The collegium NJAC contestation is usually told in one of two ways, and both ways lose the thing that matters most about it. The first telling moves case by case, from one judgment to the next, as if the story were a sequence of courtroom events with dates and citations but no connecting structure. The second telling moves government by government, treating the National Judicial Appointments Commission of 2014 as one administration’s reform project and its defeat in 2015 as that administration’s setback, as if the architecture of judicial appointments were ordinary party politics conducted by other means. This article defends a different reading, and states it as a claim that can be tested against the texts. The collegium-NJAC contestation is a four-phase architectural trajectory; the architectural content of each phase operates institutionally beyond the cases that produced it. In the first phase, running from 1993 to 1998, the Supreme Court of India built the collegium through two judgments, the Second Judges Case of 6 October 1993 and the Third Judges Case of 28 October 1998. In the second phase, running from 1998 to 2014, that architecture operated as the working method of appointment to the higher judiciary and drew sustained criticism from commissions, scholars, governments and judges themselves. In the third phase, running from 2014 to 2015, Parliament and the state legislatures attempted a wholesale substitution, enacting the Constitution (Ninety-ninth Amendment) Act and the National Judicial Appointments Commission Act. In the fourth phase, opening on 16 October 2015, a Constitution Bench struck the substitution down by four votes to one, restored the collegium, and set in motion a negotiation over a revised Memorandum of Procedure that remained unfinished for years afterwards. Each phase is defined not by its leading case or its leading minister but by the allocation of appointment power it put into operation, and each phase handed that allocation, intact or contested, to the phase that followed it.

Articles 124 and 217: the constitutional text the contestation inherited

Every argument in the contestation begins with two provisions of the Constitution of India, and the provisions repay exact reading because so much was later built on so few words. Article 124(2) provides that every judge of the Supreme Court shall be appointed by the President by warrant under hand and seal after consultation with such judges of the Supreme Court and of the High Courts as the President may deem necessary, and it adds a proviso that in the appointment of a judge other than the Chief Justice, the Chief Justice of India shall always be consulted. Article 217(1) makes the parallel provision for the High Courts: appointment by the President after consultation with the Chief Justice of India, the Governor of the state, and, for a judge other than the Chief Justice of that High Court, its Chief Justice. The word the text uses is consultation. The word the text does not use is concurrence. Between those two words the whole contestation lives. The Constituent Assembly had considered the allocation directly. On 24 May 1949, debating the draft provision that became Article 124, B. R. Ambedkar rejected two opposite designs in the same speech. Against a proposal that would have required the concurrence of the Chief Justice as a condition of appointment, Ambedkar argued that the Chief Justice was a man with all the failings, all the sentiments and all the prejudices that common people have, and that to give him a veto over appointments was to transfer authority away from the executive in a manner the Assembly should not accept. Against the opposite design of unfettered executive discretion, the Assembly’s settlement kept mandatory consultation with the senior judiciary as a structural feature of the process. The founding settlement, in other words, was a middle position held deliberately: the executive appointed, the judiciary was consulted, and neither branch held the whole power. Readers who want the founding frame set out at length can follow the constitutional rights framework through the drafting history, because the appointments question was argued there as part of the larger question of how an independent judiciary was to be staffed without being captured.

The middle position held, as practice, for roughly three decades, and its operation is part of the evidence. From 1950 onwards the convention grew that the senior-most judge of the Supreme Court would succeed as Chief Justice of India, and appointment recommendations moved through consultation between the government of the day and the Chief Justice in a manner that produced few public ruptures. The ruptures, when they came, came over the Chief Justiceship itself and they matter to the appointments story because they taught the judiciary what executive primacy could mean in practice. On 25 April 1973, the day after the judgment in Kesavananda Bharati v State of Kerala, the government of Prime Minister Indira Gandhi appointed Justice A. N. Ray as Chief Justice of India over three senior colleagues, Justices J. M. Shelat, K. S. Hegde and A. N. Grover, who resigned in protest. On 28 January 1977 the same government superseded Justice H. R. Khanna, the lone dissenter in the habeas corpus judgment of the Emergency period and the senior-most judge, and appointed Justice M. H. Beg as Chief Justice instead. The second supersession connects the appointments question to the institutional-architectural-failure-pattern framework, because the judge passed over had been passed over for a dissent that defended personal liberty against the executive at the executive’s most powerful moment. A bench and bar that had watched seniority discarded twice within four years did not need a theory to understand the stakes of the consultation clause. When the question of its meaning reached a Constitution Bench in 1981, it arrived carrying that history.

What does Article 124 actually say about appointing a Supreme Court judge?

Article 124(2) says a Supreme Court judge is appointed by the President after consultation with such judges of the Supreme Court and the High Courts as the President deems necessary, and that the Chief Justice of India must always be consulted for appointments other than his own. The text says consultation, not concurrence, and names no collegium.

The first authoritative reading of that text came in S. P. Gupta v Union of India, decided on 30 December 1981 and known afterwards as the First Judges Case. A seven-judge bench considered, among other questions, whether the consultation required by Articles 124 and 217 obliged the government to accept the opinion of the Chief Justice of India. The majority, in the leading judgment of Justice P. N. Bhagwati, held that it did not. Consultation, the majority reasoned, did not mean concurrence; the executive could depart from the Chief Justice’s opinion for cogent reasons, and the primacy of the appointment power lay with the executive, which was accountable to Parliament and through Parliament to the electorate in a way the judiciary was not. The holding rested on a textual argument and a democratic argument together: the framers had considered concurrence and rejected it in the debate of 24 May 1949, and an unelected judge’s veto over appointments sat uneasily with responsible government. For twelve years that reading governed. Appointments were made by the executive after consultation, the Chief Justice’s view carried great weight by convention, and the legal position was that the last word belonged to the government. The four-phase trajectory this article defends begins when a nine-judge bench took the same text and read it the other way, and it is essential to the trajectory reading that the First Judges Case is treated as the baseline the first phase displaced rather than as a mere preliminary. The collegium was not found in the Constitution in 1950 and applied in 1993. It was constructed in 1993 out of the same words that in 1981 had been held to mean something close to their opposite, and the fact of construction, acknowledged or not, shapes everything the later phases did with the building.

Phase One: the Second Judges Case of 1993 builds the collegium

On 6 October 1993 a nine-judge bench of the Supreme Court delivered judgment in Supreme Court Advocates-on-Record Association v Union of India, the Second Judges Case. The bench had been constituted because the correctness of S. P. Gupta had been doubted, and the reference squarely raised whether the opinion of the Chief Justice of India in appointments held primacy. The majority, speaking principally through Justice J. S. Verma, held that it did. The word consultation in Article 124 was read, in the light of the constitutional purpose of an independent judiciary, as vesting the decisive voice in the judiciary’s institutional opinion, formed by the Chief Justice of India not as an individual but in consultation with his senior colleagues. The reasoning ran in steps that later phases would quote back at each other. Judicial independence was a structural commitment of the Constitution. The appointment process was the point at which independence could be secured or compromised before a judge ever sat. The executive, a litigant before the courts in a vast share of their business, could not hold the decisive voice in choosing the judges before whom it would appear. The Chief Justice, by contrast, was best placed to assess the merit and suitability of candidates, and his opinion, formed plurally with the two senior-most judges of the Supreme Court for appointments to that court, was to be given primacy, so that a recommendation so formed would ordinarily bind the executive. The judgment also addressed the Chief Justiceship itself, holding that seniority should govern succession save in exceptional circumstances, a direct answer to the supersessions of 1973 and 1977. The minority, including Justice A. M. Ahmadi, warned against reading concurrence into a text that said consultation, which is the first recorded form of the objection that would follow the collegium for the rest of the trajectory: that the architecture had been built by interpretation past the point the words could bear.

The architectural content of the 1993 judgment is what distinguishes it from a mere reversal of S. P. Gupta, and it deserves to be stated as design rather than as doctrine. Before 1993, the appointment process had a single decision-maker, the executive, advised by consultation. After 1993, the process had a recommending body inside the judiciary whose recommendation carried binding force in the ordinary course, an executive whose role was reduced to processing, objecting on stated grounds, and returning a recommendation once for reconsideration, and a rule that a recommendation reiterated after reconsideration had to be appointed. The judgment converted a bilateral consultation into a sequenced procedure with a judicial first move and a judicial last word. It also converted the Chief Justice from a consultee into the head of a collective, because the opinion that bound was not his personal opinion but the opinion of the Chief Justice formed with his two senior-most colleagues, a grouping the judgment described and the 1998 opinion would enlarge. That is why the case-by-case telling understates the judgment. Treated as an event, the Second Judges Case is a famous reversal. Treated architecturally, it is a transfer of the appointment power’s centre of gravity from one branch to another, executed by reinterpretation, and every later phase of the trajectory is a phase in the life of that transfer: its operation, its criticism, the attempt to reverse it by amendment, and the judgment that held the amendment itself unconstitutional. The political setting of the transfer is worth recording without allowing it to swallow the analysis. The judgment was delivered in the India of the post-1991 institutional-political context, in a decade when coalition governments of shifting composition held office at the centre and the executive’s capacity to impose its will on institutions was at a low ebb. A weaker executive did not write the judgment, and the bench’s reasoning was framed in the general language of judicial independence rather than in any comment on the government of the day. But the trajectory reading notes the setting because architecture built in one distribution of political power is then operated, criticised and attacked in others, and the collegium’s first phase was completed in the same decade’s conditions.

How many judges form the collegium for a Supreme Court appointment?

For a Supreme Court appointment, the collegium is the Chief Justice of India and the four senior-most judges of the Supreme Court. The Third Judges Case fixed that five-member composition in 1998, enlarging the three-member grouping, the Chief Justice with his two senior-most colleagues, described by the Second Judges Case in 1993.

The completion of the first phase came through an unusual instrument. In 1998 the President of India, K. R. Narayanan, made a reference to the Supreme Court under Article 143 of the Constitution, Special Reference No. 1 of 1998, asking the court to clarify the operation of the 1993 judgment on nine questions, including the composition of the collegium, the weight of a Chief Justice’s opinion when his colleagues disagreed with him, the role of seniority in appointments and transfers, and the justiciability of the process. A nine-judge bench answered unanimously on 28 October 1998, in an opinion delivered by Justice S. P. Bharucha, and the answers converted the 1993 holding into a working constitution for appointments. The collegium for Supreme Court appointments would be the Chief Justice and the four senior-most judges. For High Court appointments, the Chief Justice of India would act with the two senior-most judges of the Supreme Court, and the recommendation itself would originate in the High Court’s own collegium of its Chief Justice and two senior-most judges. The opinion of the Chief Justice in his individual capacity carried no special weight; the binding opinion was the collegium’s. Consultation had to be with the plurality the opinion specified, in writing, and a recommendation made without that consultation was not binding on the government. Seniority among judges of a High Court was a legitimate and weighty consideration for elevation, though not the sole one. Judicial review of appointments was confined to the narrow ground that the prescribed consultation had not taken place or that the appointee lacked eligibility. Phase One therefore ends in 1998 not with a second transfer of power but with a codification: the architecture acquired a fixed membership, a written procedure, and defined limits of review, all of it resting on two judgments and an advisory opinion rather than on any statute or amendment. No legislature had voted for the collegium. No constituent body had designed it. Its legitimacy, its defenders said, was the legitimacy of the Constitution itself as authoritatively interpreted. Its fragility, its critics said from the beginning, was the same fact stated from the other side.

Phase Two: the collegium in operation, 1998 to 2014

The second phase is the longest of the four and the easiest to misread, because a system in operation generates fewer famous dates than a system in crisis. Between the advisory opinion of 28 October 1998 and the constitutional amendment of 31 December 2014, every appointment to the Supreme Court of India and to the twenty-four High Courts then in existence was made through the collegium architecture. The procedure settled into a recognisable sequence. For a High Court, the Chief Justice of that court, acting with his two senior-most colleagues, initiated names, drawing on the bar and the district judiciary; the state government and the Governor recorded views; the file moved to the Union government, which obtained the assessments of the Intelligence Bureau on the candidates; the Supreme Court collegium considered the proposal with the views of consultee judges who had served in that High Court; and the recommendation, once made, went to the President for the warrant of appointment. For the Supreme Court itself, the five-member collegium initiated names, ordinarily from among the Chief Justices and senior judges of the High Courts, with seniority across the High Courts weighed alongside merit and regional and institutional representation. The executive retained two real but bounded powers: it could return a recommendation once with its objections recorded, and it could sit on a file. If the collegium reiterated a returned name, the appointment followed as a matter of constitutional obligation under the 1993 and 1998 readings. The first power was used sparingly and the second, delay, was used more often than any published record acknowledged, a point that matters later in the trajectory when delay itself became the executive’s principal instrument.

Judged as output, the architecture filled the courts. The Supreme Court worked at or near its sanctioned strength of thirty-one judges for much of the phase, and elevations followed a broadly seniority-ordered pattern across High Courts that gave the system a predictability its defenders prized. The phase also produced appointments of acknowledged distinction and a line of Chief Justices whose succession by seniority was automatic and uncontroversial, in pointed contrast to 1973 and 1977. Judged as process, the architecture was a closed one. The collegium published no reasons. Its deliberations were not recorded in any document a citizen could obtain. Criteria for selection existed in the form of convention and of the judgments’ own statements about merit and seniority, but no published framework told a candidate, a government or a reader how competing claims had been weighed in a given case. The executive’s objections, when it made them, were made in confidence on the file. The result was an appointment system whose outputs were public and whose reasoning was not, and Phase Two is defined in the trajectory by the tension between those two facts. The system worked, in the sense that courts were staffed and no supersession crisis recurred. The system was also, in a sense its critics developed year by year, unaccountable: no institution outside the judiciary could test a choice, and no institution inside it explained one. The four-phase reading insists on holding both halves of that description together, because the third phase becomes unintelligible if Phase Two is remembered only as failure, and the fourth phase becomes unintelligible if Phase Two is remembered only as success.

The criticism that accumulated across the phase came from identifiable quarters and took identifiable forms, and the trajectory reading requires the criticism to be stated in its strongest versions rather than in summary. The first form was textual, and it never went away. From Justice Ahmadi’s minority reasoning in 1993 onwards, lawyers and scholars argued that consultation had been made to mean its opposite, that a power the Assembly had debated and declined to grant the Chief Justice had been granted by interpretation, and that the resulting architecture had no democratic pedigree of any kind: not enacted, not ratified, not even recommended by any commission before it was created. The second form was institutional, and it drew on the system’s own record. Appointments were alleged to favour the familiar: relatives of judges, juniors of senior advocates, candidates from a narrow band of chambers and courts. The phrase uncle judges entered the public vocabulary of the debate to describe the perception, documented in bar resolutions and in the press of the period, that kinship with sitting or former judges eased the path to a High Court bench. The third form was voiced from inside the house. Serving and retired judges said in terms that the collegium’s choices could not be defended because they could not be examined. Justice Ruma Pal, a former judge of the Supreme Court, described the process in a lecture in 2011 as one of the best-kept secrets in the country, and observed that consensus within the collegium was sometimes reached through trade-offs among its members that had nothing to do with the candidates under discussion. A former Chief Justice of India, Justice J. S. Verma, the author of the 1993 majority judgment itself, said in his later years that the judgment had been misconceived in operation and that appointments under it had not matched its expectations. When the principal author of an architecture becomes one of its critics, the criticism cannot be dismissed as the complaint of the branch that lost power under it.

Two official bodies converted the gathering criticism into reform proposals during the phase, and their reports are the documentary spine of Phase Two because they show what the political system, speaking through its own commissions, thought should replace or repair the collegium before any government attempted a replacement. The National Commission to Review the Working of the Constitution, chaired by Justice M. N. Venkatachaliah, a former Chief Justice of India, was appointed in 2000 and submitted its report on 31 March 2002. The commission recommended a National Judicial Commission for appointments, composed of the Chief Justice of India as chair, two senior judges of the Supreme Court, the Union Minister for Law and Justice, and one eminent citizen nominated by the President in consultation with the Chief Justice. The design is instructive: a judicial majority, an executive member, and a citizen member, which is a sharing of the appointment power in proportions no later proposal would repeat exactly. The Law Commission of India returned to the subject in its Two Hundred and Fourteenth Report of 2008, titled Proposal for Reconsideration of Judges Cases and Appointment of Judges. That report took the more radical position that the three judges cases had been wrongly decided as a matter of constitutional interpretation, noted the original meaning of consultation as the Assembly had used it, and recommended that the law be restored toward executive primacy with judicial consultation, whether by a larger bench reconsidering the cases or by constitutional amendment. Between the two reports, the reform spectrum of Phase Two is fully visible: repair the architecture by broadening its membership, as Venkatachaliah proposed, or reverse the architecture by revisiting its foundation, as the Law Commission proposed. What no official body of the phase proposed was the status quo defended as ideal. The collegium operated for sixteen years under a standing verdict, delivered by the state’s own commissions, that it ought to be changed. The change, when it came in 2014, came with a speed and a breadth of political support that Phase Two’s documentary record explains.

Governments of the phase tested the architecture at its edges without challenging its foundation, and the episodes are evidence of how the allocation of power worked when it was pressed. In 2004 the collegium’s recommendation concerning a Madras High Court appointment, and the government’s prolonged hesitation over a candidate against whom the Intelligence Bureau had recorded adverse material, produced a public exchange in which the court insisted on its primacy and the government of Prime Minister Manmohan Singh eventually processed the file with its objections noted rather than sustained. The pattern repeated in different registers: objections on integrity grounds were the executive’s strongest card, and where the material was concrete the collegium itself sometimes let a name drop; objections that looked political were the executive’s weakest, and reiteration answered them. Transfers of High Court judges under Article 222, which the 1998 opinion had also brought within the collegium’s consultative discipline, generated their own friction, since a transfer could be experienced by the judge transferred as a penalty imposed without a hearing. The phase’s deepest structural fact, however, was a vacancy problem that neither branch owned. High Court vacancies ran persistently at a quarter to a third of sanctioned strength across the phase, and each side attributed the shortfall to the other: governments pointed to collegiums that sent too few names too slowly, and the judiciary pointed to files that waited months in the Ministry of Law and Justice. The data supported both accusations in different states at different times, which is precisely what an architecture without published procedure produces: a failure that cannot be assigned, because the process that failed cannot be inspected. That unassignability is the bridge to Phase Three. A system whose defects could be debated but not located invited a remedy aimed at the system as a whole rather than at any located defect, and the remedy that arrived in 2014 was correspondingly total.

The measurable record of Phase Two: vacancies, composition and delay

An architectural reading earns its keep by submitting to measurement, and Phase Two’s record can be measured on three axes that the debate itself chose: how full the courts were, who the appointees were, and how long appointment took. On fullness, the Department of Justice’s own published figures across the phase show High Court working strength running persistently below sanctioned strength, with vacancies commonly between a quarter and a third of the total sanctioned posts of roughly nine hundred across the twenty-four High Courts. The Supreme Court, by contrast, sat close to its sanctioned strength of thirty-one for most of the phase, dipping sharply only when clusters of retirements coincided with slow processing, as in the months before the 2015 challenge froze the process altogether. The contrast between the two levels is itself an architectural datum. The Supreme Court collegium choosing for its own court, from a small and visible pool of High Court Chief Justices, produced a near-full bench. The distributed High Court process, in which initiation, state consultation, intelligence reporting, central processing and Supreme Court settlement each held a segment of the file, produced chronic shortage, and each segment’s holder could truthfully deny that the shortage was made in that segment. Reform proposals of the phase that targeted only the final stage, the collegium’s settlement, were therefore aimed at the segment least responsible for the measured failure, a point the Law Commission’s 2008 report made in its own way by addressing the whole chain of consultation rather than its last link.

On composition, the phase’s appointees were drawn overwhelmingly from two channels, the senior district judiciary and the bar, in proportions the Memorandum of Procedure roughly fixes at two to one, and the public argument fastened on the narrowness of the bar channel in the larger High Courts. The criticism travelled under the phrase uncle judges, and its evidentiary base was assembled by bar associations and journalists from appointment lists: sons, daughters, nephews and former juniors of judges and leading counsel appearing among the elevated with a frequency that chance and merit together strained to explain. The collegium’s defenders answered that a profession organised in chambers and families will always produce candidates known to its selectors, that knowledge of a candidate’s work over decades is information rather than favour, and that no appointment of the phase was set aside by any court for want of eligibility. Both statements can be held together, and the trajectory reading holds them together by locating the defect correctly: the architecture generated no record by which a suspicious appointment could be distinguished from a well-informed one, so every appointment of a judge’s relative stood condemned by the absence of reasons and none could be acquitted by their presence. Opacity did not merely hide favouritism where it existed. It manufactured the appearance of favouritism where it did not, which is why judges who had themselves benefited from informed selection, Justice Ruma Pal among them, attacked the secrecy rather than the selections.

On delay, the phase’s record is a record of files. The Memorandum prescribed weeks for each stage; the practice ran to months and, in contested cases, to years. The executive’s contribution to delay was the holding of files after the collegium had spoken, a practice governments defended as verification and the judiciary experienced as a pocket veto that the judgments had not authorised, since the power to return a name once, with objections, was the only brake the 1993 and 1998 readings conceded. The judiciary’s contribution was slower initiation in several High Courts, where collegiums waited for vacancies to accumulate or for seniority disputes among district judges to settle before sending names upward. A candidate’s own position in the sequence captures the human arithmetic: a district judge or advocate recommended at fifty-two and appointed at fifty-six after the file’s travels would serve a shortened tenure, reach a High Court Chief Justiceship, if at all, near its end, and enter the zone of consideration for the Supreme Court with the seniority clock already against the appointment. Delay therefore did not merely slow the system’s staffing. It silently edited the pool from which the Supreme Court itself would later be chosen, favouring early entrants and penalising the service channel through which most of the country’s trial judiciary ascends. When Phase Three’s designers promised a commission with statutory timelines, and when Phase Four’s collegium and government traded figures on pending files year after year, both were arguing over this measured inheritance of Phase Two, and the four-phase reading is the only one of the three rival readings that has a place to put it: not doctrine, not party strategy, but the operating record of an allocation of power.

The reform attempts inside Phase Two and why they failed

It is a mistake in the case-by-case telling to treat 2014 as the first attempt to replace the collegium, and the trajectory reading corrects it by recording the attempts that failed inside Phase Two itself. A Constitution Amendment Bill to establish a National Judicial Commission was introduced in the Lok Sabha in 2003 by the government of Prime Minister Atal Bihari Vajpayee, carrying the Venkatachaliah design into legislative form. The bill lapsed with the dissolution of the Lok Sabha in 2004 and was not revived by the succeeding government. The Judicial Appointments Commission Bill of 2013, introduced by the second government of Prime Minister Manmohan Singh, proposed a commission headed by the Chief Justice of India with judicial, executive and eminent-person membership, and it too failed to become law before the general election of 2014 changed the government. The failures share a structure worth naming. In each case a government of one political colour proposed a judicial-majority commission with executive and citizen participation; in each case the proposal died of legislative time and political priority rather than of principled defeat on the floor; and in each case the bar and the bench were divided, with senior advocates arguing both that reform was overdue and that the proposed designs diluted primacy past the point of safety. The significance for the trajectory is that by 2014 the idea of a commission was not an invention of the incoming government of Prime Minister Narendra Modi. It was a twenty-year-old proposal with a commission report behind it, two lapsed bills behind it, and support that crossed party lines in principle. What the incoming government added was arithmetic and speed: a single-party majority in the Lok Sabha of a kind not seen since 1984, a coalition majority adequate to carry a constitutional amendment through both Houses and the state legislatures, and a decision to move the amendment within months of taking office. Any account that reads Phase Three as a sudden partisan seizure of the appointments question has to explain away the Venkatachaliah report, the 2003 bill and the 2013 bill, and the four-phase reading declines to explain them away. They are Phase Two’s own products, and Phase Three is their enactment under new political conditions, which is a different and more demanding thing to assess.

Phase Three: the 99th Amendment and the NJAC Act, 2014 to 2015

The third phase opened with a legislative sequence whose speed was itself part of its meaning. The Constitution (One Hundred and Twenty-first Amendment) Bill was introduced in the Lok Sabha on 11 August 2014 by the Minister of Law and Justice, Ravi Shankar Prasad, alongside a companion ordinary bill to regulate the new commission’s procedure. The Lok Sabha passed both bills on 13 August 2014 and the Rajya Sabha on 14 August 2014, with support that ran across the House: the amendment carried with only a handful of dissenting voices, and members of parties in opposition to the government spoke for it on the ground, recorded across the debates, that the collegium had forfeited confidence. Because the amendment touched the Union judiciary, it required ratification by the legislatures of not fewer than half the states, and sixteen state legislatures ratified it within months. President Pranab Mukherjee gave assent on 31 December 2014, and the measure entered the Constitution as the Constitution (Ninety-ninth Amendment) Act, 2014. The National Judicial Appointments Commission Act, 2014 was enacted alongside it, and both instruments were brought into force on 13 April 2015. A constitutional amendment carried in five months from introduction to assent, on a subject where two previous bills had lapsed unconsidered, is a fact about Phase Two as much as about Phase Three: the demand for substitution had been assembled over sixteen years, and the new majority merely released it.

The architecture the amendment built must be stated by its provisions, because the fourth phase turned on their exact terms. The amendment inserted three articles into the Constitution. Article 124A constituted the National Judicial Appointments Commission and fixed its membership at six: the Chief Justice of India as chairperson; two other senior judges of the Supreme Court next in seniority to the Chief Justice; the Union Minister in charge of Law and Justice; and two eminent persons, to be nominated by a committee consisting of the Prime Minister, the Chief Justice of India and the Leader of Opposition in the Lok Sabha or, where there was no recognised Leader of Opposition, the leader of the single largest opposition party in that House. One of the two eminent persons was to be drawn, by rotation, from the Scheduled Castes, the Scheduled Tribes, the Other Backward Classes, minorities or women. Article 124B charged the commission with recommending persons for appointment as Chief Justice of India, judges of the Supreme Court, Chief Justices of the High Courts and other judges of the High Courts, and with recommending transfers. Article 124C empowered Parliament to regulate the commission’s procedure by law and to empower the commission to make regulations. The NJAC Act then supplied the operating detail. It prescribed the commission’s functions and the manner of selection, including a provision, Section 5(2) for Supreme Court appointments and its High Court parallel, that the commission shall not recommend a person if any two members of the commission do not agree to the recommendation. It required the commission to elicit the views of Governors, Chief Ministers and High Court Chief Justices for High Court appointments, and it directed that selection rest on ability, merit and other criteria the commission specified by regulation. The design’s proportions deserve exact statement, because the argument about them filled the next phase. Of six members, three were judges, one was the serving Law Minister, and two were eminent persons chosen by a committee in which the political executive held two of three voices. Judicial members therefore held half the seats, and any two members together, judicial or not, held a veto over any name.

Who would have sat on the NJAC, and who chose its lay members?

The commission had six members: the Chief Justice of India as chair, the two senior-most Supreme Court judges, the Union Law Minister, and two eminent persons nominated by a committee of the Prime Minister, the Chief Justice of India and the Leader of Opposition in the Lok Sabha. One eminent person was reserved in rotation for underrepresented categories.

The defence of the design, made at its strongest by those who built it, was a defence of shared power against monopoly. The government argued that appointment to the higher judiciary in every other mature democracy involved the executive or a mixed commission, that India’s collegium was a global outlier in excluding any voice outside the judiciary, and that a commission with a judicial chair, judicial parity and a judicial veto through the two-member rule protected independence while ending secrecy. Attorney General Mukul Rohatgi carried the argument into court in 2015 in substantially those terms. Senior advocates who supported the amendment, among them Harish Salve in public commentary of the period, argued that primacy had been confused with exclusivity, and that an appointments process in which the executive and lay members could question, delay and veto but not impose was accountability without capture. The criticism, made at its strongest by those who challenged the amendment, accepted the premise that the collegium needed reform and rejected the instrument. The petitioners in the challenge, led by the Supreme Court Advocates-on-Record Association with Fali S. Nariman among its leading counsel, argued that the composition made judicial independence structurally insecure: the Law Minister sat as a voting member of the body choosing judges before whom his government was the largest litigant; the eminent persons were undefined in qualification, unnamed in constituency, and chosen by a committee the executive dominated; and the two-member veto, neutral on its face, allowed the Law Minister and one eminent person together to block any candidate the judiciary proposed, converting the executive’s old power of delay into a constitutional power of negation. Between those two statements of the design, each coherent and each honestly held by serious lawyers, the third phase closed and the fourth began, because the challenge had been filed while the amendment was still completing ratification, and the court took it up in the spring of 2015 with the collegium in suspense and appointments to the higher judiciary effectively paused for the duration of the hearing.

Phase Four: the Fourth Judges Case of 16 October 2015

The Constitution Bench that heard the challenge consisted of Justices J. S. Khehar, J. Chelameswar, Madan B. Lokur, Kurian Joseph and A. K. Goel. The hearing ran through the spring and summer of 2015, and judgment came on 16 October 2015 in Supreme Court Advocates-on-Record Association v Union of India, the Fourth Judges Case. By a majority of four to one, the bench declared the Constitution (Ninety-ninth Amendment) Act unconstitutional and void, and the NJAC Act with it, and revived the collegium system as it had operated under the Second and Third Judges Cases. The holding, stated by its ratio rather than by its result, ran as follows. The independence of the judiciary is a part of the basic structure of the Constitution, immune from amendment under the doctrine settled in Kesavananda Bharati. Primacy of the judiciary in the appointment of judges is a constituent of that independence, because the power to choose judges is the power to shape the institution that checks every other power. The amendment damaged that constituent in three ways the majority specified: it gave the executive, through the Law Minister’s membership, a voice in selection disproportionate to the consultative role the constitutional scheme allowed; it admitted two eminent persons whose undefined status and executive-dominated selection left the process open to influence the court could not police; and it altered the appointment scheme of Articles 124 and 217 in a manner that, taken together, impaired the judiciary’s primacy rather than merely regulating its exercise. Justice Khehar’s leading opinion for the majority treated the collegium’s primacy as the settled interpretation of the appointment clauses and held that Parliament, acting under Article 368, could restructure the process only within the limit that primacy survived. The link between the appointments question and the doctrine that decided it is the link to the most-consequential-constitutional-moment framework, because the Fourth Judges Case is the basic-structure doctrine applied not to a statute or an amendment of ordinary policy but to a near-unanimous constitutional amendment about the judiciary itself. The bench then did a second thing that the trajectory reading counts as part of the fourth phase’s architecture rather than as an epilogue: in a separate order of December 2015 it invited suggestions from the government, the bar and the public for improving the collegium’s working, listed transparency, a secretariat, eligibility criteria, a complaints mechanism and the Memorandum of Procedure as the heads for reform, and thereby placed the system’s repair on the record as the court’s own undertaking.

The majority was not a single voice, and its internal composition matters to the holding’s weight. Justice Madan B. Lokur, in a separate concurring opinion, examined the appointment record under the collegium and the amendment’s design in detail and concluded that the amendment’s scheme failed the primacy test the Second and Third Judges Cases had laid down, treating the eminent persons’ undefined status as a defect the court could not cure by interpretation. Justice Kurian Joseph, concurring, recorded in striking language that the collegium itself stood in need of reform, that the system he voted to restore had failed the expectations of the judgments that created it, and that his concurrence in striking the amendment down was a vote for primacy as a constitutional compulsion and not a certificate for the collegium’s working, a distinction he pressed again in the December 2015 order on improvements. Justice A. K. Goel, concurring, addressed the argument from the amendment’s near-unanimous passage and held that the width of a parliamentary majority does not enlarge the amending power, since the basic-structure limit binds a unanimous Parliament exactly as it binds a bare one. The four concurrences thus agreed on the result and on the ratio, and disagreed, in print, on how much of the collegium’s record could be defended, with two of the four judges who saved the system using their opinions to indict its operation. That internal division is part of the judgment’s architectural legacy: the restored architecture carried, from the day of its restoration, a judicially authored repair agenda and a judicially recorded warning.

Justice J. Chelameswar, the senior-most member of the bench after the presiding judge in the order of the court’s seniority at points in the case’s life and its lone dissenter, held the opposite on the amendment and said why in terms that the trajectory must render at full strength, because his dissent became the most quoted judicial text of the fourth phase. He held that the Ninety-ninth Amendment did not violate the basic structure. His reasoning accepted that judicial independence was part of that structure and denied that primacy, as the collegium practised it, was identical with independence. The proceedings of the collegium, he wrote, were absolutely opaque and inaccessible both to the people and to history, save occasionally in a season of political controversy. A system so closed could not claim that its own continuation was a constitutional necessity, and the failures of the collegium, including appointments widely regarded within the profession as indefensible, were matters he recorded from his own knowledge of the institution. He also held the NJAC Act, as distinct from the amendment, invalid in part, treating the statutory veto provision differently from the constitutional design. The dissent’s enduring force lies in its source. The case for the collegium’s opacity had been made by scholars and politicians for twenty years and could be discounted as the resentment of the excluded. Made by a sitting judge of the court, in a judgment, against his own institution’s claim of power, it could not be discounted, and every later critic of the collegium, in Parliament, in the press and on the bench itself, has cited it as authority. The majority and the dissent therefore leave the fourth phase with a structure the rest of this article tracks: the architecture restored as a matter of law, and its legitimacy indicted as a matter of record by one of the five judges who restored it.

What happened to appointments while the NJAC challenge was being heard?

Appointments were effectively paused during the challenge. The amendment and the Act entered into force on 13 April 2015, the collegium stood displaced in law, the new commission never met to make a recommendation, and the court heard the validity question through 2015. The judgment of 16 October 2015 revived the collegium, and appointments resumed under it.

After the verdict: the Memorandum of Procedure negotiation and the open fourth phase

The fourth phase did not end with the judgment that opened it, and its longest chapter is a negotiation conducted in letters between the Supreme Court and the Government of India over the document that governs appointments in practice, the Memorandum of Procedure. The Memorandum is not a statute and not a constitutional text. It is an executive-judicial instrument, first framed in 1999 to consolidate the procedure flowing from the 1993 and 1998 judgments, which prescribes timelines, the sequence of consultations, the handling of Intelligence Bureau reports and the route by which names move between High Court collegiums, state governments, the Union government and the Supreme Court collegium. In its order of December 2015 the Constitution Bench directed that the Memorandum be finalised afresh by the government in consultation with the Chief Justice of India, with the improvements the court had listed, transparency and a permanent secretariat among them, as the framework for discussion. The government submitted a revised draft in March 2016. The collegium returned it with objections, and the negotiation settled into a dispute over clauses that went to the allocation of power in miniature. The draft proposed that the government could reject a recommended candidate on grounds of national security and that such a rejection would bind; the collegium, under Chief Justice J. S. Khehar, returned the draft in August 2016 recording that a binding executive veto dressed as a national security exception reproduced inside a procedure the negation the amendment had attempted at the constitutional level. The draft proposed language on merit and on the recording of reasons for selection and rejection that the collegium treated as an attempt to reopen settled primacy by administrative drafting. Successive drafts and letters narrowed some differences and left others standing, and no final revised Memorandum was adopted. The original Memorandum of 1999 continued to govern in the interim, which is the condition in which the fourth phase operated for years: the architecture restored by judgment, its operating manual under negotiation without conclusion.

The correspondence itself is the under-cited primary record of the phase, and it rewards the kind of attention usually reserved for judgments. The letters between the collegium and the Ministry of Law and Justice across 2016 and 2017, preserved in part in the ministry’s records and drawn on by reporters who cover the appointment process, document a negotiation in which each side stated its constitutional position in administrative language: the government pressing for recorded criteria, a screening mechanism and a decisive voice on security objections; the collegium conceding ground on timelines and on the communication of reasons in defined circumstances while refusing any clause that made an executive objection final. Journalists who reconstructed the stalled negotiation from that correspondence, much of it obtained through disclosure requests and protected sourcing of the kind discussed in accounts of how journalists use VaultBook to protect sources, established the sequence of drafts that the formal record, confined to the court’s orders, leaves incomplete. For the trajectory reading the episode carries a precise significance. The fourth phase’s central tension was not enacted in a statute or decided in a case. It was negotiated in an exchange of drafts over a procedural document, which means the allocation of appointment power after 2015 was being contested at the level of working rules by the two institutions the allocation concerned, with no third party empowered to close the contest. That is a different kind of constitutional politics from the amendment politics of Phase Three, quieter in form and identical in subject.

Why did the court revive the collegium instead of designing a new appointing body?

The majority held the amendment void for damaging judicial primacy in appointments, a constituent of judicial independence within the basic structure. A void amendment left the prior law standing, and the prior law, as interpreted in 1993 and 1998, was the collegium. The bench restored the earlier architecture by the logic of its holding.

Alongside the Memorandum negotiation, the fourth phase produced a series of episodes in which the restored architecture and the government tested each other over individual appointments and over the tempo of the process as a whole. In January 2017 the government returned the collegium’s recommendation for the appointment of K. M. Joseph, then Chief Justice of the Uttarakhand High Court, whose bench in April 2016 had quashed the imposition of President’s Rule in that state; the collegium reiterated the recommendation in a modified form in 2018, and the appointment was eventually made in August 2018 after a further exchange and after the government had processed other names from the same set ahead of his, with a consequent loss of seniority that the bar publicly criticised. In 2018 the elevation of Justice Indu Malhotra directly from the bar to the Supreme Court, the first such appointment of a woman advocate, was processed at a different speed from the Joseph file within the same recommendation, and the contrast was entered into the public argument by senior advocates on both sides. Delay, the executive instrument that Phase Two had already identified, became the fourth phase’s most debated practice. Successive Chief Justices stated the judiciary’s position in open court and in addresses: that files pending with the government beyond the Memorandum’s timelines were appointments withheld in effect, and that the collegium could recommend but could not appoint. Successive Law Ministers stated the government’s position in Parliament and in letters: that verification of antecedents was a legitimate and time-consuming executive function, that the court was being sent names at a rate the process could not clear, and that the government had in fact appointed judges in large numbers across the phase. Both statements were supported by figures each side selected, and the figures were rarely computed on the same base, which the scholarly commentary of the phase repeatedly noted. The episode that carried the contest furthest into public view came from within the judiciary itself. On 12 January 2018 the four senior-most judges of the Supreme Court after the Chief Justice, Justices J. Chelameswar, Ranjan Gogoi, Madan B. Lokur and Kurian Joseph, held a press conference in New Delhi and stated that the administration of the court, including the assignment of sensitive cases to particular benches, had departed from convention in ways that threatened the institution, and Justice Chelameswar said that democracy in the country was at stake. The press conference concerned the Chief Justice’s roster power rather than appointments directly, and it belongs in this article because three of its participants had sat in the Fourth Judges Case and because its most senior participant was that case’s dissenter: the judge who had called the collegium opaque in judgment in 2015 was, by 2018, telling the country that the court’s internal government had become a public question. The fourth phase’s tensions, that day, were visible as a single fabric: appointments, roster, seniority and the executive’s reading of all three.

The phase continued past 2020 without architectural settlement, and its later record is part of the evidence the brief for this article requires to be carried. In November 2022 and the months following, the Minister of Law and Justice, Kiren Rijiju, stated publicly and repeatedly that the collegium system was opaque and unaccountable, that it was the only system in the world in which judges appointed judges, and that the government intended to press for a role in appointments consonant with what he described as the spirit of the Constitution; the Vice President of India, Jagdeep Dhankhar, speaking in the Rajya Sabha in December 2022 in his capacity as its Chairman, questioned how a constitutional amendment passed by Parliament and ratified by the states could be undone by a judgment, and returned to the theme in later addresses. The judiciary’s stated position in the same period was given by a bench of the Supreme Court hearing a petition on appointment delays in late 2022 and 2023, which recorded in open court that the government was bound to process collegium recommendations within the framework the judgments laid down, that keeping names pending while appointing others from the same set disturbed the seniority the system presupposed, and that the court’s restraint in the face of delay should not be read as the absence of a remedy. In January 2023 the collegium took a step without precedent in the trajectory: it published, on the Supreme Court’s website, the reasons recorded in its resolutions for a set of recommendations, including the government’s objections to particular candidates and the collegium’s answers to them, in the cases of candidates whose files had become matters of public controversy. Transparency, demanded of the collegium since the 1990s by its critics and promised in the court’s own order of December 2015, arrived in partial form through the collegium’s own practice rather than through the unfinished Memorandum. The fourth phase at its later end therefore shows movement on the very ground the dissent of 2015 had marked, achieved by the restored architecture acting on itself, alongside a parliamentary and executive critique of that architecture stated as openly as at any point since 2014. Neither movement closed the phase.

The collateral powers: transfers, seniority and the Chief Justiceship

The appointment power at the centre of the contestation does not stand alone, and the four phases cannot be understood without the collateral powers that travel with it, because each collateral power was at some point the real subject of a dispute that wore the dress of an appointments dispute. The first is transfer. Article 222 empowers the President, after consultation with the Chief Justice of India, to transfer a judge from one High Court to another, and the judges cases brought transfer within the collegium’s consultative discipline, with the 1998 opinion requiring the collegium’s recommendation and the views of the affected Chief Justices. A transfer can serve the system, moving a judge whose continuance in a home court has become untenable for reasons of propriety or local entanglement. A transfer can also punish, and the receiving grammar of the service is that an out-of-turn transfer, to a smaller court or at a career’s delicate point, is read by the transferred judge and by the profession as a penalty imposed without charge or hearing. The fourth phase supplied the episode the whole country read in those terms. In February 2020 the collegium recommended the transfer of Justice S. Muralidhar from the Delhi High Court to the Punjab and Haryana High Court, and the government’s notification of the transfer issued on the night of 26 February 2020, days after a bench headed by Justice Muralidhar had passed orders concerning the violence in Delhi and the registration of cases arising from speeches made during it. The collegium’s recommendation predated those orders, the government stated that the sequence was routine and consensual in the manner transfers require, and Justice Muralidhar himself, at his farewell reference, made no complaint and spoke of the transfer system in terms of institutional duty. The episode is recorded here at the level of the dated acts and the stated positions, and its lesson for the trajectory is structural rather than personal: a transfer system that publishes no reasons leaves every transfer, routine or otherwise, available for the darkest reading, and the executive’s speed in notifying this one, set against its delays elsewhere, became part of the public argument about the allocation as a whole.

The second collateral power is seniority, which the appointments architecture manufactures as a by-product and then treats as a fact of nature. Because succession to the Chief Justiceship follows seniority, and because the zone of consideration for the Supreme Court follows the seniority of High Court Chief Justices, the order in which a government processes a single set of recommendations decides careers. The fourth phase’s Joseph episode turned on exactly this: the appointment of one recommended judge ahead of another from the same set altered their relative seniority, and with it their prospects, without any institution having decided that the alteration should occur. The collegium protested the sequencing in its resolutions; the government answered that the warrant of appointment was the President’s act and that seniority among judges appointed on different dates had always followed the dates of appointment. Both positions were available within the texts, which is the point. An architecture that leaves the sequencing of its own outputs to the branch that lost the appointment power has left a residue of that power in the loser’s hands, and the residue is exercised, lawfully and consequentially, every time a file is cleared out of turn. The published resolutions of January 2023, in which the collegium recorded the government’s stated objections and its own replies, were in part an attempt to close that residue by publicity: a sequence performed in the open is harder to arrange in the dark.

The third collateral power is the Chief Justiceship itself, the office the supersessions of 1973 and 1977 had made the system’s exposed point and the 1993 judgment had tried to protect by entrenching seniority. Across the four phases the seniority convention held unbroken after 1993, and its holding is one of the restored architecture’s genuine achievements: no government of Phase Two, Three or Four attempted a supersession, and the amendment of 2014, for all its reach, left the convention untouched, since the commission it created would have recommended the Chief Justice as it recommended other judges, with seniority as the settled practice the government’s own counsel did not disown. The convention’s survival across a phase in which everything else about appointments was contested shows the trajectory’s selectivity. The contest was never about whether judges should be independent in office; every text in the record, including the amendment’s, professes that commitment. The contest was about who holds the pen at the single moment when the institution’s future composition is written, and the collateral powers matter because each of them is a way of holding a corner of that pen after the headline question has been answered.

What the scholars made of it: the argument stated at full strength

The scholarly literature of the trajectory is not commentary on the four phases from outside; it is one of the arenas in which the phases were fought, and the stance of this article requires the leading positions to be anchored to named works. Arghya Sengupta’s Independence and Accountability of the Higher Indian Judiciary, published in 2019, supplies the most complete institutional treatment and the one this article’s reading most directly engages. Sengupta, a founder of the Vidhi Centre for Legal Policy, argues that the Indian debate has been deformed by treating independence and accountability as opposites, that the higher judiciary’s institutional design secures independence from the executive while leaving its internal accountability, in appointments, administration and discipline, underdeveloped, and that reform should proceed by redesigning the appointment process to combine judicial participation with recorded reasons, criteria and a secretariat, rather than by restoring executive primacy or by defending the collegium unchanged. The book’s force for the trajectory is that it refuses both of the framings this article also refuses: it will not treat 2015 as a simple victory of independence, and it will not treat the collegium’s critics as enemies of independence. Pratap Bhanu Mehta, writing in essays across the post-2015 period in The Indian Express and elsewhere, pressed the institutional critique from the standpoint of constitutional morality: that the court’s defence of its own appointment power sat uneasily with its silence or selectivity elsewhere, that the collegium’s claim to embody independence was weakened each time the institution appeared to bargain with the executive over names, and that the judiciary’s authority rested on reasons given in public, a discipline the appointment process conspicuously lacked. Suhrith Parthasarathy, writing in The Hindu and in legal commentary across the same period, examined the architecture at the level of individual appointments and transfers, arguing from the record of particular files that the opaqueness of the process injured candidates, governments and the court alike, since a decision without reasons could be neither defended when it was right nor corrected when it was wrong, and that the Memorandum negotiation’s failure was a failure for which both institutions bore responsibility. Against these positions the defenders of the restored architecture stated their case at matching strength. Fali S. Nariman, who had argued the 2015 challenge, maintained in his writing and addresses that the history from the supersessions of 1973 and 1977 to the amendment of 2014 taught a single lesson, that every generation of the political executive sought a hand in choosing its judges, and that primacy, for all the collegium’s faults, was the only arrangement that had kept that hand off the institution’s composition. Senior advocates and former judges who took this side pointed to the four-phase record itself: the executive’s instruments in every phase had been delay, objection and amendment, in ascending order of ambition, and the failure of the most ambitious attempt was the system’s proof rather than its embarrassment.

The disagreement this article must adjudicate is narrower than the whole scholarly quarrel, and it is the disagreement the brief names: whether the contestation is best read case by case, as doctrine succeeding doctrine; or politically and strategically, as the record of one government’s reform and its defeat; or as the four-phase architectural trajectory defended here. The case-by-case reading is not foolish, and its evidence is real. The doctrines genuinely differ: S. P. Gupta’s consultation without concurrence, the Second Judges Case’s primacy, the Third Judges Case’s plurality and procedure, the Fourth Judges Case’s basic-structure invalidation are distinct propositions, and a lawyer who cannot state them separately cannot practise in the area. But the case-by-case reading cannot explain the phenomena that fall between its events: why two commissions and two governments of different parties spent Phase Two proposing commissions; why the amendment of 2014 passed a Parliament of rivals in days; why the defeated architecture’s repair was negotiated in letters about a memorandum for years after the doctrine was settled; why a dissenting judge’s paragraph on opacity out-argued the majority’s hundred pages in the public life of the judgment. The political-strategic reading explains the amendment and cannot explain the rest. It cannot explain why the government of 2003 and the government of 2013, of opposed coalitions, carried near-identical designs; it cannot explain why the critique of the collegium was voiced most damagingly by judges, including the author of the 1993 judgment and the dissenter of 2015; and it cannot explain why the executive after 2015, holding the same majority that had passed the amendment, contested appointments through delay and correspondence rather than through a second amendment, unless the constraint was architectural rather than merely political. The four-phase reading explains what the other two leave over, because it tracks the thing each phase actually transmitted to its successor: an allocation of power, with its procedure, its legitimation and its accumulated grievances, which the successor phase operated, criticised, attempted to substitute or restored under judgment. Doctrine and politics both appear in that account, in their places, as the instruments by which allocations were made and contested. Neither is the story. The allocation is the story, and the story is unfinished.

The four-phase map: the contestation on one page

The reading defended in this article can be set out in a single citable form, and the table below is that form. It is labelled here as Table 1, the Four-Phase Architectural-Trajectory Map of the Collegium NJAC Contestation, and each row states a phase by its dates, its governing instrument, the allocation of appointment power the instrument put into operation, and the tension the phase transmitted to its successor.

Phase	Dates	Governing instrument	Allocation of appointment power	Tension transmitted
Phase One	1993 to 1998	Second Judges Case, 6 October 1993; Third Judges Case, 28 October 1998	Judicial primacy: collegium recommends, executive processes and may return once, reiteration binds	An architecture built by interpretation, with no legislative pedigree
Phase Two	1998 to 2014	Collegium practice under the 1993 and 1998 readings; Memorandum of Procedure of 1999	Judicial primacy in operation, executive voice confined to objection and delay	Opacity and unassignable vacancies; commissions of 2002 and 2008 recommend change
Phase Three	2014 to 2015	Constitution (Ninety-ninth Amendment) Act, 31 December 2014; NJAC Act, 2014	Shared power: six-member commission, judicial parity, executive membership, two-member veto	A substitution carried near-unanimously and challenged before it could operate
Phase Four	2015 onwards	Fourth Judges Case, 16 October 2015, by four votes to one; collegium revived	Judicial primacy restored by judgment; repair negotiated through a revised Memorandum	Restoration without settlement: delay, correspondence and public critique continue

Three features of the map deserve emphasis, because they are the features the rival readings miss. The first is that the phases are joined by transmission rather than by mere succession. Phase Two’s criticism was not background noise to Phase Three; it was Phase Three’s enabling condition, assembled in commission reports and lapsed bills over sixteen years. Phase Four’s Memorandum negotiation was not an administrative sequel to the judgment; it was the contest over allocation continuing by other instruments, as this pillar’s treatment of the institution-under-stress framework shows for a sister institution whose founding design is likewise contested in operation long after its text was settled. The second feature is the asymmetry the map makes visible between the two branches’ instruments across all four phases. The judiciary acted through interpretation and judgment, in 1981, 1993, 1998 and 2015. The executive acted through practice, delay, commission, amendment and correspondence, in every phase. Each branch reached for the instruments its position supplied, and the trajectory is the record of those instruments meeting. The comparison with the Election Commission’s design, where the appointment of the watchdog was left to the executive alone for seven decades, is drawn out in the architectural-asymmetry-question framework, and it sharpens the point: India’s constitutional institutions divide sharply between those whose staffing the judiciary captured and those whose staffing it never contested, and the appointments of judges fall on the contested side of that line for reasons the four phases supply. The third feature is that the map has an open final row. Phase Four has no closing instrument, and the reading does not pretend otherwise, which brings the argument to its hardest test.

The road not taken: the designs available between collegium and NJAC

Between the architecture the court restored and the architecture Parliament enacted there stood, in 2015, a set of designs that neither side adopted, and the trajectory is incomplete without them, because their existence disciplines the argument in both directions. The first was the Venkatachaliah design of 2002, described earlier in this article: a five-member commission with a judicial majority of three, the Law Minister, and one eminent citizen nominated by the President in consultation with the Chief Justice. Set beside the NJAC’s six members and judicial parity, the 2002 design gave the judiciary a majority rather than a half, gave the citizen member a nominating process anchored in the presidency rather than in a committee the executive dominated, and contained no statutory two-member veto. A constitutional amendment in that form would have presented the 2015 bench with a harder question than the one it decided, since the majority’s reasoning fastened on the dilution of primacy below parity’s line and on the eminent persons’ undefined selection. The second design was procedural rather than compositional, and it was available without any amendment at all: the collegium reformed from within, with published eligibility criteria, a permanent secretariat to prepare dossiers and record decisions, recorded reasons for selection and rejection communicated to the government, and a complaints process for the conduct of the process itself. Every element of that design appeared in the court’s own list of heads in its order of December 2015, which is to say the bench that restored the collegium also wrote the specification for its repair. Elements of it arrived piecemeal in the fourth phase, the publication of resolutions with reasons from January 2023 being the largest, and the secretariat and the criteria remaining, on the public record, unbuilt. The third design was the Law Commission’s road of 2008, reconsideration of the judges cases by a larger bench, which would have returned the question to doctrine and, if the cases fell, restored the appointment clauses to the meaning the Assembly’s debates support, leaving Parliament free to legislate a commission of any composition consistent with independence as a general principle rather than with primacy as its synonym. No government moved for that reference, and the court did not make it of its own motion in 2015, preferring to decide the amendment’s validity on the law as the earlier benches had settled it.

The three unchosen designs matter to the adjudication this article owes its readers, because they mark the distance between the two framings it rejects. The political-strategic framing cannot accommodate them: a contest between a government seizing power and a court defending it has no place for the observation that the government enacted the bluntest of the available instruments when a subtler one, carrying a judicial majority and a defensible pedigree from a commission headed by a former Chief Justice, lay ready to hand, nor for the observation that the court, having won, left its own repair specification largely unexecuted for years. The case-by-case framing cannot accommodate them either, since each design is a proposition about architecture, not about the correctness of any single judgment’s reading of the word consultation. The four-phase reading accommodates them as what they are: the unrealised branches of the trajectory, which show that the allocation of appointment power in India was contested at every phase among at least three coherent settlements, of which the enacted and restored ones were, at the moment of decision, the two most absolute. A settlement that shares power in recorded proportions, or a collegium that disciplines itself with the instruments of an accountable institution, remains available to a future phase precisely because the arguments for both were fully made inside this one and answered, where they were answered, on grounds narrower than their merits.

The comparative argument and the outlier charge

One argument recurred in every phase and deserves separate treatment, because it carried the amendment in Parliament and it survives the amendment’s fall: the claim that India stood alone. The government of 2014 stated it in the debates on the amendment, and the Law Minister of the fourth phase, Kiren Rijiju, restated it in his public interventions of 2022, that India was the only country in the world in which judges appointed judges. Stated as a debating point the claim is effective. Examined as comparative law it requires the kind of precision the trajectory reading applies to every other claim in the contest, taking the comparison seriously and stating what it does and does not show. The United Kingdom, whose Judicial Appointments Commission was created by the Constitutional Reform Act of 2005, staffs its senior courts through a commission of mixed judicial, professional and lay membership that selects on merit and certifies a single name to the Lord Chancellor, whose power to reject is confined and reasoned. South Africa’s Judicial Service Commission, created by the Constitution of 1996, interviews candidates for the Constitutional Court and the superior courts in public sessions, with membership drawn from the judiciary, the legal profession, Parliament and the executive, and advises the President, who appoints Constitutional Court judges from its list. The United States stands at the other pole: federal judges are nominated by the President and confirmed by the Senate in open political contest, and the system’s defenders treat that publicity as the accountability the Indian debate seeks elsewhere. Appointment commissions of the British and South African kinds share two features the Indian collegium lacks and the NJAC possessed in form: membership that includes persons answerable outside the judiciary, and proceedings that generate a record. They share one feature with the collegium that the NJAC’s critics said the Indian design abandoned: the executive cannot impose a name. In the British design the minister’s rejection is exceptional and explained; in the South African design the President’s choice is confined to a list the commission supplies.

The comparative record therefore cuts in two directions at once, and honesty about it is a test of the evenhanded treatment this subject requires. Against the collegium’s defenders, the comparison shows that no comparable democracy entrusts appointment wholly to the appointees’ own institution, and that the features foreign systems use to reconcile independence with accountability, lay membership, published criteria, interviews and records, are features the Indian system could adopt without surrendering primacy, as Sengupta’s 2019 institutional treatment argues in detail from the same comparisons. Against the amendment’s defenders, the comparison shows that foreign commissions operate inside conventions and statutory disciplines that the Indian instruments of 2014 did not reproduce: no foreign design examined in the 2015 hearing gave a serving minister of the government a vote on the selection of judges before whom that government litigates daily, and none left its lay membership’s qualifications wholly undefined. The court in the Fourth Judges Case had the comparative material before it through the parties’ written submissions, and the majority’s treatment of it was brief compared with its treatment of text and precedent, a thinness the scholarly commentary of the fourth phase noted from both directions. The trajectory reading assigns the comparison its correct weight. It explains the persistence of the reform demand better than any domestic document, since every commission that examined the question, in 2002, in 2008 and in the lapsed bills, worked with the foreign models in view; and it cannot by itself settle the constitutional question the 2015 bench had to decide, which was not whether a commission was wise but whether the Constitution, as its basic structure had been construed, left Parliament room to enact this one. A reader who keeps those two propositions separate has the comparative argument in its strongest and its only defensible form, and the argument remains available, in that form, to whichever institution opens the contestation’s next phase.

The strongest objection: a trajectory without a settlement

The strongest counter to the four-phase reading is that it claims more pattern than the record holds. On this objection, the post-2015 years show not a phase but a stalemate: the collegium restored and unreformed in its essentials, the Memorandum unrevised, the executive’s critique unanswered by any new architecture, and individual appointments fought file by file with no principle settled beyond the ones the 2015 judgment had already stated. A trajectory, the objection runs, implies direction, and a decade of non-resolution has none; the honest description of the period is drift, and a framework that dignifies drift as a phase is extending itself past the evidence. The objection is the correct one to press, and it is pressed here in the form its best exponents would recognise, since scholars who defend the collegium and scholars who condemn it have both used the stalemate description against the other’s complacency. The answer the four-phase reading gives is that a phase in an architectural trajectory is defined by the allocation in operation and the mode of contest over it, not by movement toward resolution. Phase Two lasted sixteen years and resolved nothing; its unity is the unity of an allocation operating under criticism. Phase Four’s unity is of the same kind, with the mode of contest changed: where Phase Two’s contest ran through commissions and lapsed bills aimed at a future substitution, Phase Four’s runs through correspondence, delay and published resolutions inside a restored allocation whose legitimacy both sides continue to dispute in public. The framework is robust to non-resolution because non-resolution is one of the states an allocation can be in, and the framework’s claim is about the continuity of the contested allocation, not about progress. What would falsify the reading is not stalemate but rupture of a kind the record does not show: a second amendment carried and upheld, a larger bench reconsidering primacy itself, or a negotiated settlement embodied in a final Memorandum accepted by both sides as closing the question. Any of those would open a fifth phase, and the map has a row waiting for it. Until one of them occurs, the contestation’s fourth phase continues, and its continuation is a datum, not an embarrassment, for the trajectory that contains it.

The same discipline governs how the phase’s evidence past 2020 should be weighed, and a note on method belongs in the open. The episodes of 2022 and 2023 recorded in this article, the ministerial statements, the parliamentary remarks and the collegium’s published resolutions, are included because the fourth phase is defined as running from 2015 onwards and because the brief for this pillar’s appointments article fixes the post-2015 trajectory as part of its subject. They are included at the level of stated positions and dated acts, attributed to the office-holders who made them, and no inference about motive is drawn from them in either direction. The government of the phase stated a reforming intention and an accountability argument; the judiciary stated a constitutional obligation and a seniority argument; the record supports both statements and neither caricature. That restraint is the evenhandedness the subject demands, and it is consistent with the wider frame of the integrated three-tensions framework, in which the appointments contest appears as one durable tension among the several that Indian politics renegotiates without concluding. One further discipline follows from the map, and it concerns how the next development in the contestation, whatever it is, should be reported when it occurs. A recommendation published with reasons, a file returned with objections, a ministerial speech, a bill introduced: each will be announced, in the habits the rival framings have taught, either as a fresh case to be catalogued or as a fresh manoeuvre to be scored. The trajectory reading instructs a different first question. Ask which allocation the development presupposes, which instrument it uses, and which grievance inherited from an earlier phase it answers. A development that answers an inherited grievance is a phase working itself out. A development that changes the allocation, and survives, is a phase ending. The distinction will not settle the argument about whether any given development is wise, and it is not meant to. It will settle the prior question of what kind of event the development is, and in a contestation that has been misdescribed for thirty years by the two framings this article rejects, settling that prior question is the beginning of analysis rather than its end.

A reader who finishes this article persuaded that the collegium is indefensible has read only its critics’ chapters; a reader who finishes persuaded that the amendment was a mere power grab has read only its defenders’. The article’s claim is prior to both verdicts and survives either of them: the contestation is one architectural trajectory in four phases, the phases are joined by the transmission of an allocation and its grievances, and any future settlement, whenever it comes, will be intelligible only as the fifth phase of the sequence this article has mapped.

Who else holds a stake: the bar, the litigant and the district judiciary

The contestation is usually narrated as a duel between two institutions in New Delhi, and the four phases look different when the other holders of a stake are brought into the frame, because each phase redistributed something among them as well. The organised bar is the first of these holders. The Supreme Court Bar Association and the Bar Council of India appear in the record at every turn: as petitioners and interveners in the judges cases, as the constituency whose resolutions on individual appointments supplied Phase Two’s criticism with its evidence, and as a divided house in 2014 and 2015, when senior members of the same profession argued the amendment’s validity on opposite sides. The bar’s stake is direct. Its senior members are the pool from which Supreme Court judges are occasionally elevated and from which High Court judges are routinely drawn, and its members practise before the appointees the system produces. A selection process conducted in confidence among judges asks the bar to trust the selectors’ knowledge of the selected, and the bar’s institutional answer across the phases was to demand a record: criteria it could read, reasons it could test, and a secretariat it could petition. That demand, more than any party’s programme, is the thread connecting the Venkatachaliah commission’s citizen member, the amendment’s eminent persons and the collegium’s own published resolutions of 2023, three instruments of three different phases aimed at the same absence.

The litigant is the second holder, and the easiest to forget because no litigant is consulted in any version of the design. The Union of India and the state governments together appear in a very large share of the cases the higher judiciary decides, as tax authorities, land acquirers, regulators, prosecutors and respondents to challenges against their own legislation. The identity of the judges before whom the state litigates is therefore a question of the state’s interest as a party, and the constitutional problem of the appointment power can be stated without any imputation of motive: a litigant that chooses the bench holds an advantage no rule of procedure can neutralise, and a bench that chooses itself holds an insulation no record of integrity can fully justify to the losing party. Every design in the trajectory is an attempt to weight that dilemma, and the four-phase reading’s sympathy is with the dilemma rather than with any weighting of it. The district judiciary is the third holder, and its stake is the largest in numbers and the least heard. The subordinate judiciary of the states, recruited through state judicial services and promoted through the High Courts’ administrative control, supplies the majority of High Court elevations through the service channel, and its members’ careers are governed by the same collegiums that dominate this article, acting on the administrative side under Articles 233 to 235 of the Constitution. A district judge’s prospect of elevation depends on the timing and seniority questions this article has tracked, and the delay that edits the Supreme Court’s future pool edits the district judiciary’s present incentives first. When the phases are read with these three holders in view, the contestation stops resembling a duel and takes its true shape: a permanent negotiation among the bench, the executive, the profession and the service over a single scarce good, the power to decide who shall judge, conducted across thirty years in the languages of doctrine, amendment, correspondence and delay, and not concluded.

Frequently Asked Questions

Q: What is the collegium system?

The collegium system is the method by which judges of the Supreme Court and the High Courts of India are appointed on the recommendation of a body of senior judges. For the Supreme Court the body is the Chief Justice of India with the four senior-most judges of the court; for a High Court appointment the recommendation begins with that court’s Chief Justice and two senior-most judges and is settled by the Chief Justice of India with the two senior-most Supreme Court judges. The system rests on the judgments of 1993 and 1998 rather than on any statute, and its recommendations bind the government when reiterated.

Q: What was the Second Judges Case?

The Second Judges Case is the judgment in Supreme Court Advocates-on-Record Association v Union of India, delivered by a nine-judge bench on 6 October 1993. The majority, speaking through Justice J. S. Verma, held that the consultation required by Article 124 of the Constitution vested primacy in the opinion of the Chief Justice of India formed with his two senior-most colleagues, departing from the First Judges Case of 1981, which had given the decisive voice to the executive. The judgment thereby created the collegium as the effective appointing authority for the higher judiciary and held that seniority should govern succession to the Chief Justiceship save in exceptional circumstances.

Q: What is NJAC?

NJAC is the National Judicial Appointments Commission, the body created by the Constitution (Ninety-ninth Amendment) Act of 2014 and the National Judicial Appointments Commission Act of 2014 to replace the collegium as the appointing authority for the higher judiciary. It had six members: the Chief Justice of India as chair, the two senior-most Supreme Court judges, the Union Law Minister, and two eminent persons chosen by a committee of the Prime Minister, the Chief Justice of India and the Leader of Opposition. The commission never made an appointment, because a Constitution Bench struck down both instruments on 16 October 2015 before the body began to function.

Q: What was the 99th Amendment?

The 99th Amendment is the Constitution (Ninety-ninth Amendment) Act, 2014, which received the assent of President Pranab Mukherjee on 31 December 2014 after passage by both Houses of Parliament and ratification by sixteen state legislatures. It inserted Articles 124A, 124B and 124C into the Constitution, constituting the National Judicial Appointments Commission, charging it with recommending appointments and transfers in the higher judiciary, and empowering Parliament to regulate its procedure. The amendment was the legislative half of the 2014 substitution of the collegium, and it was declared unconstitutional by the Supreme Court on 16 October 2015 in the Fourth Judges Case.

Q: What did the Fourth Judges Case decide?

The Fourth Judges Case, decided on 16 October 2015 by a Constitution Bench of five judges, held by four votes to one that the Constitution (Ninety-ninth Amendment) Act and the National Judicial Appointments Commission Act were unconstitutional and void. The majority held that the independence of the judiciary forms part of the basic structure of the Constitution, that judicial primacy in appointments is a constituent of that independence, and that the amendment damaged primacy through the Law Minister’s membership, the undefined eminent persons and the design taken as a whole. The collegium system of 1993 and 1998 stood revived by the judgment.

Q: Who is Justice Chelameswar?

Justice J. Chelameswar is a former judge of the Supreme Court of India who wrote the lone dissent in the Fourth Judges Case of 16 October 2015. He held that the Ninety-ninth Amendment did not violate the basic structure of the Constitution, reasoning that judicial independence did not require the collegium’s primacy and recording that the collegium’s proceedings were opaque and inaccessible to the public and to history. In January 2018 he was the senior participant in the press conference held by the four senior-most judges of the court on the administration of the institution. He retired from the Supreme Court in June 2018.

Q: How does the post-2015 collegium operate?

The post-2015 collegium operates under the procedure consolidated in the Memorandum of Procedure, with recommendations moving from High Court collegiums through state governments and the Union government to the Supreme Court collegium, whose reiterated recommendations the government appoints. After the judgment of October 2015 the court directed that the Memorandum be revised in consultation with the Chief Justice of India, and drafts exchanged between the government and the collegium from 2016 onwards narrowed some differences without producing an adopted revision. From January 2023 the collegium also began publishing its resolutions with reasons in contested cases on the Supreme Court website.

Q: What did the First Judges Case decide in 1981?

The First Judges Case, S. P. Gupta v Union of India, decided on 30 December 1981 by a seven-judge bench, held that the consultation required by Articles 124 and 217 of the Constitution did not mean concurrence. The majority, in Justice P. N. Bhagwati’s leading judgment, reasoned that the Constituent Assembly had considered giving the Chief Justice of India a veto over appointments and had declined to do so, and that the executive could therefore depart from the Chief Justice’s opinion for cogent reasons. The holding gave the government of the day the decisive voice in appointments for twelve years, until the nine-judge bench in the Second Judges Case of 1993 read the same provisions the other way and vested primacy in the judiciary.

Q: How did the Third Judges Case change the size of the collegium?

The Third Judges Case, the advisory opinion in Special Reference No. 1 of 1998 delivered on 28 October 1998, enlarged the collegium and fixed its composition. The Second Judges Case of 1993 had formed the Chief Justice of India’s binding opinion in consultation with his two senior-most colleagues. The 1998 opinion held that for appointments to the Supreme Court the collegium would consist of the Chief Justice and the four senior-most judges of the court, and that the opinion of the Chief Justice in his individual capacity, unsupported by that plurality, was not binding on the government. For High Court appointments the Chief Justice of India acts with the two senior-most judges. The opinion also required the consultation to be recorded in writing.

Q: What would Articles 124A, 124B and 124C have done?

The three articles inserted by the Ninety-ninth Amendment in 2014 were the constitutional skeleton of the National Judicial Appointments Commission. Article 124A constituted the commission and fixed its six members: the Chief Justice of India, the two senior-most Supreme Court judges, the Union Law Minister and two eminent persons. Article 124B charged the commission with recommending appointments of the Chief Justice of India, judges of the Supreme Court, Chief Justices and judges of the High Courts, and with recommending transfers of High Court judges. Article 124C empowered Parliament to regulate the commission’s procedure by law and to authorise the commission to frame its own regulations. All three articles fell with the amendment when the Fourth Judges Case struck it down on 16 October 2015.

Q: Why did the majority hold that the NJAC damaged judicial independence?

The majority in the Fourth Judges Case held that judicial independence is part of the basic structure of the Constitution and that primacy of the judiciary in appointing judges is a constituent of that independence, since the power to choose judges shapes the institution that checks every other power. It found the damage in the design taken together: the Union Law Minister sat as a voting member of the body choosing judges before whom the government was the largest litigant; the two eminent persons had no defined qualifications and were selected by a committee in which the executive held two of three voices; and the arrangement displaced the primacy the appointment clauses had been held to secure. Parliament could regulate the process, the majority reasoned, but could not amend away the primacy itself.

Q: What did Justice Chelameswar argue in his dissent in the Fourth Judges Case?

Justice J. Chelameswar, dissenting alone in the Fourth Judges Case, held that the Ninety-ninth Amendment did not violate the basic structure of the Constitution. He accepted that judicial independence belonged to that structure and denied that the collegium’s primacy was the same thing as independence. The proceedings of the collegium, he wrote, were absolutely opaque and inaccessible both to the people and to history, save occasionally in a season of political controversy, and a system so closed could not claim constitutional necessity for its own continuation. He recorded appointments under the collegium that the profession regarded as indefensible and held the NJAC Act, as distinct from the amendment, invalid in part. His dissent became the most cited judicial criticism of the collegium in the years that followed.

Q: What is the Memorandum of Procedure and why did its revision stall?

The Memorandum of Procedure is the document, first framed in 1999, that prescribes the working sequence for appointments to the higher judiciary: how names originate in High Court collegiums, how state governments and the Union government record views and obtain antecedent reports, how the Supreme Court collegium settles recommendations, and within what timelines each step should occur. After the Fourth Judges Case the Supreme Court directed in December 2015 that the Memorandum be revised in consultation with the Chief Justice of India. The government’s draft of March 2016 proposed that a candidate could be rejected on national security grounds with binding effect and added language on merit and recorded reasons. The collegium objected that a binding executive rejection reproduced within a procedure the veto the amendment had attempted, and successive exchanges narrowed some points without settling that one, so no revised Memorandum was adopted.

Q: What did the Venkatachaliah Commission recommend for judicial appointments?

The National Commission to Review the Working of the Constitution, chaired by Justice M. N. Venkatachaliah, a former Chief Justice of India, was appointed in 2000 and reported on 31 March 2002. It recommended a National Judicial Commission for appointments to the Supreme Court and the High Courts, chaired by the Chief Justice of India and composed of two senior judges of the Supreme Court, the Union Minister for Law and Justice, and one eminent citizen nominated by the President in consultation with the Chief Justice. The design kept a judicial majority while admitting an executive member and a citizen member, and it became the template for the Constitution Amendment Bill introduced in the Lok Sabha in 2003, which lapsed when that House was dissolved in 2004.

Q: What did the Law Commission recommend in its 2008 report on judges cases?

In its Two Hundred and Fourteenth Report of 2008, titled Proposal for Reconsideration of Judges Cases and Appointment of Judges, the Law Commission of India held that the three judges cases had departed from the original meaning of the Constitution. The report reasoned that consultation in Articles 124 and 217 had not meant concurrence to the Constituent Assembly that debated the provisions in 1949, that the collegium created by interpretation had produced an appointment process without transparency or accountability, and that the position should be corrected either by a larger bench reconsidering the judges cases or by constitutional amendment restoring executive primacy tempered by judicial consultation. The recommendation set out the most far-reaching option on the reform spectrum of the period between 1998 and 2014.

Q: What happens when the government returns a collegium recommendation?

Under the procedure settled by the judgments of 1993 and 1998 and consolidated in the Memorandum of Procedure, the government may return a collegium recommendation once, recording its objections and supplying the material on which they rest, including antecedent reports. The collegium then reconsiders the name in the light of that material. If the collegium drops the name, the process ends. If the collegium reiterates the recommendation, the President appoints the candidate, and the reiteration is treated as binding on the executive. The sequence was tested in the case of K. M. Joseph, whose recommendation as a Supreme Court judge was returned in 2017, reiterated in a modified set in 2018, and given effect in August 2018 after further exchanges over the order in which names from the same set were processed.

Q: Who selects the two eminent persons under the NJAC design?

Under Article 124A as inserted by the Ninety-ninth Amendment, the two eminent persons on the National Judicial Appointments Commission were to be nominated by a committee of three: the Prime Minister, the Chief Justice of India and the Leader of Opposition in the Lok Sabha or, where no opposition leader held formal recognition, the leader of the single largest opposition party in that House. The Constitution prescribed no qualifications for eminence and no field from which the persons were to be drawn, and it required one of the two to be drawn in rotation from the Scheduled Castes, the Scheduled Tribes, the Other Backward Classes, minorities or women. The absence of defined qualifications and the executive’s two voices on the nominating committee were central to the challenge the Supreme Court upheld in 2015.

Q: What role do High Court collegiums play in appointments?

Appointments to a High Court begin in that court’s own collegium, consisting of its Chief Justice and the two senior-most judges of the court. That body initiates names from the bar and from the state’s district judiciary, weighing seniority, merit, integrity and the court’s composition. The proposal then travels through the state government and the Governor, who record their views, and to the Union government, which obtains antecedent reports, before the Supreme Court collegium of the Chief Justice of India and the two senior-most judges settles the recommendation after consulting Supreme Court judges who have served in that High Court. The President then issues the warrant of appointment. The High Court collegium’s initiating role makes it the point at which most information about candidates enters the system, and its choices frame every later stage.

Q: How does Arghya Sengupta frame the independence and accountability question?

In Independence and Accountability of the Higher Indian Judiciary, published in 2019, Arghya Sengupta argues that the Indian debate has been deformed by treating independence and accountability as opposites to be traded against each other. His institutional treatment finds that the higher judiciary secured strong independence from the executive in appointments and administration while leaving its internal accountability underdeveloped, and he proposes redesign rather than reversal: an appointment process that combines judicial participation with published criteria, recorded reasons and a permanent secretariat, so that choices can be examined without being transferred to the political executive. The book criticises the collegium’s closed working in detail and equally rejects the restoration of executive primacy, which is why it is cited by reformers on both sides of the argument that followed the judgment of 2015.

Q: Why do serving judges’ own criticisms of the collegium carry special weight?

Criticism of the collegium from politicians can be answered as the resentment of the branch that lost the appointment power in 1993, and criticism from scholars can be answered as theory. Criticism from within the bench removes both answers. Justice J. S. Verma, who wrote the majority judgment in the Second Judges Case, said in later years that the system it created had not worked as he had intended. Justice Ruma Pal, a former judge of the Supreme Court, described the process in 2011 as one of the best-kept secrets in the country. Justice J. Chelameswar recorded the collegium’s opacity in a judgment of the Supreme Court itself in 2015. Each statement came from a holder of the power being criticised, speaking against institutional interest, and each is cited in the debate for exactly that reason.