one instrument, and it did so, the Court noted, in conditions of asymmetry. The anonymity was never anonymity against the State. The issuing bank held the purchaser’s identity under its know-your-customer obligations, law enforcement could reach the record through the clause’s own exceptions, and the bank was an instrumentality of the very government whose party stood to receive the largest share. The veil fell between the donor and the voter, not between the donor and the ruling establishment, and a privacy protection that operates in one direction only is a weak candidate for the status of a fundamental-rights trump.

Necessity, and the inverted transparency of the amended code

Two features of the necessity analysis repay a slower reading, because between them they explain why the scheme failed at narrow tailoring rather than at any earlier stage. The first is the structure of the alternative the Court used. A necessity enquiry is only as strong as the alternatives it can point to, and courts are properly cautious about inventing policy alternatives of their own. The electoral trust was decisive precisely because it was not invented. It existed in the statute book, created by the same polity that later enacted the bonds, recognised under the Income Tax Act, and equipped with machinery that the Court could inspect rather than imagine. Its anatomy mattered. A trust formed for electoral purposes could receive contributions only through banking channels; it was required to keep records of its contributors and of its distributions; it reported to the Election Commission the contributions received and the contributions disbursed to parties; and because it stood between many donors and many parties, pooling what it received and distributing what it chose, the public record showed its receipts and its payments without identifying which of its donors stood behind which of its payments. In the Court’s analysis the trust therefore separated the two things the scheme had fused: the channel, which could be clean, and the correspondence between donor and party, which could be softened without being extinguished. A donor who feared the displeasure of a party could give to a trust; the money entered the disclosed system; the party received funds in the open; and no member of the public could say that this donor had funded that party. Measured against that device, the scheme’s total anonymity was not the least restrictive means of protecting donor privacy. It was a blunter instrument than the one the legislature already had, and Article 19(1)(a) does not permit the State to choose the blunter instrument when the cost of the choice is a constitutional right.

The second feature is an anomaly inside the amended code itself, which the petitioners pressed and which the judgment’s structure quietly vindicates. The Finance Act of 2017 did not only anonymise; it also tightened. The same statute that removed the donor’s name from bond contributions reduced the ceiling on anonymous cash donations, amending Section 13A so that a party’s exemption was imperilled by a cash contribution above two thousand rupees from one person. The result was an inverted transparency. A citizen who handed a party five thousand rupees in cash stood at the edge of the party’s tax exemption and within the record-keeping regime. A citizen, or a company, that bought bonds worth five crore rupees through the State Bank stood outside the reporting regime altogether, its name sealed by clause 7(4), the transaction exempted from the contribution report, the particulars beyond the reach of the voter. The larger the donation, the deeper the statutory silence. A code that regulates the small gift and immunises the large one is not drafted to a single intelligible principle of disclosure, and that inversion corroded the Union’s characterisation of the package as a transparency reform. The petitioners’ phrase for the design was that it legalised opacity at scale, and the judgment, without adopting the rhetoric, adopted the substance: whatever the scheme was for, the informational right of the voter was burdened most heavily exactly where the stakes for electoral influence were highest.

There is a further point about burden that the necessity stage exposes. Under a structured proportionality enquiry, the State that restricts a right must place before the court the material from which necessity can be judged: that the measure works, and that nothing less costly would work as well. The Union’s material on the black money limb consisted, in substance, of the assertion that channelled money is cleaner than cash and that anonymity was needed to draw donors into the channel. The first proposition was not in dispute, and it supported the channel, not the anonymity. The second was an empirical claim about donor behaviour, and the record contained a natural experiment against it: for the whole period of the interim order, donors had continued to buy bonds in very large volumes although particulars were being collected by parties and filed with the Commission, and contributions through trusts and electronic transfers had continued under full record-keeping throughout. Nothing in the record showed that the prospect of disclosure, as distinct from publication, had emptied the channel. The judgment does not belabour the point in the language of evidence law, but the structure of its reasoning is exactly this: at the stage where the State had to show that total anonymity was necessary, the showing was not made, and the less restrictive alternative stood in the statute book as a standing refutation of the claim that it could not have been made.

The black money answer

There remained the Union’s most politically resonant justification: that the scheme curbed black money in political finance by drawing contributions into the banking system. The Court’s answer operated at two levels, and the first is the one most often missed. The freedoms guaranteed by Article 19 may be restricted only on the grounds enumerated in the Constitution for each freedom. For the freedom of speech and expression, those grounds are listed in Article 19(2): the sovereignty and integrity of India, the security of the State, friendly relations with foreign States, public order, decency or morality, contempt of court, defamation and incitement to an offence. The purpose of curbing black money, the lead opinion held, is not traceable to any of the grounds in Article 19(2). However important the objective as policy, it is not a constitutionally permissible ground for restricting the freedom of expression, and a restriction that cannot be referred to a permitted ground cannot be saved by demonstrating that it pursues the impermissible one efficiently.

The second level made the first nearly academic. Even if the anti-black-money purpose were brought within the enquiry, the scheme was not narrowly tailored to it. The design did not merely channel money through banks; it severed the public link between giver and receiver, and the channeling objective was fully achievable without the severance. The Court pointed to the electoral trust as the standing proof. An electoral trust, a device the Income Tax Act had recognised for a decade before the bonds, collects contributions from donors and distributes them to one or more parties. The trust route cleans the money in the same way: contributions are received only through banking channels, never in cash, the trust maintains records of contributors, and it files annual reports of contributions received and disbursed with the Election Commission. Yet because a trust pools the gifts of several donors and distributes to several parties, the precise donor-to-party correspondence is not disclosed, and a measure of practical privacy survives. The comparison disposed of the necessity claim in both directions at once. Against the voter, the trust surrendered less information than total anonymity was claimed to require. Against the donor, it protected the interest that was genuinely at stake, the inability of a disappointed party to identify and punish its non-supporters, without extinguishing the public record of who was funding the system. A State that had the trust mechanism available, and chose instead an instrument of complete concealment, had not adopted the least restrictive means, and the scheme failed the proportionality enquiry at that step.

The channel and the cloak: what exactly was unconstitutional

A persistent misreading of the judgment treats it as a ruling against the modernisation of political donations, as though the Court had preferred cash and nostalgia to banking and records. The text supports no such reading, and the distinction on which the judgment actually turns is the most practically important in the case. The scheme, as designed, was a bundle of two separable things. The first was a channel: a defined instrument, bought through the banking system after know-your-customer verification, expressed in fixed denominations, valid for fifteen days, transferable to a single class of registered parties maintaining notified accounts. The second was a cloak: the statutory and scheme provisions that severed the transaction from the disclosure code, sealing the purchaser’s identity in clause 7(4), exempting the receipt from the contribution report, and relieving the party of the record-keeping condition of its tax exemption. The channel moved money visibly to the bank and invisibly to everyone else; the cloak was the invisibility.

The Court’s objections attach to the cloak, and to the cloak alone. No passage of either opinion holds that a donation instrument must not exist, that banking channels are constitutionally suspect, or that the State may not design the plumbing of political finance. The fifteen-day validity, offered by the Union as an anti-circulation safeguard, was not condemned; the eligibility threshold for recipient parties was not condemned; the know-your-customer discipline was the opposite of condemned, since the existence of complete bank records is what made the disclosure remedy possible. What failed, at each stage of the analysis, was the severance from disclosure. The informational holding is a holding about the exemption provisions and the confidentiality clause. The proportionality failure is a failure of the cloak: total anonymity was not the least restrictive means, which presupposes that a lesser anonymity, or none, attached to the same channel would have passed. Even the equality holding, though it struck at the ceiling’s deletion rather than at secrecy, is a holding about the terms on which the channel could be used by companies, not about the channel’s existence.

The distinction disciplines the judgment’s aftermath in both directions. Against the scheme’s defenders, it forecloses the claim that the Court left parties with no lawful way to receive large donations: the cheque, the draft, the electronic transfer and the electoral trust all survived the judgment untouched, because none of them carries the cloak. Against the scheme’s critics, it forecloses the claim that the judgment constitutionalised a complete code of political finance: the Court did not fix a disclosure threshold, did not cap individual giving, did not regulate party expenditure, and did not hold that every design feature of a future scheme must resemble the pre-2018 law. It held that the funding of parties must be visible to the voter and that corporate money must be rationed by substance, and it left the engineering of a system satisfying those two requirements to Parliament. Judicial interpretation articles exist to mark exactly this kind of boundary, between the principle a court lays down and the design space it leaves open, and few modern Indian judgments draw the boundary as cleanly as this one: the channel was the legislature’s, the cloak was unconstitutional, and the difference between the two is the case.

The second holding: equality and the corporate ceiling

The challenge to the Companies Act amendments stood on a different constitutional footing, and the Court kept it there. Section 182 of the Companies Act, 2013 had, before the Finance Act, contained a calibrated regime. A company other than a government company could contribute to a political party up to seven and a half per cent of its average net profits during the three immediately preceding financial years, with board authorisation, and was required to disclose in its profit and loss account the total amount contributed and the name of the party to which each contribution was made. The Finance Act deleted the ceiling and removed the party-wise disclosure. The architecture of the Companies Act 2013 as a whole is treated in this series’ guide to that statute; the question here is what the Court held about the 2017 changes to it.

The lead opinion held the deletion of the ceiling manifestly arbitrary and violative of Article 14. Manifest arbitrariness, the doctrine drawn from the line of authority running through Shayara Bano, condemns legislation that is capricious, irrational or without an adequate determining principle. The Court identified the arbitrariness in a comparison. The amended provision treated two contributors who are not alike as though they were alike. An individual’s political contribution and a company’s political contribution differ in kind. An individual contributes, in the paradigm case, from conviction: the contribution expresses support for a party’s programme. A company has no convictions. Its contribution is, in the Court’s analysis, by its nature a business transaction, made with the expectation of influencing the policy environment in which the company operates, and carrying an inherent potential for quid pro quo arrangements between the contributor and the governing party. A legal regime that licenses both forms of giving on identical terms, in unlimited amounts, ignores the difference and thereby fails the elementary requirement of equality law that unequals not be treated as equals.

Two features of the amended provision aggravated the defect. The first was the absence of any ceiling whatsoever. The deleted proviso had tied the permissible contribution to profits, which both limited the scale of corporate influence and confined it to companies with profits to give. Its deletion permitted contributions without limit. The second was the admission of loss-making companies. Because the ceiling had been expressed as a fraction of average net profits, a company without profits could contribute nothing under the old regime. After the amendment, a company with no profits, including a shell company with no genuine business, could contribute any amount its controllers chose to route through it. The lead opinion treated that consequence as destructive of the provision’s own rationale: a regime defended as governing corporate political participation in fact created a channel through which money of entirely external origin could enter the electoral process wearing a corporate form, and it infringed, in the Court’s words, the principle of free and fair elections itself.

The consequence of the holding is the part of the judgment with the longest reach. Because the deletion fell, the pre-existing provision revived: the ceiling of seven and a half per cent of average net profits, and with it the structure of Section 182 as it had stood. Because the disclosure amendment fell on the Article 19(1)(a) ground discussed earlier, party-wise disclosure in the profit and loss account revived as well. But the reasoning reaches beyond restoration. The equality holding does not depend on secrecy. A future legislature that enacted unlimited corporate contributions inside a regime of perfect transparency would still have to answer the objection that a company and an individual are unlike contributors, that the company’s gift is a transaction, and that treating the two alike is arbitrary. That is the sense in which the series’ thesis thread holds here: the judgment’s reputation rests on the information holding, which produced the disclosure and the headlines, while its durable constraint on future legislation rests on the equality holding, which is quoted far less and binds far more.

Political equality: the strand beneath the equality holding

The Article 14 holding is sometimes read as a technicality, an arbitrariness point that happened to succeed where broader challenges might have failed. That reading misses the strand of election law into which the holding is woven. Free and fair elections are not a statutory value in India; since Indira Nehru Gandhi v Raj Narain they have been treated as part of the basic structure of the Constitution, a feature the amending power itself cannot abrogate. The content of the value has been elaborated case by case, and its recurring core is political equality: the premise that the democratic process belongs to citizens on equal terms, one person and one vote, and that advantages converted from outside the process, from wealth, office or organisation, are constitutionally suspect when they deform the equality of the process itself. The ceiling in Section 182 was company law’s expression of that premise. It did not prohibit corporate money; it rationed it, by a formula tied to the company’s own substance, so that the scale of a company’s political voice bore some relation to the scale of the enterprise whose interests it claimed to speak for, and it excluded by arithmetic the company that had no substance at all.

Read against that strand, the deletion of the ceiling did three things at once, and the judgment’s arbitrariness analysis is a compressed account of all three. It removed the ration, so that the volume of corporate political money became a function of nothing but the willingness of those who controlled companies. It removed the substance test, so that the privilege of contributing detached from any genuine business whose interests might legitimate the contribution, and attached instead to the mere fact of incorporation, a legal shell that can be had cheaply and in numbers. And, in combination with the amendment of Section 182(3), it removed the trace, so that even the discipline of publicity, which might have substituted in part for the ration, was withdrawn at the same moment. A legislature might defend any one of those removals. Defended together, they amounted to a decision that the channel of corporate influence over elections should be unlimited in volume, open to entities without substance, and invisible in operation. It is that combination the Court held the Constitution would not sustain, and the holding is best understood as the basic structure value of free and fair elections doing work inside an Article 14 analysis, the doctrine of manifest arbitrariness supplying the form and the election jurisprudence supplying the content.

The Union’s answer to this strand deserves its place, because it identifies the live question for any successor scheme. Corporate donations, the answer ran, are not inherently corrupting; they are a feature of political finance in democracies generally, companies have legitimate interests in the policies that govern the economy, and a ceiling pegged to profits is an arbitrary instrument of its own, favouring the old and the profitable over the young enterprise. The judgment does not deny the premise. It restores a ceiling; it does not prohibit contributions; and its reasoning is careful to locate the vice in the erasure of distinctions, of profit from loss, of substance from shell, of company from citizen, rather than in corporate giving itself. The question left for a future legislature is therefore one of design rather than of permission: what distinctions a scheme draws, between contributors and between amounts, and what publicity attends them, will decide its fate under the two holdings, and a scheme that rationed corporate money by substance and disclosed it by party would present the Court with a materially different case from the one decided here.

The concurrence of Khanna J

Khanna J agreed with the conclusions of the lead opinion and differed in parts of the reasoning, and the differences are doctrinally instructive rather than decorative. Where the lead opinion struck down the deletion of the corporate ceiling on the ground of manifest arbitrariness, Khanna J applied the principle of proportionality, observing that the proportionality enquiry would subsume the test of manifest arbitrariness: a measure that fails structured proportionality does not need a second label for its defect. He also examined the privacy claim from the corporate side, and held that the claim of privacy by a company, particularly a public limited company, rests on very limited grounds. A company’s affairs must be open to its shareholders and to the public who deal with it; it would be difficult for such a body to claim a violation of privacy in respect of its political contributions, and any residual privacy interest is properly the interest of the individuals conducting its business, not an institutional cloak for the contribution itself.

On one provision, the concurrence went further than the lead opinion. Section 31(3) of the Reserve Bank of India Act, inserted by the Finance Act, had been enacted to permit the issuance of the bonds as instruments payable to the bearer on demand, which is what made them transferrable in anonymity from hand to hand. Khanna J held that provision unconstitutional and struck it down, reasoning that a bearer instrument of this character was itself impermissible. The lead opinion, having struck down the scheme the provision existed to serve, did not rest a separate holding on it. The practical outcome is unaffected: the scheme is gone, and with it the market for the instruments the Reserve Bank amendment enabled. But for the completeness of the record, the position is that the bearer-bond provision of the Reserve Bank Act stands struck down on the authority of the concurrence, in a judgment whose conclusions were otherwise unanimous.

Khanna J also framed the remedial consequence in a form worth noting. His directions contemplated the Election Commission ascertaining the details of bond contributions from the political parties, from the issuing bank and from the bankers of the parties, and thereupon disclosing the names of donors and purchasers and the amounts donated. The operative directions in the event were those of the lead opinion, set out in the next part of this article, which placed the furnishing duty on the State Bank of India and the publishing duty on the Commission. The two remedial visions converged on the same destination: the data held in the bank’s two carefully separated records, of purchase and of encashment, was to be brought into the open.

What fell, on what ground, and what survives

The table below is the record of the case in one view. Each provision the Court considered is set against what it did, the ground on which it was struck down or left undecided, the constitutional provision applied, and the position that follows from the judgment.

Provision	What it did	Ground of decision	Constitutional provision	Position after the judgment
Electoral Bond Scheme, 2018	Created anonymous bearer bonds for political donations, sold by the State Bank of India, with donor identity sealed by the confidentiality clause	Violates the voter’s right to information; fails the double proportionality test because total anonymity was not the least restrictive means; curbing black money is not a permitted ground of restriction	Article 19(1)(a)	Struck down; issue of bonds stopped forthwith; valid unencashed bonds returned and refunded
Proviso to Section 29C(1), Representation of the People Act, 1951	Exempted contributions received through electoral bonds from the contribution reports parties must file	The exemption anonymised contributions and infringed the voter’s right to information	Article 19(1)(a)	Struck down; the reporting obligation covers bond-era contributions as it covers others above the threshold
Proviso to Section 13A, Income Tax Act, 1961	Relieved parties of the duty to record contributors’ names and addresses for bond receipts as a condition of tax exemption	The exemption removed the record on which disclosure depends and infringed the right to information	Article 19(1)(a)	Struck down; the record-keeping condition applies to all contributions
Deletion of the first proviso to Section 182(1), Companies Act, 2013	Removed the ceiling of seven and a half per cent of average net profits on corporate political contributions	Manifest arbitrariness: companies and individuals treated alike though unlike; loss-making and shell companies admitted; free and fair elections infringed	Article 14	Deletion struck down; the profit-linked ceiling stands restored
Amendment of Section 182(3), Companies Act, 2013	Required disclosure only of the total contributed, removing party-wise disclosure in the profit and loss account	The voter cannot identify which party a company funded, defeating the right to information at the point of use	Article 19(1)(a)	Struck down; party-wise disclosure stands restored
Section 31(3), Reserve Bank of India Act, 1934	Permitted the bonds to be issued as instruments payable to bearer on demand	Held unconstitutional in the concurring opinion of Khanna J as permitting bearer instruments of this character	Articles 19(1)(a) and 14 framework	Struck down on the concurrence; the scheme it served struck down unanimously
The Finance Act, 2017 as a Money Bill	Supplied the legislative vehicle for every amendment above	Not examined; the scope of Article 110 stands referred to a seven-judge Bench in Rojer Mathew and is sub judice	Article 110	Open; the amendments fell on rights grounds without any ruling on the vehicle

The table’s final row is the one that does not belong to the pattern of the others, and it points to the section of this article that follows the disclosure narrative: the judgment decided the fate of the amendments without deciding whether they had been competently enacted, because the vehicle question was already before a larger Bench in another case.

The sealed covers: how the archive survived

The disclosure remedy of February 2024 was possible because of a procedural device adopted five years earlier, and the device deserves its own account in any complete study of the case. When the challenge was young and the scheme was new, the Court faced the classic dilemma of interim relief in a case about a running system. To stay the scheme would have decided much of the case before it was argued, on a prima facie view, and would have disrupted a funding channel parties had begun to use. To refuse a stay without conditions would have let the transactions the petitioners called unconstitutional proceed unrecorded in any usable form, so that even a complete victory at the final hearing would have arrived too late to reconstruct what had happened. The interim order of 12 April 2019 chose the middle course that defines the case’s procedural character: the scheme would run, but the record would be preserved. Political parties were directed to submit to the Election Commission, in sealed cover, the particulars of the contributions they received through electoral bonds, donor identities included, both for the period already elapsed and on a continuing basis as the litigation proceeded.

Sealed cover procedure has attracted sustained criticism in Indian public law, and the criticism is not difficult to state: material placed before a court or an authority under seal is material the opposing party and the public cannot test, and a jurisprudence built on such material corrodes the openness on which adjudication depends. The device in this case, though, was directed to a different end from the uses that attract the criticism. The sealed covers were not evidence on which the Court proposed to decide while keeping the parties in ignorance. They were an escrow. Their function was preservative rather than adjudicative: to hold the informational record intact, in the custody of the constitutional authority responsible for elections, until the Court could decide, with full argument, whether the information belonged to the public or to the confidence of the scheme. Either answer at the final hearing would then be effective. If the scheme had been upheld, the covers could have remained sealed and the confidentiality the statute promised would have been substantively intact. If the scheme fell, as it did, the information would exist to be disclosed, and the informational right would not be defeated by the simple passage of the five years the litigation consumed.

The March 2024 sequence completed the device’s logic and also revealed its limits. The judgment’s directions ran against the State Bank rather than the sealed covers, because the bank’s records were the master record: purchases as well as encashments, with the bond numbers as the joining key, where the parties’ filings recorded only what each party had received. The sealed covers thus receded in importance at the moment of disclosure, their preservative work done, while the bank’s archive supplied the publishable whole. There is a lesson in the sequence for the conduct of constitutional litigation over running systems. Interim orders are usually treated as temporary accommodations, forgotten when the judgment arrives. Here the interim order determined the value of the final one. A differently drawn order in April 2019, a stay refused without a record condition, would have left the Constitution Bench of 2024 pronouncing on rights whose violation it could no longer repair, and the most celebrated feature of the judgment, the opening of the bond record to the electorate, would have been beyond its power. The case is therefore a study in remedies twice over: in the proportionality of the final directions, and in the foresight of the interim one that kept the final directions worth giving.

The remedy: disclosure rather than a merely prospective invalidity

Constitutional courts possess a quiet instrument for softening the consequences of invalidation, and its absence from this judgment is one of the judgment’s most deliberate features. Where a law has governed transactions for years and parties have ordered their affairs on its faith, a court may declare the law invalid only for the future, leaving the past untouched. The doctrine of prospective overruling exists for exactly the reliance interest that the donors under this scheme could plead: they had given money under a statute, on an express statutory assurance of confidentiality, and a purely prospective judgment would have protected that assurance for every completed transaction while ending the scheme. The Bench did not take that course. It invalidated the scheme and the amendments, and then reached backwards, ordering the disclosure of six years of transactions to the Election Commission and through it to the public. An article on judicial interpretation must ask why, because the choice was the hinge between the judgment as a statement of doctrine and the judgment as an event.

The reasoning is distributed across the judgment and the March orders rather than collected in one passage, but it assembles into three strands. The first is the nature of the right. What had been infringed was an informational right, and an informational injury is not remedied by a declaration that the infringement will stop. The voter who had voted in the elections of 2019 and 2022, and in the state elections of the intervening years, had voted without information the Constitution entitled the voter to have. A prospective-only ruling would have left that injury entirely unrepaired and would have meant that the scheme, though unconstitutional from its first day, had successfully withheld from the electorate the very information whose withholding made it unconstitutional, for the whole of its working life. Disclosure was the only remedy that answered the violation in kind.

The second strand is notice. The reliance interest that prospective overruling protects is reliance free of warning, and this reliance was not free of warning. From 12 April 2019 the scheme operated under the shadow of the interim order: the challenge was pending, the Court had refused a stay but had required particulars to be collected and filed, and every purchaser after that date bought a bond knowing that the donor’s identity existed in records held under the authority of a court that had the scheme under active constitutional examination. The Bench made the point expressly in March, when the industry bodies intervened against the disclosure of bond numbers: everyone had been put on notice in April 2019. The cut-off in the directions is the remedial expression of the same point. Disclosure runs from the date of notice, not from the date of the scheme, and the first year of the scheme’s life, in which bonds were bought without that warning, was left outside the order. The remedy thus drew its own line between reliance that deserved protection and reliance that had been warned, and drew it at the interim order.

The third strand is feasibility, and it is the strand the bank’s extension application tested and lost. Retrospective remedies fail when the past cannot be reconstructed. Here the past was fully documented, because the scheme was a banking product: every bond carried a unique alphanumeric number, every purchase had passed know-your-customer verification, every encashment had moved through a notified party account. The archive existed, complete and exact, in the custody of a single public sector bank. The bank’s difficulty was never that the record did not exist; it was that the record had been deliberately divided into two silos so that no ready-made table joined purchaser to party. The Court treated that division as a design choice made to serve the scheme’s confidentiality, and refused to let the design choice mature into a remedial impossibility once the confidentiality itself had been held unconstitutional. There is a principle embedded in that refusal which reaches beyond this case: a State that engineers opacity into a system it operates cannot then plead the opacity it engineered as the reason a court cannot order light. The directions that follow in the next section of this article should be read with that principle in view, because each of them, the deadline, the refusal of extension, the insistence on the bond numbers and the affidavit of completeness, is an application of it.

The directions and the disclosure sequence

A judgment’s directions are part of its holding in a case of this kind, and here they were unusually specific. In paragraph 219 of the lead opinion, the Court directed, first, that the issuing bank stop the issuance of electoral bonds forthwith. Second, the State Bank of India was to submit to the Election Commission the details of electoral bonds purchased since the interim order of 12 April 2019: the date of purchase of each bond, the name of the purchaser and the denomination. Third, the bank was to submit the details of political parties that had received contributions through electoral bonds in the same period, including for each bond encashed the date of encashment and the denomination. Fourth, the information was to reach the Commission within three weeks of the judgment, by 6 March 2024, and the Commission was to publish it on its official website within one week of receipt, by 13 March 2024. Fifth, electoral bonds still within their fifteen-day validity period and not yet encashed were to be returned by whoever held them, the party or the purchaser, to the issuing bank, which was to refund the amount to the purchaser’s account.

The compliance that followed is a sequence worth narrating exactly, because it is itself an exercise of supervisory power of a kind courts rarely have occasion to repeat. Two days before the deadline, the State Bank of India filed a miscellaneous application seeking an extension of time until 30 June 2024. Its case, argued by senior counsel, was one of operational difficulty: the bank had maintained the purchase records and the encashment records in two separate silos, deliberately unlinked to preserve the scheme’s confidentiality, and the work of decoding and matching them could not be completed in three weeks. The petitioners responded with a petition invoking the Court’s contempt jurisdiction. By order dated 11 March 2024, reported at 2024 INSC 195, the Constitution Bench dismissed the application. The reasoning was that the information the directions required was information the bank held; the directions of 15 February had specified the fields to be furnished, and the bank’s internal architecture of secrecy, adopted to serve the scheme, could not be converted into a reason for withholding what the judgment had ordered disclosed. The bank was directed to furnish the details by the close of business on 12 March 2024, the Commission was directed to publish them by 15 March 2024, and the bank’s Chairman and Managing Director were required to file an affidavit of compliance.

The bank furnished the data on 12 March and the Commission published it on 14 March 2024. What the publication revealed, however, was the practical effect of the two silos. The purchase data and the encashment data appeared as separate datasets. A reader could learn who had bought bonds of a given denomination on a given date, and could learn which party had encashed bonds of a given denomination on a given date, but could not connect a particular purchase to a particular encashment, because the one field common to both records, the unique alphanumeric number borne by each bond, had not been disclosed. For anyone working through the published files, including the spreadsheets and slide decks through which newsrooms circulated the two datasets side by side and read them in a free online PowerPoint viewer and similar tools, the limitation was immediate: the aggregates were public, the correspondence between them was not. On 15 March 2024 the Court issued notice to the bank on the omission of the bond numbers.

On 18 March 2024 the Bench heard the matter and closed the gap. It held that there was no manner of doubt that the bank was required to make complete disclosure of all details in its possession; the word “including” in the original directions, preceding the specified fields, showed that the fields were illustrative and not exhaustive, and the bank was not to be selective in what it furnished. The bank was directed to disclose all details, including the unique alphanumeric number and the serial number of each bond. The Chairman of the bank was directed to file an affidavit by 5 p.m. on 21 March 2024 stating that the bank had disclosed all details in its custody and that no details had been withheld, and the Commission was directed to upload the details on its website forthwith upon receipt. Industry bodies, the Federation of Indian Chambers of Commerce and Industry and the Associated Chambers of Commerce and Industry, sought to be heard against the disclosure of the bond numbers; the Court observed in response that everyone had been put on notice in April 2019, when the interim order directed the collection of details. On 21 March 2024 the bank furnished the complete data with the bond numbers and the Chairman’s affidavit, and the Commission published it the same day. The two datasets could then be joined, bond by bond, across the whole period from 12 April 2019 to 15 February 2024.

What the disclosed archive could and could not establish

Because the disclosure is the feature of the case that reached citizens who never read judgments, its evidentiary character deserves a precise statement, and this article’s register, fixed by its brief, is to state it without crossing it. The completed archive established, for every bond in the window, four facts: that a named purchaser bought a bond of a stated denomination on a stated date; that a named party encashed a bond of a stated denomination on a stated date; and, through the bond number, that these two events were the two ends of one instrument. Assembled across the window, those facts support findings of aggregate and structure: the total volume of money that moved through the scheme in the window, the distribution of encashments among parties, the distribution of purchases among purchasers, the timing of purchases and encashments in relation to the calendar of elections and of bond sale windows. Findings of that order are findings about a system, and they are the findings the informational right was vindicated to make possible: an electorate, a press and a research community able to describe how political money actually flowed, at the level of pattern, during six years of elections.

The archive does not establish the further facts that public discussion most eagerly sought from it. A correspondence between a purchase and an encashment is consistent with many purposes: conviction, habit, access, insurance, expectation, pressure. The record carries no field for purpose, no record of any communication between purchaser and party, no evidence of any act of government connected in time or substance with a donation. To move from the archive to a conclusion about the conduct of any party, company or individual requires evidence the archive does not contain, gathered by processes, investigation, prosecution, adjudication, that the judgment neither initiated nor prejudged. The Court itself was careful on the point: the disclosure directions were framed as the publication of a record held by a bank and a Commission, the petitioners’ request for a court-monitored investigation into the disclosed material was not converted into any direction against any person, and no finding in the judgment or the March orders attributes a motive to a single transaction among the thousands disclosed.

That restraint is not a technicality, and it shapes the responsible uses of the data. Verification is one: the encashment record can be laid beside the contribution reports parties filed for the same years, and agreements or silences between the two public records are themselves facts of public interest about the reporting system. Study is another: the concentration of purchases, the recurrence of particular denominations, the rhythm of sale windows against election schedules, are materials for the political science of campaign finance that India has historically had to conduct on estimates. Accountability of the third kind, the kind that names a donor and a party and asserts a bargain, belongs to other institutions and other evidence, and commentary that performs it on the archive alone misdescribes its source as much as it wrongs its subjects. This article has accordingly drawn no such conclusion anywhere in its pages, and it records the boundary here because the boundary is part of what the judgment, correctly read, decided: the voter was given the record because the record was the voter’s, and what the record means for any individual transaction is a question the disclosure was never designed, and must not be used, to answer by itself.

Two opinions, one result: the precedential structure

Because the case will be cited for decades, its internal structure deserves a section of its own. The judgment is unanimous in result and dual in reasoning, and the two facts interact. Where a Constitution Bench speaks with one voice on the operative conclusions, lower courts and later Benches receive the holdings at full strength: the scheme is unconstitutional, the four amendments are unconstitutional, the directions bind. Those conclusions carry the authority of all five judges, and no argument is open to a future litigant that the case was carried by a fractured court. The reasoning that supports the conclusions, however, is distributed between two opinions that agree on the destination and differ at two identifiable junctions, and a reader citing the case must know which opinion carries which proposition.

At the first junction, the standard for corporate contributions, the lead opinion travels by Article 14 and manifest arbitrariness, while Khanna J travels by a proportionality enquiry that he holds subsumes the arbitrariness analysis for provisions of this kind. The holdings converge, but the routes have different futures. A future Bench confronting a fresh restriction on corporate funding can cite the lead opinion for the proposition that the classification of companies with individuals is unconstitutional in itself, a defect of the law’s design that no justification can cure. It can cite the concurrence for a more graduated proposition: that the State must justify the restriction or the permission as proportionate to a legitimate electoral aim. Which route governs the next case is a question the bonds Bench deliberately left distributed, and counsel in the next case will argue it from both texts.

At the second junction, Section 31(3) of the Reserve Bank of India Act, the distribution matters more sharply. The lead opinion does not rest a holding on the provision; the concurrence strikes it down. The practical consequence is nil, because a provision that exists only to serve a struck-down scheme has no field of operation left, and no future bond can issue under a dead scheme whether or not the provision formally stands. The doctrinal consequence is the kind of point on which examinations and later judgments fasten. The bearer character of the bonds, an instrument payable to a bearer on demand, transferable by delivery, sitting outside the discipline that the law of negotiable instruments ordinarily imposes, was in the concurrence a free-standing constitutional vice, because it was the bearer design that made the bond a currency of anonymous influence rather than a mere banking channel. That strand of reasoning is available to any future court confronting anonymous financial instruments in regulated spaces, and it travels on the authority of a single judge’s opinion in which the result was unanimous but the point was not joined. The careful reader will therefore cite the case in layers: unanimous for the invalidity of the scheme and the amendments and for the rights holdings, majority-reasoned for the double proportionality route and the arbitrariness route, and single-opinioned for the bearer instrument holding. The judgment rewards that care, because its two authors wrote as judges conscious that they were building doctrine for cases that had not yet been imagined, and the layering is part of the building.

Reading the judgment: a guide to its structure

The reported judgment runs to two hundred and nineteen paragraphs in the lead opinion, with a substantial concurrence beside it, and a reader approaching it for the first time benefits from a map. The lead opinion moves in five discernible movements, and each movement does a different kind of work. The first movement is narrative and architectural: the scheme, the amendments, the petitions and the course of the litigation, including the interim order, are set out so that the legal analysis that follows has its facts fixed in one place. The second movement is jurisdictional in the small, decisive sense discussed earlier in this article: it settles the character of the measure and the standard of review, rejecting the deference claimed for economic legislation and framing the questions as questions about the democratic polity. A reader who skips this movement will misunderstand everything after it, because the intensity of the review that follows is established here, in the classification of the subject rather than in the analysis of the rights.

The third movement is the informational one, and it is the opinion’s centre of gravity. It builds the right in stages: the candidate cases and what they actually decided, the step from candidate to party and the institutional reasons for taking it, the connection between funding information and the detection of quid pro quo arrangements, and then the structured proportionality enquiry in its doubled form, with the black money objective tested against Article 19(2) and the necessity of clause 7(4) tested against the electoral trust. The fourth movement is the equality one, shorter and more densely doctrinal, in which the Companies Act amendments are taken apart: the character of a corporate contribution, the collapse of the distinction between the profitable and the loss-making, the shell company as the reductio of the amended design, and the infringement of free and fair elections. The fifth movement is remedial and conclusory: the findings are gathered provision by provision, in the form reproduced in the artifact table earlier in this article, and the directions issue, ending at paragraph 219 with the disclosure regime and the refund of valid unencashed bonds.

The concurrence is organised on a different plan, and the difference is instructive. Khanna J proceeds issue by issue in a more compressed register, restating the challenge and the Union’s three-limbed defence, and then applying to the whole matter a unified proportionality analysis in which the manifest arbitrariness enquiry is absorbed rather than run in parallel. Two features of his opinion make it the one practitioners reach for on particular points. The treatment of Section 31(3) of the Reserve Bank of India Act appears only here, as this article has recorded. And the discussion of the standard of review for electoral legislation, including the refusal to import the deference of the insolvency and economic cases into the law of democracy, is stated here with a bluntness the lead opinion’s more architectural manner does not attempt. A citation practice has already formed around the case in consequence: the lead opinion for the rights holdings and the directions, the concurrence for the standard of review and the bearer instrument point, and both together for the unanimity of the result.

Three cautions belong in any guide to the text. The first concerns headnotes. The report in the Supreme Court Cases carries an editorial headnote of unusual length, and it is a finding aid, not authority; the holdings are those stated in the opinions, and where the headnote’s compression and the opinion’s text diverge, the text governs. The second concerns the March orders. The orders of 11 and 18 March 2024 are reported and reasoned separately from the February judgment, the first of them at 2024 INSC 195, and they are part of the case’s operative law: the dismissal of the extension application construed the directions, and the order on the bond numbers construed the word “including” in them, so a complete account of the disclosure regime requires all three documents. The third caution concerns the undecided questions. Both opinions are scrupulous in marking what they do not decide, the Money Bill reference above all, and the reader who treats the judgment’s silences as holdings will misstate the law in both directions, crediting the case with a ruling on Article 110 it refused to give and with a ruling on the foreign contribution amendments it expressly reserved. The map this section has drawn is, in the end, the discipline of the whole article: two holdings, one concurrence with a holding of its own, a remedial regime completed by two later orders, and a boundary of open questions that the Court drew itself and that honest commentary must preserve.

What the judgment did not decide

The discipline of a judicial interpretation article is to mark the boundary of the holding as carefully as its content, and this judgment has an unusually consequential boundary. The largest open question is the one flagged in the table. Every amendment the Court struck down had been enacted through the Finance Act, 2017, which had been introduced and passed as a Money Bill under Article 110 of the Constitution, a route that excludes the Rajya Sabha from any power beyond recommendation. The petitioners had challenged that route as well. The Court did not examine it. Both opinions record the reason: the scope of Article 110 stands referred to a seven-judge Bench in Rojer Mathew v South Indian Bank and is sub judice, so the bonds Bench tested the amendments on fundamental-rights grounds and left the vehicle question untouched. The consequence for the practical lawyer is recorded at the treatment of the money bill and finance bill distinction, and it is this: a future Parliament that re-enacted donor anonymity in a free-standing bill passed by both Houses would face the rights holdings of this judgment, but the distinct objection that electoral reform of this magnitude was smuggled past the upper House remains unadjudicated, available for the day the seven-judge Bench speaks.

Four further limits deserve statement. First, the challenge to the amendments made to the Foreign Contribution (Regulation) Act by the Finance Acts of 2016 and 2018, which had altered the treatment of contributions from foreign-linked companies, was expressly not decided; a batch of petitions raising it was pending, and the Bench left it there. Second, the Court laid down no design for a replacement. It invalidated a scheme and its statutory supports and it identified, in the electoral trust, an existing mechanism that satisfied the standard the scheme had failed; it did not enact a funding code, fix a disclosure threshold of its own, or hold that any particular ceiling on individual contributions would be valid. The space of lawful political finance after the judgment is the pre-2018 space, restored, not a new settlement drafted by the Court. Third, the judgment ordered disclosure and refunds of unencashed bonds; it did not order any political party to disgorge donations already received and spent, it attached no consequence to any individual donation disclosed, and it converted no part of the disclosed data into a finding against any person. Fourth, the cut-off of 12 April 2019 was left as drawn. The first year of the scheme’s operation lies outside the ordered disclosure, and the Court declined the invitation to extend the window backwards, regarding the date as a conscious choice of the Bench rather than an error open to correction in miscellaneous proceedings.

The criticism, stated fairly

The register of this article is to report the judgment and the sequence it produced as facts, and to give the criticism of the judgment the same full-strength statement the stance charter requires, without adjudicating between the positions. Five criticisms recur in serious commentary and were, in substance, anticipated by the Union’s submissions in the case itself or by the sequence that followed the judgment.

The first is textual. The right to information about political funding, the criticism runs, is a judicial extension of the freedom of expression rather than a right the text confers: the 2002 and 2003 decisions concerned candidates, whose antecedents bear directly on the vote for that candidate, and the move from the candidate to the party’s treasury crosses from interpretation into legislation. The answer recorded in the judgment is institutional, as set out earlier in this article: parties select, finance and discipline the candidates, so a line at the candidate’s person cannot be maintained without incongruity, and the informational right had always been grounded in the conditions of an effective vote rather than in the identity of the information’s subject.

The second criticism concerns the weight given to retaliation. Donor anonymity, it is said, protects the small and the vulnerable contributor against the displeasure of parties in power, and the judgment dissolves that protection in favour of an informed electorate whose actual use of funding data is speculative. The answer in the judgment is twofold: the anonymity the scheme conferred was asymmetric, since the State’s own bank held the identities and the State could reach them, so the protection ran against voters and opposition parties rather than against power; and the electoral trust mechanism demonstrates that a practical measure of protection, the dissociation of a donor from a specific recipient, can be preserved without extinguishing the public record.

The third criticism is about the judicial role. A necessity test that asks whether the legislature chose the least restrictive means, it is argued, substitutes the court’s assessment of policy design for the legislature’s, in a field, political finance, where every design trades values that a democracy might weight differently. The answer in the judgment is that proportionality is the settled standard for rights adjudication since Puttaswamy, that the burden it placed on the Union was the ordinary burden of justifying a rights restriction, and that the Court in fact invalidated the scheme at the point of narrow tailoring rather than at the level of objective: the objective of clean money was never held illegitimate, only unusable as a ground under Article 19(2) and unachieved by the means chosen.

The fourth criticism is about retrospectivity. Donations were made, between 2018 and the judgment, under a scheme enacted by Parliament and operative in law, by donors who relied on a statutory promise of confidentiality; ordering the disclosure of their identities afterwards, the criticism runs, defeats a legitimate expectation and is unfair to the donor even if the scheme was vulnerable. The answer that prevailed in the March proceedings is the one the Chief Justice gave to the industry bodies: the interim order of April 2019 had put everyone on notice that particulars were being collected under the Court’s authority, so reliance on permanent secrecy after that date was reliance against a visible risk; and a scheme held unconstitutional is not ordinarily capable of generating a protected expectation that its concealment will survive its invalidity. The fifth criticism is about timing. The judgment was delivered on 15 February 2024 and its disclosure directions ran through the following five weeks; the Election Commission announced the schedule of the general election on 16 March 2024, in the middle of the disclosure sequence, and polling opened the following month. A ruling that abolished the principal channel of party funding and published six years of its record on the eve of a general election, the criticism runs, intervened in the electoral contest it purported to protect, unsettling the finances of every party at the moment of maximum dependence on them and handing the disclosed record to the campaign itself as material. The answer is one of chronology and of institutional role. The petitions had been pending since 2017; the hearing had concluded and judgment had been reserved in November 2023, months before any election schedule existed; and a court that timed its judgments by the electoral calendar would be performing the political calculation the criticism attributes to it. The sequence that followed, judgment, disclosure, election, is the ordinary consequence of a case decided when it was ready, and the alternative urged by the criticism, that constitutional adjudication should pause as elections approach, would hand every incumbent scheme a recurring immunity in a country that is seldom far from an election of consequence.

This article records both the criticism and the answer without preferring either, and it observes the boundary the brief of this case imposes on all commentary: the disclosed data establish that purchases and encashments occurred, in stated amounts on stated dates, and a correspondence between a purchase and an encashment establishes nothing about the purpose of either. No inference about the conduct of any party, company or individual is drawn in this article from the disclosed record.

The questions that return to Parliament

A judgment that restores an older law does not settle the policy that produced the newer one, and the pressures that carried the Finance Act of 2017 through Parliament did not dissolve with the scheme. Political parties require money on a scale that membership and small donations have never supplied in India, campaigns cost more with each cycle, and the demand for funds will press against whatever rules stand. The judgment changes the terms on which that pressure may lawfully be answered, and it is worth identifying, with the two holdings in hand, the questions that return to the legislature and the constraints the answers will face.

The first question is whether any form of donor anonymity can be rebuilt. The information holding forecloses the design that failed: total anonymity, for contributions of every size, against the voter while the State retains access. What the holding does not foreclose, because it was not asked, is a calibrated regime: a disclosure threshold set at a level that protects the genuinely small donor, publication of contributions above it, aggregation rules that prevent the splitting of one large gift into many small ones. The proportionality grammar of the judgment supplies the test any such regime must pass. The State would have to show that the anonymity it retained served the protection of donors against retribution rather than the convenience of parties, that the threshold bore a rational relation to that purpose, and that the electoral trust route did not already achieve the protection at less cost to the voter’s right. Those are answerable questions, and a legislature that answers them on a record, with evidence rather than assertion, would litigate the next case from a stronger position than the Union occupied in this one.

The second question is the individual contributor. Indian law has never capped what an individual may give to a party, and the bonds judgment does not create such a cap; its informational logic is indifferent to amount provided the gift is disclosed, and its equality logic was directed at the corporate form. Whether unlimited disclosed individual giving is compatible with political equality in the long run is a debate the judgment opens without joining, and it will be conducted, if it is conducted, in Parliament rather than in court, because no holding in the case supplies a court with the materials to impose a ceiling the legislature has not chosen.

The third question is the demand side. The law polices what candidates may spend through the election expenditure regime and leaves party expenditure in a far looser position, an asymmetry reform commissions have remarked on for decades. Committees of the past, from the Indrajit Gupta Committee on state funding of elections onward, have proposed meeting part of the legitimate cost of politics from public funds in exchange for tighter discipline on private money. The judgment neither adopts nor obstructs that agenda. It observes, by its structure, only this: a Parliament that wishes to reduce parties’ dependence on large private donations has instruments available that do not require secrecy, and the existence of those instruments was, in the necessity analysis, part of the reason secrecy failed. The reform debate after the judgment is therefore returned to the ground on which the Court insisted it belonged, the design of electoral law by the elected, conducted within boundaries the two holdings draw and policed, when the boundaries are crossed, by the standards this case has settled.

Two parameters of the restored law will frame that debate, and both are worth naming because the judgment leaves them exactly where it found them. The first is the reporting threshold. Since the Election and Other Related Laws (Amendment) Act of 2003, Section 29C has required parties to report contributions above twenty thousand rupees received from any person or company, with the donor identified; the threshold is the level at which Indian law has long struck its own balance between the voter’s information and the small donor’s privacy, and the bond scheme’s deepest offence against the code was that it suspended that balance for the largest contributions while leaving it in force for the smallest. A legislature revisiting the threshold after the judgment, raising it to meet the retribution concern for modest donors or lowering it to widen the disclosed field, would be legislating inside the judgment’s logic, because the proportionality enquiry the Court conducted is sensitive to where the line is drawn and indifferent only to its abolition at the top. The second parameter is the reporting rhythm. Contribution reports are annual, filed after the financial year, and the bond archive’s disclosure in the middle of an electoral season demonstrated, by contrast, the difference that timing makes to the value of information: information that arrives after the vote informs history, while information that arrives before it informs the vote itself. Whether disclosure should move closer to real time is a question of administration and design on which the judgment lays down no rule, and it may prove, in the long work of reform that follows the case, to be the question on which the informational holding has its most practical sequel.

What survives, and the position for the reader who must act

What survives the judgment is the older architecture of political finance law, restored to the shape it had before the Finance Act of 2017. Under the Representation of the People Act, a party’s obligation to report contributions above the statutory threshold, and the connected disclosure regime that the pillar account of the 1951 Act describes, applies to contributions by every mode of transfer, the bond exemption having fallen. Under the Income Tax Act, a party’s exemption again depends on maintaining the prescribed records of its contributors. Under the Companies Act, a company’s contribution is again capped by reference to its average net profits and is again disclosable party-wise in its accounts. Contributions by cheque, draft and electronic transfer remain lawful, cash contributions remain confined by the limits the Finance Act itself had tightened and which the judgment did not disturb, and the electoral trust remains available as the pooling vehicle the Court identified as the less restrictive alternative.

What does the electoral bonds judgment mean for a company planning a political contribution?

A company may contribute only within the restored ceiling of seven and a half per cent of its average net profits over the three preceding financial years, with the authorisations Section 182 requires, and must disclose in its profit and loss account the amount given to each party. A company without profits has no headroom to contribute at all.

For the compliance reader that answer carries the practical core of the case, and it illustrates why the equality holding outlasts the disclosure story. The general account given here is legal information about the position the judgment restored; it is not advice on any individual company’s contribution, for which the company’s own records, profit computation and board process are decisive. The researcher and the student will take a different payload: a worked example of structured proportionality operating on two rights at once, in which the State lost not because its objective was condemned but because a less restrictive instrument lay ready to hand in the statute book and had lain there for a decade.

The judgment in the longer doctrinal line

No constitutional judgment stands alone, and this one is best placed as the meeting point of three lines of authority that had developed separately for decades. The first line is the informational line, running from the candidate disclosure cases of 2002 and 2003 through the right to know jurisprudence of the intervening years: its premise is that the franchise is a form of expression and that expression without information is a formality. The bonds case extends that line from the person of the candidate to the finances of the party, and in doing so it completes a logic that the earlier cases had stated but not exhausted. A voter choosing between candidates chosen by parties, financed by interests the voter cannot see, was exercising the form of the right the earlier cases protected while being denied its substance, and the judgment says so in terms.

The second line is the proportionality line, running from the structured test adopted for privacy in Puttaswamy through its applications to speech restrictions and surveillance-adjacent measures in the years that followed. The bonds case is that line’s most instructive modern application, for two reasons. It shows the test operating on legislation whose objective was conceded to be legitimate in the abstract and yet failing, because the objective could not be translated into a permitted ground of restriction and because the tailoring failed against a real alternative. And it refines the test itself, by carrying the double proportionality variant from the concurrence jurisprudence of an earlier decade into a Constitution Bench holding, establishing that where the State defends a restriction as a balance between two rights, both sides of the balance must survive structured scrutiny. Future rights litigation in fields far from elections, wherever the State pleads one citizen’s right against another’s, will be argued in the grammar this judgment stabilised.

The third line is the equality line: Shayara Bano’s recognition of manifest arbitrariness as a ground for striking down legislation, and its gradual absorption into the Court’s equality practice. Before the bonds case, the doctrine’s applications had been episodic and its status periodically doubted, with some Benches preferring the discipline of the classification test. The bonds judgment is the doctrine’s most consequential modern deployment, because it is the ground on which a fiscal amendment in a Finance Act, a species of provision that courts historically approach with reluctance, was struck down. The analysis did not ask whether the legislature had a conceivable reason for freeing corporate donations; it asked whether treating the company and the citizen as one class, for the purpose of influencing elections, was a determination a Constitution committed to political equality could sustain, and it answered that it could not. Legislatures drafting in the shadow of the case know that the equality review of electoral finance is substantive, that the profit-linked ceiling is not a regulatory detail but the visible form of a constitutional distinction, and that the distinction between the citizen who gives from conviction and the company that gives from interest will structure the validity of whatever scheme is attempted next.

The verdict of this article is the namable claim with which it began. The electoral bonds judgment is two judgments wearing one citation. The first, on information, holds that the funding of political parties is part of what a voter is entitled to know before voting, and it produced the scheme’s invalidation and the disclosure of the bond record. The second, on equality, holds that corporate political money is constitutionally unlike a citizen’s gift and may not be licensed as though it were the same thing, and it will constrain every future scheme of corporate funding whether or not that scheme is transparent. Accounts of the case that merge the two grounds misdescribe both: the first is about secrecy, the second is about the character of the contributor, and only the second binds the legislature that tries again.

Frequently Asked Questions

Q: What did the Supreme Court decide about electoral bonds?

A Constitution Bench of five judges held unanimously on 15 February 2024 that the Electoral Bond Scheme of 2018 was unconstitutional. The Court struck down the scheme and the amendments that supported it: the exemption of bond contributions from party reporting under the Representation of the People Act, the relief from record-keeping under the Income Tax Act, and the Companies Act changes that had removed the ceiling on corporate contributions and the duty of party-wise disclosure. It directed the State Bank of India to stop issuing bonds and to disclose the record of purchases and encashments to the Election Commission for publication.

Q: Why did the scheme violate the right to information?

The Court held that the right to information under Article 19(1)(a) extends beyond the antecedents of candidates to the funding of political parties, because parties select, finance and discipline the candidates a voter chooses between, and because information about a party’s funding is essential to the effective exercise of the vote, including the detection of possible quid pro quo arrangements. By sealing the identity of every bond purchaser and exempting bond contributions from the reporting regime, the scheme deprived the voter of that information entirely, and the Court held that the infringement could not be justified as a proportionate balance against donor privacy.

Q: Why was the removal of the corporate donation cap struck down?

The ceiling in Section 182 of the Companies Act, which had limited a company to seven and a half per cent of its average net profits, was deleted by the Finance Act of 2017. The Court struck the deletion down as manifestly arbitrary under Article 14. It reasoned that a company’s contribution and an individual’s contribution are not alike: the individual’s expresses political conviction, while the company’s is in its nature a business transaction made with an expectation of influence. Treating the two alike, without any ceiling, and admitting even loss-making companies that could serve as shells for outside money, failed the equality guarantee and infringed free and fair elections.

Q: Did the Court accept the black money justification?

No. The Union defended the scheme as a measure to curb black money in political finance by moving contributions into banking channels. The Court answered at two levels. First, curbing black money is not among the grounds on which the freedom of speech and expression may be restricted under Article 19(2), so it could not justify infringing the voter’s right to information however worthy the objective. Second, the scheme was not tailored to the objective in any event: channelling money through banks did the anti-cash work, and the additional anonymity was unnecessary to it, as the electoral trust mechanism demonstrated.

Q: What electoral bond disclosure did the Court order?

The Court directed the State Bank of India to submit to the Election Commission the details of every electoral bond purchased since the interim order of 12 April 2019, including the date of purchase, the name of the purchaser and the denomination, and the details of every bond encashed by a political party, including the date of encashment and the denomination. The bank was given until 6 March 2024, and the Commission was directed to publish the information on its official website within one week of receipt, by 13 March 2024. Bonds still within their validity period and unencashed were to be returned for refund.

Q: What happened when the bank sought more time for electoral bond disclosure?

The State Bank of India applied for an extension until 30 June 2024, explaining that purchase records and encashment records had been kept in two separate silos to preserve confidentiality and that matching them would take months. The Constitution Bench dismissed the application on 11 March 2024, holding that the information required was information the bank held and that its internal arrangements could not defeat the judgment’s directions. The bank was ordered to furnish the details by the close of business on 12 March 2024, and the Election Commission to publish them by 15 March 2024, with an affidavit of compliance from the bank’s Chairman and Managing Director.

Q: What did the electoral bond data reveal after disclosure?

The disclosure placed two categories of record in the public domain for the period from 12 April 2019 to 15 February 2024: purchases, naming the purchaser, the date and the denomination of each bond, and encashments, naming the receiving party, the date and the denomination. Once the Court directed disclosure of the unique alphanumeric number borne by each bond, the two records could be matched bond by bond. The data establishes that purchases and encashments occurred in stated amounts on stated dates. It does not, without more, establish the purpose of any donation or any understanding between a purchaser and a party.

Q: What political funding law remains after the judgment?

The judgment restored the position that stood before the Finance Act of 2017. Parties must report contributions above the statutory threshold under Section 29C of the Representation of the People Act whatever the mode of payment, and must keep the contributor records the Income Tax Act requires for their exemption. Companies may contribute only within the restored ceiling tied to average net profits, and must disclose the amount given to each party in the profit and loss account. Contributions by cheque, draft and electronic transfer remain open, cash contributions remain restricted, and electoral trusts continue as a lawful vehicle for pooled donations.

Q: Who were the petitioners who challenged the electoral bond scheme?

The lead petition was Writ Petition (Civil) No. 880 of 2017, filed by the Association for Democratic Reforms, an electoral reform organisation whose earlier litigation had produced the candidate disclosure decisions of 2002. Connected petitions in the batch included those of Common Cause, a public interest organisation, and the Communist Party of India (Marxist), which participated as a political party subject to the funding regime it was challenging. The respondents were the Union of India and the Election Commission of India, and the Election Commission’s submissions supported disclosure rather than the defence of anonymity mounted by the Union.

Q: What standard of review did the Court refuse to apply to the electoral bonds scheme?

The Court refused to test the scheme and the Finance Act amendments on the deferential parameters applicable to economic policy. The Union had argued for the latitude that decisions such as Swiss Ribbons afford to economic legislation. The Bench held that contributions to political parties relate to the democratic polity, the citizen’s right to know and accountability in democracy, and that the Union had itself classified the package as electoral reform. To import the deference owed to trade and commerce legislation into electoral law, Khanna J added, would dilute the principle of free and fair elections, so the Bench undertook an in-depth rights review instead.

Q: Does a political donor have a constitutional right to keep a contribution private?

The Court recognised a right to informational privacy that includes privacy over political affiliation: which party a citizen supports is protected, and the State cannot compel its disclosure at large. It held, however, that a financial contribution is not protected in the same way. A contribution transfers resources into the electoral contest, may carry an expectation of return, and had always been disclosable above the statutory threshold under the pre-existing law. Applying the double proportionality standard, the Court held that whatever privacy interest a contributor holds in a donation is outweighed, for contributions of this character, by the voter’s right to information, and cannot sustain total anonymity.

Q: What happened to electoral bonds that were bought but never encashed?

Electoral bonds carried a validity of fifteen days from issue, after which an unencashed bond was worthless to the party holding it. The judgment directed that bonds still within their validity period on the date of the judgment, and not yet encashed by the political party, be returned by whoever was in possession of them, the party or the purchaser, to the issuing bank. On return of a valid bond, the bank was to refund the amount to the purchaser’s account. Bonds whose validity had already expired before the judgment received no direction and no refund under the operative part.

Q: Did the Supreme Court decide whether the Finance Act 2017 was a valid money bill?

No. The petitioners had challenged the routing of the amendments through a Money Bill under Article 110 of the Constitution, which confines the Rajya Sabha to a recommendatory role. The Bench expressly declined to examine that question, recording that the scope of Article 110 stands referred to a seven-judge Bench in Rojer Mathew v South Indian Bank and is sub judice. The amendments were therefore tested and struck down on fundamental-rights grounds, on Articles 19(1)(a) and 14, and the distinct question whether electoral changes of this kind could competently be enacted through the Money Bill route remains open.

Q: Did Justice Khanna’s opinion strike down any provision the main judgment did not?

Yes. Khanna J concurred in every conclusion of the lead opinion but reasoned partly differently, applying proportionality in a form that he held subsumes the manifest arbitrariness test for the Companies Act amendment. On the Reserve Bank of India Act, he went further than the lead opinion. Section 31(3), inserted by the Finance Act to permit the bonds to be issued as instruments payable to bearer on demand, was held unconstitutional and struck down in his opinion. The lead opinion struck down the scheme that the provision existed to serve without resting a separate holding on the provision itself.

Q: Why did the Court order disclosure of the unique bond numbers after the first release of data?

The data the bank furnished on 12 March 2024 and the Commission published on 14 March arrived as two unconnected datasets: purchases on one side, encashments on the other, with no shared field. A reader could not tell which purchase corresponded to which encashment. On 18 March 2024 the Bench held that the original directions, which specified fields by way of inclusion, required complete disclosure of all details in the bank’s possession, and that the bank was not to be selective. It directed disclosure of the unique alphanumeric and serial numbers of the bonds, with the Chairman’s affidavit of completeness, and the bank complied on 21 March 2024.

Q: Can a loss-making company still contribute to a political party after the judgment?

On the restored text of Section 182 of the Companies Act, a company that is not a government company may contribute up to seven and a half per cent of its average net profits during the three immediately preceding financial years. A company without profits has no amount computed by that formula available to it, which is one of the consequences the Court relied on in striking down the deletion of the ceiling: the amended provision had allowed loss-making companies, including shells without genuine business, to contribute without limit. The restored ceiling therefore closes the channel the equality holding identified, subject to the section’s authorisation requirements.

Q: Why does the ordered disclosure begin in April 2019 rather than at the start of the scheme?

The scheme began operating in March 2018, but the disclosure directions run from 12 April 2019. That is the date of the Court’s interim order in the challenge, by which political parties were directed to submit particulars of bond contributions to the Election Commission in sealed cover. From that date, purchasers, parties and the bank were on express notice that the details were being collected under the Court’s authority. When the Court was later asked to extend the disclosure backwards to the scheme’s first year, it declined, treating the cut-off in the judgment as a conscious choice of the Constitution Bench rather than an error correctable in miscellaneous proceedings.

Q: What part did electoral trusts play in the Court’s reasoning?

Electoral trusts supplied the less restrictive alternative on which the necessity stage of the proportionality analysis turned. A trust recognised under the Income Tax Act collects contributions through banking channels only, keeps records of its contributors and reports contributions received and disbursed to the Election Commission, yet because it pools several donors and distributes to several parties, the link between a particular donor and a particular party is not disclosed. The Court held that the trust mechanism achieved the objective of curbing black money in electoral financing while surrendering far less of the voter’s information than total anonymity, so the Union could not show that the scheme’s confidentiality was necessary.

Q: Were political parties ordered to return the donations they had received through electoral bonds?

No. The only refund direction in the judgment concerns bonds that were still within their fifteen-day validity and had not been encashed: those were to be returned to the issuing bank and the amount refunded to the purchaser’s account. Donations that parties had received and encashed during the scheme’s operation were not ordered to be disgorged, no party was directed to repay any donor, and the judgment attached no penalty or disqualification to any disclosed receipt. The operative consequences were prospective cessation of the scheme and retrospective disclosure of the record, not restitution of the money.

Q: What are the main criticisms of the electoral bonds judgment?

Four criticisms recur. The right to funding information is said to be a judicial extension of Article 19(1)(a) beyond its candidate-centred origins. Donor anonymity is said to protect contributors against retaliation, an interest the judgment is argued to undervalue, particularly for smaller donors. The least-restrictive-means enquiry is said to substitute the Court’s policy design for the legislature’s in a field of competing democratic values. And retrospective disclosure is said to be unfair to donors who gave under a statutory promise of confidentiality. The judgment answers each: parties finance and control candidates, anonymity ran against voters rather than the State, proportionality is the settled rights standard, and the 2019 interim order put donors on notice. A fifth criticism, of timing on the eve of a general election, is answered by the chronology: the case had been pending since 2017 and judgment was reserved months before any election schedule was announced.