Social justice laws for UPSC Mains Paper Two frighten candidates out of proportion to their difficulty, and the fear itself is the first thing to correct. The marks in this area do not sit in the candidate’s opinion about reservation, untouchability or tribal land. They sit in a small, stable set of constitutional provisions, six statutes, three commissions and a dozen judgments, and in one skill that can be practised like any other: separating the constitutional question a paper is asking from the judicial question it is also asking and from the empirical question it is quietly asking underneath both. A candidate who answers those three questions as one produces an editorial. A candidate who separates them produces an analysis, and the difference between the two is the difference the marking rewards.

The anxiety has a specific shape, and naming it helps. Aspirant communities worry that an examiner will penalise a view, so preparation drifts toward memorised balance, a paragraph for and a paragraph against with no conclusion at all. That is the weakest structure available in this field. An answer that refuses to conclude reads as evasion, and an answer that endorses or denounces reads as advocacy. The strong structure concludes firmly on the question the evidence can settle, which is almost always a constitutional or institutional question, and names the remaining disagreement as a disagreement about values, which no candidate is required to resolve and no examiner can fairly demand be resolved in one direction. This article teaches that structure by building the material it operates on: the provisions that generate the questions, the four moments that organise the history, the judgment set with the precise proposition each case is for, the statute set as citation stock, the discipline for handling figures, and a revision scaffold that converts the whole field into six answer shapes.

Where this field sits in the paper, and what the paper rewards

The syllabus places this field inside General Studies Paper Two under two heads. The first is the Constitution head, which covers significant provisions, amendments and basic structure, and which generates the doctrinal questions on reservation, the ceiling and the commissions. The second is the welfare head, whose syllabus language asks for the mechanisms, laws, institutions and bodies constituted for the protection and betterment of vulnerable sections, and which generates the institutional and enforcement questions on the atrocities statute, the civil rights statute, tribal self-governance and sanitation. A third head, on social justice as a subject in its own right, draws on both. The same material reappears in Paper One under Indian society, where caste is examined as a social institution rather than a legal one, and in the preliminary examination as polity fact. One body of knowledge therefore serves four surfaces, which is why this cluster repays organisation out of proportion to its size.

The answer-structure method used here is the one this series applies in its other study guides, and a candidate revising across subjects should keep the method constant while the material changes. The method in the tax study article separates the constitutional competence question from the design question and the revenue question; the method in the agriculture study article separates the List question from the statute question and the implementation question. The reader who has worked through the tax laws study guide for the economy paper or the agriculture laws study guide will recognise the discipline: every demand in this field decomposes into layers, each layer has its own authorities, and marks are lost when authorities from one layer are offered as answers to another. A conviction-rate figure cannot answer a question about what Article 15 permits. A holding cannot answer a question about whether a statute works. A commission’s composition cannot answer a question about identification. Layer discipline is the whole technique, and the rest of this article is the stock the layers draw on.

What the paper rewards, stated plainly, is precision about holdings. Preparation material in this field is full of propositions that are nearly right: the ceiling stated as absolute, observations in a judgment described as its decision, a commission’s recommendation described as law, a pending report described as implemented policy. Each of these is a mark surrendered by a candidate who knew the area but not the exact status of the proposition. The examination register therefore requires a habit the rest of the series also enforces: state the proposition in one sentence, attach the case name only where the attribution is certain, and mark the status of everything, decided, observed, recommended, pending or repealed, as part of the sentence itself.

For the teacher, this field presents a distinct problem that preparation should anticipate, because aspirants arrive having absorbed a classroom version of it. The subject is politically charged, and it is commonly taught as a controversy to be summarised rather than a structure to be operated. The confusions worth pre-empting in any teaching of the field are the ones this article is built around: the ceiling taught as a number in the Constitution when the Constitution contains no number; the commissions taught as a list of names when their examinable content is the difference in their bases and powers; the judgments taught as outcomes when their content is propositions with statuses; and the statutes taught as prohibitions when their content is machinery. A candidate whose instruction has installed the structure can be given any live controversy in the field, a survey, a judgment, a commission’s report, and will locate it in the spine within a paragraph. A candidate taught the controversy must begin again with each new event. The examination rewards the first candidate, and the difference is pedagogical before it is intellectual.

The Essay paper draws on the same stock under a different discipline. Essay topics on caste, reservation and social justice recur across years in thematic form, and the temptation they create is to write the mains answer at greater length with warmer rhetoric. The three-layer separation disciplines the essay as it disciplines the answer: the constitutional position stated from text, the judicial position stated from holdings, the empirical dispute attributed, and the value disagreement named rather than adopted. What the essay adds is the historical and comparative dimension, the pre-history of communal orders and the Constituent Assembly’s change of basis, which this article supplies, and a conclusion that identifies the deciding institution. An aspirant who has built the anchor file for Paper Two owns the essay on this theme without separate preparation, which is a further return on the field’s compactness.

A word on register, because this article’s register differs from the rest of the cluster and the difference is itself examinable content. The guide article in this cluster writes as analysis addressed to a citizen. The examination register writes as a demonstration addressed to a marker, and its defining skill is the one the aspirant community finds hardest to credit: the ability to argue a position at full strength without disclosing whether it is the candidate’s own. That skill is not hypocrisy and it is not neutrality. It is the professional discipline of the advocate and the civil servant, both of whom must be able to state the case for a policy they did not choose, and the examination tests it deliberately by setting questions on subjects where every candidate is presumed to have a view. The technique taught in this article, separation of the layers, attribution of every position to a holder, conclusion on the tractable question, is the mechanism by which the skill is performed on paper. A candidate who performs it has answered the register question as well as the subject question, and the two are marked together.

The constitutional spine: the provisions that generate the questions

Almost every question in this field is generated by a short list of provisions, and the first revision task is to know what each one does, in one sentence each, without recourse to commentary.

Article 14 guarantees equality before the law and is the background norm against which every special provision is a stated exception rather than a contradiction. Article 15(1) prohibits discrimination on grounds only of religion, race, caste, sex and place of birth. Article 15(3) permits special provision for women and children. Article 15(4), inserted by the First Amendment after the Supreme Court struck down a communal allocation of medical seats in the Champakam Dorairajan case, permits special provision for the advancement of socially and educationally backward classes, the Scheduled Castes and the Scheduled Tribes. Article 15(5), inserted by the Ninety-third Amendment and in force from 20 January 2006, extends the special-provision power to admission in educational institutions including private institutions, aided or unaided, other than minority institutions, and it is the provision under which reservation reached the unaided private college. Article 15(6), inserted by the One Hundred and Third Amendment, permits special provision for economically weaker sections up to a maximum of ten per cent, in addition to existing reservation.

Two neighbouring provisions complete the equality code for examination purposes, and both are frequent prelims material. Article 29(2) provides that no citizen shall be denied admission into a state-maintained or state-aided educational institution on grounds only of religion, race, caste or language; it was the provision, read with Article 15(1), on which the communal order fell in Champakam Dorairajan, and it explains why the First Amendment’s answer had to be placed in Article 15 rather than left to implication. Article 30(1), guaranteeing minorities the right to establish and administer educational institutions of their choice, is the boundary provision of the field: it is the express exception in Article 15(5), so the special-provision power stops at the minority institution’s gate, and questions on the reach of reservation into private education are answered by reading the two provisions together. The word “only” in Article 15(1) and Article 29(2) also carries doctrine worth a sentence in an answer: a classification that rests on a listed ground among other relevant grounds is not, without more, discrimination on grounds only of that ground, which is the textual foothold for the treatment of caste as one factor in identification rather than a disqualification from considering it.

Article 16 runs parallel for public employment. Article 16(1) guarantees equality of opportunity. Article 16(4) permits reservation of appointments for a backward class which, in the opinion of the state, is not adequately represented in the services. Article 16(4A), inserted by the Seventy-seventh Amendment, permits reservation in promotion for the Scheduled Castes and the Scheduled Tribes. Article 16(4B), inserted by the Eighty-first Amendment, takes vacancies carried forward from an earlier year outside the fifty per cent ceiling for the year in which they are filled. The Eighty-fifth Amendment wrote consequential seniority into clause (4A), and the Eighty-second Amendment added a proviso to Article 335 permitting relaxation in qualifying marks and standards of evaluation for Scheduled Caste and Scheduled Tribe candidates in promotion. Article 16(6) is the employment counterpart of the economically weaker sections power. Article 335 itself, which requires the claims of the Scheduled Castes and the Scheduled Tribes to be taken into consideration consistently with the maintenance of efficiency of administration, is the text on which the promotion litigation turned, and a candidate should be able to say so.

Three further provisions carry whole question families. Article 17 abolishes untouchability and forbids its practice in any form; it is one of the very few fundamental rights enforceable against private persons, and the civil rights statute of 1955 is the penal legislation that gives it effect. Article 46, a directive principle, directs the state to promote with special care the educational and economic interests of the weaker sections, in particular the Scheduled Castes and the Scheduled Tribes, and to protect them from social injustice and all forms of exploitation; it supplies the purpose language for answers across the field. Articles 341 and 342 govern the specification of the Scheduled Caste and Scheduled Tribe lists by Presidential order, with variation thereafter only by Parliament, and Article 342A, inserted by the One Hundred and Second Amendment, does the same for the socially and educationally backward classes, which is why the Maratha litigation and the One Hundred and Fifth Amendment, restoring the states’ power to identify their own lists, belong in the same answer family.

The three commissions are the third leg of the spine, and confusing them is among the costliest errors in this paper. The National Commission for Scheduled Castes rests on Article 338 and the National Commission for Scheduled Tribes on Article 338A, inserted when the Eighty-ninth Amendment split the earlier combined commission in 2003. Both are constitutional bodies with powers of a civil court for inquiry, and both report to the President. The National Commission for Backward Classes began as a statutory body under an Act of 1993, created because the Indra Sawhney judgment required it, and became a constitutional body under Article 338B only when the One Hundred and Second Amendment took effect in 2018. Article 340, under which the President may appoint a commission to investigate the conditions of backward classes, is the power under which the Mandal Commission was appointed, and it must not be confused with Article 339, which concerns a commission on the administration of the Scheduled Areas and the welfare of the Scheduled Tribes and carries the ten-year language that preparation notes routinely attach to the wrong article.

When did the backward classes commission become a constitutional body?

The National Commission for Backward Classes became a constitutional body in 2018, when the One Hundred and Second Amendment inserted Article 338B. Before that it was a statutory body under an Act of 1993, created following the direction in Indra Sawhney. The Scheduled Caste and Scheduled Tribe commissions had been constitutional far longer, under Articles 338 and 338A.

One quantum doctrine completes the spine, because candidates persistently ask the text for a number it does not contain. Neither Article 15 nor Article 16 fixes any percentage. The figures in the field come from three different kinds of source, and an answer should keep the kinds apart: from executive and legislative choice within the judicial ceiling, as with the twenty-seven per cent of the 1990 memorandum and the shares in the central institutions statute; from the constitutional text itself only in the economically weaker sections clauses, where ten per cent is written as a maximum; and from the lists’ population proportions in the legislative reservation of Articles 330 and 332, the one place where proportionality is the textual rule. Article 16(4)’s condition is qualitative, a backward class which in the opinion of the state is not adequately represented, and the Court’s review of quantum has policed the ceiling and the creamy layer rather than the state’s assessment of adequacy within them. The candidate who understands that the Constitution distributes the quantum decision, to the state under a ceiling for services, to the text for the economic quota, and to population for legislatures, can answer any how-much question in the field without inventing a figure.

The final element of the spine is territorial. The Fifth Schedule, under Article 244(1), governs the administration of Scheduled Areas in ten states and gives the Governor powers, including over land transfer, that no ordinary state administration holds. The Sixth Schedule, under Article 244(2), creates autonomous district and regional councils with legislative powers in the tribal areas of Assam, Meghalaya, Tripura and Mizoram. The distinction between the two models, protective administration in the Fifth Schedule and self-governing institutions in the Sixth, is the constitutional frame into which the panchayats extension statute of 1996 fits, and it is the frame examiners test when a question speaks of tribal self-governance.

Two further pieces of the spine reward attention because they convert descriptive answers into institutional ones. The commissions were not born in their present form. Article 338 as enacted provided for a Special Officer for the Scheduled Castes and the Scheduled Tribes, an investigating and reporting office rather than a commission. The Sixty-fifth Amendment substituted a multi-member National Commission for the two groups together, and the Eighty-ninth Amendment divided that body into the two commissions of Articles 338 and 338A. The backward classes commission followed a different route, statutory first and constitutional later, because its creation was directed by a judgment rather than designed by the constituent process. All three commissions share the same structural limits, and those limits are the examinable content: each may inquire and investigate with the powers of a civil court for summoning and evidence, each reports to the President, whose office causes the report to be laid before Parliament with a memorandum of action taken, and none can issue a binding direction to a government. A question on the commissions is therefore never a question about their findings; it is a question about whether a recommendatory design can discipline an administration, and the strong answer argues from the reporting line and the inquiry powers rather than from any commission’s conclusions.

The list provisions have a mechanics of their own that candidates garble. The Scheduled Castes and Scheduled Tribes lists were first specified by Presidential orders in 1950, and Article 341 and Article 342 then lock the lists: the President specifies after consultation with the Governor of the state concerned, and thereafter only Parliament may include in or exclude from a list. A state legislature cannot add a caste to the Scheduled Castes list, a state executive cannot treat a community as listed by analogy, and a court cannot extend a list entry to a synonym the order does not contain. Article 342A applies the same architecture to the backward classes, with the complication that the One Hundred and Second Amendment’s text was read as creating a single central list and the One Hundred and Fifth restored the separate state lists that had existed in practice. The definitions in Articles 366(24) and 366(25), which tie the constitutional meaning of the two expressions to the Articles 341 and 342 lists, complete the point: in this field, membership of a protected class is a legal fact created by notification under a locked procedure, and that is why identification questions are institutional questions rather than sociological ones.

Legislative reservation completes the spine and is tested separately from service reservation, a distinction answers frequently blur. Articles 330 and 332 reserve seats for the Scheduled Castes and the Scheduled Tribes in the House of the People and the state assemblies in proportion to population, and Article 334 fixes the duration of that reservation, extended by successive amendments and taken to eighty years by the One Hundred and Fourth Amendment, which at the same time ended the nomination of Anglo-Indian members. The service provisions and the legislative provisions rest on different reasoning: the first on representation in the administration as an aspect of equality of opportunity, the second on representation in lawmaking as an aspect of political presence, and an answer that justifies one with the other’s arguments has committed a category error the examiner can see. The legislative line has a recent addition that the paper has already absorbed. The One Hundred and Sixth Amendment of 2023, the women’s reservation amendment, inserted Articles 330A and 332A reserving as nearly as may be one-third of the seats in the House of the People and the state assemblies for women, including within the seats reserved for the Scheduled Castes and the Scheduled Tribes, with its operation tied to the delimitation following the first census taken after its commencement. For this field its significance is structural rather than numerical: it extends the legislative reservation model beyond the two listed groups for the first time, it does so without a separate identification problem because the category is defined by sex, and its deferred commencement makes it the standing example of a provision whose status, enacted but not operative, must be stated as part of any answer that cites it. It also completes the Article 15(3) line for examination purposes: the special provision for women that began as a permissive clause in services and welfare has its largest expression in legislative reservation, reached seventy-three years later by a different route.

Article 46, finally, is the provision to cite when a question asks about purpose rather than power: it is unenforceable, as all directive principles are under Article 37, but it is the text the Court and the commissions themselves use to state what the protective scheme is for, and placing it at the head of an answer’s purpose paragraph, before the enabling clauses are analysed, is the conventional and rewarded order.

Four moments that organise the field

Chronology is the enemy of revision in this field because the material spans seventy years of amendments and litigation. Four moments organise it, and a candidate holding the four can place any provision, statute or controversy without memorising a timeline.

Beneath the four moments sits a pre-history the paper assumes. Reservation in India is older than the Constitution: the Madras Presidency’s communal government orders of the 1920s distributed public appointments among communities by quota, Mysore acted on the Miller Committee’s recommendations for its backward classes, and Kolhapur under Shahu Maharaj had reserved posts for non-Brahmin communities decades earlier. The Constituent Assembly worked in that inheritance and deliberately changed its basis, rejecting communal proportionality in favour of backwardness linked to the named classes, and the Champakam Dorairajan case that forced the First Amendment was itself a challenge to the Madras communal order surviving into the constitutional period. A candidate who knows the pre-history can answer the recurring question whether reservation is a constitutional innovation or a constitutional settlement of an older administrative practice, and the answer, that it is a settlement with its basis deliberately changed, is the kind of framing sentence that earns its place in an introduction.

The first moment is 1955, when the first statute giving penal effect to the abolition of untouchability was enacted, and with it the template that the whole field repeats: a constitutional promise in Article 17, a statute converting the promise into offences, and an enforcement record that becomes the standing criticism of both. The second moment is 1989 and 1990, when the atrocities statute was enacted for the Scheduled Castes and the Scheduled Tribes and the Mandal Commission’s recommendations were implemented for the backward classes in central services. The two events belong together in an answer because they mark the field’s transformation from a predominantly constitutional subject into a predominantly statutory and administrative one, and because the litigation of the following decade was litigation about both. The third moment is 1992 and the sequence from 1995 to 2001: the nine-judge judgment in Indra Sawhney that settled the architecture of reservation, followed by Parliament’s answers to it in the Seventy-seventh, Eighty-first, Eighty-second and Eighty-fifth Amendments, each directed at a specific holding on promotion, carried-forward vacancies, qualifying standards and seniority. The fourth moment runs from 2018 to 2024 and contains the atrocities amendment and the judicial response to it, the One Hundred and Second and One Hundred and Third Amendments, the One Hundred and Fifth Amendment’s correction on state lists, the economically weaker sections judgment, and the sub-classification judgment of 2024. A candidate who can assign any item in this field to its moment has the structure of most long answers before writing a word.

Each moment also has an internal logic worth one layer more of detail. The first moment begins before the statute: the Constitution’s own commencement in 1950 was preceded within a year by the Champakam Dorairajan decision striking down a communal seat allocation and by the First Amendment answering it with Article 15(4), so the pattern of judgment and amendment that defines the field is present from its first two years, and the 1955 statute is the same pattern in penal form, Article 17 having been left by its own terms to be enforced by law made by Parliament. The second moment turns on a delay: the Mandal Commission reported at the end of 1980, and its central recommendation waited a decade for the office memorandum of 13 August 1990, which confined the first phase of implementation to the castes common to the Commission’s list and the state lists. The memorandum of 25 September 1991 then modified the scheme by adding a preference for the poorer sections within the reserved category and a separate ten per cent quota on economic criteria, and the treatment of those two elements in the judgment that followed, one absorbed and one destroyed, is the hinge between the second and third moments. The third moment is best learned as a mapping rather than a narrative: the Seventy-seventh Amendment answers the holding against reservation in promotion, the Eighty-first answers the problem of unfilled reserved vacancies counting against the following year’s ceiling, the Eighty-second answers the constraint that Article 335 was thought to place on relaxed qualifying standards, and the Eighty-fifth answers the seniority consequence of accelerated promotion. A candidate who can run the mapping in either direction, holding to amendment and amendment to holding, holds the doctrinal core of the paper. The fourth moment is distinguished by speed and by the range of instruments: a parliamentary amendment answering a two-judge decision within months in 2018, two constitutional amendments on commissions and economic reservation in the same period, a corrective amendment in 2021, a five-judge resolution of the economic quota in 2022, and a seven-judge resolution of sub-classification in 2024. The lesson the fourth moment teaches for the examination is that the field’s open questions close through different instruments at different speeds, and an answer that identifies which instrument is competent to close the question in front of it, court, Parliament or commission, has already done most of the analytical work.

Between the third and fourth moments sits an interlude the four-part scheme should not be allowed to hide, because the paper tests it as heavily as any moment: the admissions settlement of 2005 to 2008. In three years the field acquired a new enabling provision in Article 15(5), a new central statute fixing admission shares, and a Constitution bench decision upholding both while exporting the creamy layer from services into admissions. The interlude matters structurally because it is the one episode in the field where the sequence ran amendment first and litigation second on an enabling design Parliament chose in advance, rather than amendment as a reply to a judgment’s prohibition; even here the pattern partly holds, since the amendment itself answered Inamdar. For revision the interlude is best held as a fifth, smaller moment with its own triad, Inamdar, the Ninety-third Amendment with the 2006 statute, and Ashoka Kumar Thakur, and a candidate who can state why it does not fit the reply pattern has understood the pattern rather than memorised it.

The judgment set, and what each case is for

A judgment earns its place in revision only if the candidate can state what it is for in one sentence. The set below is deliberately short. Everything else in this field is commentary on one of these decisions, an amendment answering one of them, or a later judgment applying one of them.

The ceiling comes from M. R. Balaji v. State of Mysore, decided in 1963, which held that caste could not be the sole criterion of backwardness and that reservation should ordinarily remain below fifty per cent. The candidate cites Balaji for the origin of the ceiling and for the proposition that caste is a factor, not a synonym, in identifying backward classes. The master case is Indra Sawhney v. Union of India, decided on 16 November 1992 by nine judges, six to three. It upheld twenty-seven per cent reservation for the backward classes in central services, required exclusion of the creamy layer, restated the fifty per cent ceiling as the governing rule subject to extraordinary exception, accepted caste as a legitimate starting point for identifying a backward class, struck down the additional ten per cent quota on an economic criterion alone, and held that reservation operates at initial appointment and not in promotion. Almost every doctrinal question in the paper is a question about one of those holdings, which is why the full treatment in the Indra Sawhney article on the ceiling and the creamy layer and the passage record in the Mandal Commission article carry the depth a candidate needs, and why this article states the holdings without re-arguing them.

Which case should a candidate master first in this field?

Indra Sawhney, decided on 16 November 1992 by nine judges sitting six to three. It is the source of the creamy layer rule, the fifty per cent ceiling in its operative form, the identification principles for backward classes, and the bar on reservation in promotion that four amendments then answered. No other judgment generates as many questions across this field.

The promotion sequence follows from Indra Sawhney’s bar. Parliament answered with the Seventy-seventh Amendment, and in M. Nagaraj v. Union of India in 2006 a Constitution bench upheld the promotion amendments while conditioning their exercise on quantifiable demonstration of backwardness and inadequacy of representation, consistency with efficiency under Article 335, and exclusion of the creamy layer. In Jarnail Singh v. Lachhmi Narain Gupta in 2018 the Court removed the requirement that the state collect quantifiable data to prove the backwardness of the Scheduled Castes and the Scheduled Tribes for promotion, holding that Nagaraj had not required what Indra Sawhney had already settled, while retaining the creamy layer condition for promotion within those groups. A candidate who can state what Nagaraj demanded, what Jarnail Singh withdrew and what it kept has the promotion question entire, and the commonest error in that answer is to report Jarnail Singh as having removed the creamy layer condition, which it did the opposite of.

The admission line has a link that candidates omit and examiners notice. In P. A. Inamdar in 2005 a seven-judge bench held that the state could not impose its reservation policy on unaided professional institutions, state or minority, treating such imposition as an unreasonable restriction on the institutions’ autonomy. The Ninety-third Amendment was Parliament’s answer, inserting Article 15(5) to enable precisely the provision Inamdar had placed beyond legislative reach, while expressly preserving minority institutions under Article 30(1). The central educational institutions statute followed on that enabling power, and Ashoka Kumar Thakur tested both the amendment and the statute together. The chain matters because it is the admission-side instance of the field’s master pattern, judgment, amendment, statute, judgment, and because it locates the minority institution exclusion in the amendment’s own text rather than in judicial solicitude, a precision answers routinely reverse.

Nagaraj itself became the subject of the next reference, which is how Jarnail Singh arose: a bench considering promotion reservation for the Scheduled Castes and the Scheduled Tribes doubted whether Nagaraj’s requirement of quantifiable data on backwardness could stand with Indra Sawhney’s treatment of those classes as identified by the lists themselves, and the five-judge bench in 2018 resolved the doubt by discarding the data requirement on backwardness while preserving the rest of the Nagaraj conditions. The sequence is the field’s cleanest example of doctrine correcting itself by reference, and it gives the candidate a usable rule for every promotion answer: the conditions that survive are inadequacy of representation, which the state must still show, the efficiency consideration of Article 335, and the creamy layer exclusion, while backwardness of the two listed groups is not the state’s to prove afresh in each case. Answers that reproduce Nagaraj’s original triad without the 2018 subtraction are six years out of date in substance, and the error is of the kind examiners describe as reliance on superseded notes.

Two admission-side cases complete the doctrinal set. Ashoka Kumar Thakur v. Union of India, decided on 10 April 2008, upheld the Ninety-third Amendment and the central educational institutions statute of 2006, applying the creamy layer exclusion to backward class admissions in central institutions; it is the authority for reservation in admissions as distinct from appointments. Janhit Abhiyan v. Union of India, decided in 2022 by three judges to two, upheld the One Hundred and Third Amendment, and it is the authority for two propositions that changed how the ceiling must be stated: that reservation on an economic criterion is constitutionally permissible, and that the fifty per cent ceiling is not an inflexible part of the basic structure, so the economically weaker sections quota standing above the ceiling does not fall for that reason. The dissent, which treated the exclusion of the already-reserved classes from the new category as a basic structure violation, is examinable as the counter-position and should be attributed as a dissent, not blended into the holding.

The Maratha sequence teaches the list question. In Jaishri Laxmanrao Patil v. Chief Minister of Maharashtra, decided on 5 May 2021, the Court struck down the Maharashtra statute granting Maratha reservation for exceeding the ceiling without extraordinary circumstances, and a majority read the One Hundred and Second Amendment as withdrawing the states’ power to identify socially and educationally backward classes. Parliament answered with the One Hundred and Fifth Amendment restoring that power. The sequence is the cleanest modern illustration of the dialogue between Court and Parliament that the third and fourth revision moments describe, and it belongs in any answer on federalism and identification.

What did the seven judges decide on sub-classification in 2024?

In Davinder Singh, decided on 1 August 2024 by seven judges, six to one, the Court held that states may sub-classify within the Scheduled Castes and the Scheduled Tribes to favour the more disadvantaged groups inside the lists, provided the classification rests on empirical material, not executive preference alone. The creamy layer observations of four judges were not the holding.

The sub-classification judgment, cited as 2024 INSC 562, overruled the contrary position in E. V. Chinnaiah and requires two further precisions that preparation material routinely loses. The majority permitted sub-classification for reservation in appointments and admissions; it did not alter the Presidential lists, and a state may not use sub-classification to add or remove a caste from them. And the views expressed by four of the judges favouring application of a creamy layer within the Scheduled Castes and the Scheduled Tribes were observations offered for future consideration, not part of the operative decision, which is why the widespread claim that the judgment applied the creamy layer to the Scheduled Castes is wrong and why examiners probe for exactly that error.

Three cross-cutting lines inside the judgment set deserve separate revision because questions are framed on the lines rather than on the individual cases. The first is the creamy layer line, which runs through four decisions and changes character as it runs. In Indra Sawhney the exclusion is a condition of validity for backward class reservation, a means test applied to a caste-based scheme to keep it within Article 16(4). The executive criteria followed in the office memorandum of 8 September 1993, and the income ceiling within those criteria has been revised by executive order over time, standing at eight lakh rupees in the revision of 2017 on the delivered record. In Ashoka Kumar Thakur the same exclusion was applied to admissions in central institutions, so the line crosses from services to education. In Nagaraj the exclusion was carried into promotion for the Scheduled Castes and the Scheduled Tribes as one of the conditions on the promotion amendments, and in Jarnail Singh it was expressly retained for that context while the backwardness data condition was removed. The line therefore ends in a differentiated position: creamy layer exclusion is a condition for the backward classes generally, a condition for the Scheduled Castes and the Scheduled Tribes in promotion, and an undecided proposal, supported by observations in 2024, for those groups generally. Any answer that states the position at a higher level of generality than that has lost the line’s whole point.

The second line is the basic structure thread, which examiners use to test whether a candidate understands what kind of limit the ceiling is. The framework comes from Kesavananda Bharati in 1973, under which constitutional amendments are reviewable against the Constitution’s basic structure, with equality among its features. Indra Sawhney worked inside that framework without treating the fifty per cent figure as itself a basic structure element: the figure is a working rule derived from the balance between reservation and the open category, and the majority contemplated extraordinary departure from it. The Maratha bench applied the figure strictly on the facts before it while the larger question of its status remained. Janhit Abhiyan then squarely presented the argument that breaching the ceiling for the economically weaker sections quota violated the basic structure, and the majority rejected it, holding the ceiling not inflexible in that sense, while the dissent located the violation not in the number but in the exclusion of the reserved classes from the new category. The thread’s lesson for answer writing is that basic structure in this field operates on the design of a category, its criteria and its exclusions, rather than on arithmetic, and answers that treat the number fifty as a basic structure talisman misdescribe every decision in the line.

The third line is the dissents, which this paper rewards when they are attributed precisely and punishes when they are blended. Three dissents carry examination weight. In Janhit Abhiyan, the dissent of Bhat J., with the Chief Justice concurring, would have struck the amendment down for excluding the Scheduled Castes, the Scheduled Tribes and the backward classes from the economically weaker sections, treating that exclusion as discrimination dressed as classification; it is the source of the strongest form of the argument against the category’s design. In Davinder Singh, Bela Trivedi J. dissented alone against sub-classification, on the ground that the Presidential lists are a single indivisible class for reservation and that grading within them is a power the Constitution reserves to Parliament through the list procedure; her reasoning is the strongest form of the indivisibility argument and should be cited as the dissent it is. In the Maratha case the division ran the other way on one question, three judges to two holding that the One Hundred and Second Amendment had withdrawn the states’ identification power, a reading Parliament then reversed by amendment; it is the standing example of a majority proposition in this field being undone by the amending power rather than by a later bench. A candidate who can place these three dissents, their authors, their ratios and their fates, commands the argumentative range of the field, because the majority positions are in every set of notes and the dissents are where preparation is thinner.

Citation practice in the hall follows from the status discipline. The candidate cites a case for a proposition, and the citation earns its place only where the proposition is the case’s own: the name, the year and, where it matters, the bench strength and the division. Bench strength is not decoration in this field. A proposition from a nine-judge bench closes an argument; the same proposition from a two-judge bench opens one, and the atrocities sequence turns entirely on that difference, a two-judge reading of a statute answered by Parliament and then reviewed by a bench of three. Where a neutral citation exists, as with the 2023 and 2024 judgments cited in this article by their INSC numbers, it is the form least likely to be mistyped from memory, and a candidate uncertain of a report citation should give the name and year with confidence rather than a volume number with doubt, since a wrong citation is an invented fact under examination conditions exactly as it is in print. The proposition is the marks; the citation is its credential, and a credential that cannot be verified should not be offered.

One discipline closes the judgment set: knowing what each case is not authority for. Balaji is not authority for a rigid numerical cap; it spoke of a broad proposition about the permissible extent of reservation in the context of a state order it struck down on other grounds. Ashoka Kumar Thakur is not authority on state institutions or on unaided institutions generally; it decided the central statute and the amendment behind it. Mahajan is not the settled reading of the atrocities statute; its directions were recalled in review, and the position in the field is the one the review and Prathvi Raj Chauhan left. Chinnaiah is not authority against sub-classification; it was overruled on the point by the larger bench in 2024, though it remains the clearest statement of the position that lost. Stating the negative proposition for each leading case, in one clause, is a revision exercise that prevents the most expensive class of error in this paper, the deployment of a famous name for a proposition it does not support.

The atrocities sequence teaches the boundary between interpretation and correction. In Dr. Subhash Kashinath Mahajan v. State of Maharashtra, decided on 20 March 2018, a two-judge bench read safeguards into the atrocities statute: no absolute bar on anticipatory bail, a preliminary enquiry before registration, and approval before the arrest of a public servant. Parliament answered with the Amendment Act 27 of 2018, inserting Section 18A to restore the statute as enacted. In the review judgment of 1 October 2019 the Court recalled the 2018 directions, and in Prathvi Raj Chauhan v. Union of India on 10 February 2020 it upheld Section 18A while clarifying that anticipatory bail remains available where a complaint discloses no prima facie offence under the Act. The full sequence, and what it shows about which institution has the last word on a protective statute’s procedure, is treated in the article on the 2018 amendment and the Court; for examination purposes the sequence is the standard answer to any question about judicial legislation, because each institution’s move and its justification can be stated in a sentence.

Identification as a method: Kalelkar, Mandal, Rohini

Identification is the field’s oldest unsolved problem, and the paper tests it as a history of methods rather than as a list of results. Three commissions define the history, and each is examinable for the method it chose and the fate the method met.

The first backward classes commission, chaired by Kaka Kalelkar, was appointed on 29 January 1953 and reported on 30 March 1955. It identified 2,399 castes as backward, of which 837 were classed as most backward, working substantially through caste as the unit of social position. Its fate is the examinable point: the report was not acted upon, and the chairman’s own letter transmitting it recorded his unease at caste as the criterion, a hesitation the Union government of the day shared in declining to build a central list upon it. The Kalelkar episode teaches the candidate that identification by caste was contested from the first attempt, by the very commission that performed it, and that the states’ own lists consequently became the operative instruments for a generation.

The Mandal Commission, appointed in 1979 and reporting at the end of 1980, was built to answer the Kalelkar hesitation with method. It constructed eleven indicators of backwardness in three groups, four social, three educational and four economic, weighted at three points for each social indicator, two for each educational and one for each economic, so that a community could score twenty-two points at most, and it treated communities scoring above half the available points as backward. The weighting is the answer to the criticism that the Commission merely counted castes: social indicators dominate the score by design, economic indicators count least, and caste functions as the unit to which the indicators are applied rather than as the criterion itself, which is the construction Indra Sawhney later accepted when it held caste a legitimate starting point rather than a forbidden one. The Commission’s population estimate of fifty-two per cent, and the gap between that estimate and the twenty-seven per cent it recommended, belong to the same analysis: the recommendation was set by the ceiling arithmetic of the period and by the exclusion of the shares already reserved, not by the estimate, and a candidate who can explain why a commission that found fifty-two per cent recommended twenty-seven has understood the interaction of identification with the ceiling better than most commentary states it.

The Rohini Commission represents the method’s third generation: identification is settled for the central list, and the question has become distribution within it. Its remit was to examine the inequitable distribution of reservation benefits among the communities on the central backward classes list and to work out a scientific basis for sub-categorisation, using the administrative records of admissions and appointments because no population instrument exists for the purpose. The progression across the three commissions is itself an answer structure. Kalelkar asks which communities are backward and answers by social position. Mandal asks the same question and answers by a weighted index. Rohini asks a different question, which communities inside the recognised class the benefits reach, and answers, when its report is published, from records. A candidate who narrates the progression as a change in the question, rather than as three attempts at the same question, has the identification answer at the level the paper sets it.

A thread runs beneath the identification history that answers should surface, because it connects the field’s beginning to its present state. The Constitution’s own design anticipated revision: Article 339 contemplated a commission on the Scheduled Areas and tribal welfare at the expiration of ten years from commencement, and the identification judgments absorbed the same instinct, Indra Sawhney contemplating that the lists and the creamy layer criteria would be revisited periodically as conditions changed, since a classification true of one decade hardens into privilege in the next if it is never re-examined. The field’s practice has honoured the instinct irregularly. The creamy layer income ceiling has been revised by executive order at intervals, the lists have been amended by Parliament for inclusion and exclusion of particular communities, and the two standing commissions review safeguards continuously, but no general decennial re-identification of the backward classes has been performed since Mandal, which is why a commission appointed in 2017 was still working from a list architecture built in the 1990s. The examinable proposition is that the scheme assumes its own revision and supplies the instruments for it, commission, list amendment and criteria revision, and that the politics of the field consists substantially in the decision whether to use them. That proposition, placed at the end of an identification answer, converts a historical recital into an institutional analysis.

The promotion question, stated whole

Promotion is the doctrinal thread candidates most often learn in fragments, and it repays one consolidated statement. Indra Sawhney confined reservation to initial appointment for a reason internal to its logic: reservation was justified as securing representation of a class in the services, and the majority treated that justification as exhausted at entry, leaving advancement to the ordinary rules of seniority and merit as the Constitution’s text in Article 16(1) was read to require. The Seventy-seventh Amendment displaced that reading for the Scheduled Castes and the Scheduled Tribes by inserting Article 16(4A), and the seniority consequence immediately generated the next problem: a reserved candidate promoted earlier would, on ordinary seniority principles, carry that seniority upward permanently, overtaking general candidates senior at entry. The Court’s answer in the Virpal Singh Chauhan and Ajit Singh line was the catch-up rule, under which a general candidate who is promoted later recovers seniority over the reserved candidate promoted earlier. The Eighty-fifth Amendment displaced the catch-up rule in its turn by writing consequential seniority into Article 16(4A) with retrospective effect from 17 June 1995, the date of the Seventy-seventh Amendment itself. Nagaraj then upheld the whole architecture on its conditions, and Jarnail Singh pruned the conditions, as the judgment section records. Stated whole, the thread is four institutional moves and two judicial settlements in thirty years over a single question, the rate at which representation at entry should convert into representation at the top, and it is the field’s best illustration that constitutional doctrine in this area is made by alternation between bench and Parliament rather than by either alone.

The statute set as citation stock

Statutes enter an answer as citations of mechanism: the candidate names the statute to show what machinery exists, not to narrate its contents, which belong to the statute articles in this cluster. Six statutes constitute the stock, and each has one examination function.

The Protection of Civil Rights Act, Act 22 of 1955, in force from 1 June 1955, is the penal statute under Article 17. Enacted as the Untouchability (Offences) Act, it was renamed and its penalties recast by Act 106 of 1976 with effect from 19 November 1976; the Section 3 scale of one to six months’ imprisonment with a fine of one hundred to five hundred rupees is a 1976 scale, and a candidate who attributes it to 1955 has repeated the commonest dating error in this statute’s literature. Its examination function is the gap: a prohibition drawn in the widest formal terms with a minimal enforcement record, which the delivered cluster documents from the ministry’s own reporting. The Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, Act 33 of 1989, in force from 30 January 1990, is the field’s principal protective statute, strengthened by the amendment Act 1 of 2016, in force from 26 January 2016, which widened the offence list and added victim and witness rights, and by the 2018 amendment described above. Its examination function is the enforcement question: special courts, exclusive special courts, and the distance between the statutory scheme and disposal outcomes.

The civil rights statute’s internal architecture has one feature candidates should be able to name, because it distinguishes the statute from ordinary penal law. Its offences, set out in Sections 3 to 7A, punish the enforcement of disabilities arising out of untouchability across worship, access to public places and services, refusal of sale and services, and insult, and Section 12 raises a presumption that an act was done on the ground of untouchability where it relates to a member of a Scheduled Caste, placing the evidential initiative on the accused in the manner protective statutes use when the prohibited ground is the very thing least likely to be documented. The presumption is part of the original design in substance, as the delivered pillar establishes from the case law, and not a later graft, which is the precision its literature usually misses.

The tribal statutes divide by model. The Panchayats (Extension to the Scheduled Areas) Act, Act 40 of 1996, assented to on 24 December 1996, extends Part IX of the Constitution to the Fifth Schedule areas subject to the Gram Sabha’s competence over consultation before land acquisition, minor minerals, minor forest produce and village markets; its examination function is self-governance within a protective frame. The Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act, Act 2 of 2007 under its 2006 short title, in force from 31 December 2007, recognises individual and community forest rights and makes the Gram Sabha the initiating authority for determining them; the Supreme Court’s Niyamgiri judgment of 2013, in which the Gram Sabha’s decision on religious and cultural rights in the bauxite hills was treated as determinative, is its companion case, and Samatha in 1997, voiding mining leases to non-tribals in Scheduled Areas, is the earlier authority on tribal land alienation that answers still cite.

How do the Fifth and Sixth Schedule models differ for a tribal governance answer?

The Fifth Schedule is protective administration: the Governor holds powers, including over land transfer and moneylending, and the Tribes Advisory Council advises, while ordinary institutions govern. The Sixth Schedule is self-government: autonomous district councils in four north eastern states legislate on land, forests and custom. The 1996 panchayats extension statute adds Gram Sabha competence inside the Fifth Schedule model only.

The dignity statute is the Prohibition of Employment as Manual Scavengers and their Rehabilitation Act, Act 25 of 2013, assented to on 18 September 2013 and in force from 6 December 2013, preceded by a statute of 1993, Act 46 of 1993, that addressed dry latrines without a rehabilitation architecture. Its examination function is the conversion of a sanitation question into a dignity question under Articles 17 and 21, carried by the compensation jurisprudence: ten lakh rupees for sewer deaths since 1993 in Safai Karamchari Andolan in 2014, and thirty lakh rupees in the judgment of 2023, cited as 2023 INSC 950, which also quantified the hazardous cleaning death record from the material before the Court. The Central Educational Institutions (Reservation in Admission) Act, Act 5 of 2007 under its 2006 short title, fixes the admission shares in central institutions, fifteen per cent for the Scheduled Castes, seven and a half for the Scheduled Tribes and twenty-seven for the backward classes in Section 3, and is the statute Ashoka Kumar Thakur upheld; its function is to let a candidate move from the enabling provision, Article 15(5), to the operating instrument in one step.

One design feature of that statute is worth the candidate’s attention because it recurs in policy questions. The shares were coupled with an expansion of the total seats in the covered institutions, phased over three years, so that the number of seats available outside the reserved categories was not reduced by the introduction of the backward class share. The device converted a zero-sum controversy into an expansion programme, at the price of an infrastructure and faculty burden that the implementation record then debated. The examination use is precise: when a question asks how reservation can be extended without displacing existing expectations, the seat-expansion device is the field’s standing answer, and when a question asks why the 2006 statute was litigated on the creamy layer rather than on its existence, the answer is that the enabling amendment had settled existence and left design, which is where Thakur was fought. The implementation was supervised through an oversight committee chaired by Veerappa Moily, whose phased seat-expansion schedules the institutions followed, and the episode as a whole is the field’s clearest instance of a reservation measure administered as an expansion programme rather than a redistribution, which is why it is cited in policy argument by both sides: by those who present expansion as the solvent of the ceiling conflict, and by those who answer that expansion at this scale was purchasable only once, in a period of rapid growth in public higher education, and cannot be the general answer to a distributive question.

The enforcement record of the statute set is itself examinable content, and it should be learned in the attributed form in which the cluster’s statute articles present it. For the civil rights statute, the ministry’s annual reporting under Section 15A of that Act records a register of cases so small that the delivered pillar article treats disuse as the statute’s central fact: thirteen cases recorded for 2022 in that series, against twenty-four for 2021 and twenty-five for 2020, with disposal figures showing a single conviction across the reported disposals and pendency above ninety-seven per cent. The contrast the pillar draws, and the contrast an answer should draw, is with the atrocities statute’s reporting for the same period, which records cases in the tens of thousands annually, the ministry’s compilation for 2022 giving 51,656 cases for the Scheduled Castes and 9,735 for the Scheduled Tribes. The two statutes thus present opposite examination problems under the same constitutional promise: one prohibition so complete in form that it is barely invoked, the other so heavily invoked that the system’s capacity to dispose of its cases is the question. A candidate who can state that contrast, with the source of the figures named, has the enforcement paragraph of most answers in this field.

The atrocities statute’s machinery rewards one further level of precision. The Act works through special courts designated under Section 14, and the amendment of 2016 added exclusive special courts for districts with a higher incidence of offences, together with a chapter on the rights of victims and witnesses in Section 15A, covering protection, information, and relief at each stage. Relief itself is delivered under the Rules of 1995 through a staged compensation scale paid by the district administration, a design that makes the district magistrate and the superintendent of police, rather than the court alone, the statute’s operating officers, and makes the vigilance and monitoring committees the Act prescribes at district and state level the institutional forum in which its working is reviewed. Those features are the substance of an institutional answer on the statute: the candidate who names the machinery can discuss why outcomes diverge from the text without resorting to general claims about implementation, and the candidate who cannot is left with the text alone, which the pillar article demonstrates is the weaker half of the subject. The relief scale illustrates the same point about where operative law lives in this field. The sums a victim receives are fixed not in the Act but in its Rules of 1995, revised in 2016 to a scale running, on the pillar’s verification, from eighty-five thousand rupees to eight lakh twenty-five thousand rupees according to the offence, so the provision a practitioner consults and the provision a citizen feels are subordinate legislation. A candidate should generalise the observation carefully: across this cluster, commencement notifications, rules and office memoranda carry operative content that the parent text only frames, the 1990 and 1991 memoranda in the Mandal sequence being the field’s largest example, and an answer that reads only the Acts has read the skeleton of the law rather than the law.

For the tribal statutes, the record has the same shape. The panchayats extension statute required every state with Scheduled Areas to bring its panchayat legislation into conformity within a year of its commencement, and the rules and conformity position across the ten states remained incomplete for decades on the account the ministry itself gives in Parliament, with the pillar article recording the ministry’s statements that eight states had notified rules by late 2022 and two of the largest, Jharkhand and Odisha, accounted for differently in that record. The forest rights statute’s record turns on the claims process: recognition depends on the Gram Sabha’s initiation and the committees above it, rejection rates have been the standing controversy in the ministry’s status reporting, and the fragility of recognised and rejected claims alike was displayed in 2019, when an eviction direction in the Wildlife First proceedings, founded on rejected claims, was stayed by the same bench within weeks after the scale of the affected households became clear. The examination use of both episodes is identical: they show that in the tribal statutes the operative question is not what the right is but which institution’s determination makes it effective, and that question is institutional, which is why it recurs.

The manual scavenging statute’s judicial record supplies the data discipline’s best worked example. In Balram Singh v. Union of India, decided in 2023 and cited as 2023 INSC 950, the Court recorded from the material placed before it 1,035 deaths in hazardous cleaning since 1993, with compensation paid to the families in 948 of them, and recorded at the same time that no death was reported as resulting from manual scavenging as the 2013 Act defines it. The pillar article on that statute identifies the apparent contradiction as a definitional artefact: the deaths occur in sewers and septic tanks, which the Act treats as hazardous cleaning, while manual scavenging in the statutory sense is the handling of human excreta from insanitary latrines, a practice the surveys address separately. A candidate who reproduces the two counts without the definitional point has misunderstood the field’s most instructive statistic; a candidate who explains why both counts can be accurate has demonstrated exactly the handling of figures this paper rewards.

A last discipline on the statute set concerns commencement, because the cluster’s statutes display every commencement pattern the series knows and the examination tests the differences. The civil rights statute was assented to and published on 8 May 1955 and commenced on 1 June 1955, a short clean interval. The atrocities statute is dated 1989 by its assent and commenced on 30 January 1990. The forest rights statute carries a 2006 short title, an assent at the end of December 2006, the number of an Act of 2007, and a commencement a full year after assent, on 31 December 2007, with its rules following. The panchayats extension statute had no commencement clause at all and so operated from assent. The lesson generalises beyond this cluster and is worth stating in an answer wherever a statute’s working is in issue: the year in a short title is the year of the legislative session that produced the bill, not a statement of when the law began to bind, and arguments about a statute’s age, delay or retrospectivity that run from the short title rather than the commencement notification are built on the wrong date.

These six statutes sit inside the wider map drawn by the social justice laws guide, whose three-layer frame, constitutional permission, statutory machinery and enforcement record, is the organising structure this article assumes and the examination scaffold below operationalises.

The distinctions that get tested

Template knowledge in this field is mostly distinction knowledge. The paper returns to a small set of pairs whose members candidates merge, and each pair below is stated the way an answer should state it, as a difference in legal operation rather than a difference in emphasis.

A special provision is not a reservation. Articles 15(4) and 16(4) are enabling provisions of different widths: the first permits any special provision for the advancement of the named classes, of which reservation is one species alongside scholarships, hostels, fee concessions and remedial programmes, while the second permits reservation of appointments specifically, conditioned on inadequate representation. An answer that treats the two clauses as interchangeable loses the reason the First Amendment was drafted widely and the employment clause narrowly, and loses the further point that a special provision which is not a reservation is not measured against the fifty per cent ceiling at all, since the ceiling governs the reserved share of a competitive field. Reservation is also not a right: the provisions are enabling, the Supreme Court has repeatedly treated them as conferring no fundamental right to claim reservation, and a state that declines to reserve acts within the text, which is the proposition that decides a recurring prelims statement and disciplines mains answers that write as though the quotas were entitlements enforceable against a reluctant government.

Sub-classification is not sub-categorisation, and the difference has become a difference in legal status. Sub-classification, after Davinder Singh, is the judicially approved grading of groups within the Scheduled Caste and Scheduled Tribe lists by a state, for the distribution of the reserved share, on empirical material. Sub-categorisation is the corresponding proposal for the backward classes, examined by the Rohini Commission, which has reported without publication and without implementation. The first is decided law with conditions; the second is pending policy with a completed report. An answer that uses the words interchangeably, or that transfers the 2024 holding to the backward classes, commits the precise error the 2024 paper’s discourse was built around, and the distinction costs one clause to state and saves the answer.

Identification is not enumeration. Enumeration counts a population; identification determines which communities belong to a constitutionally recognised class. The Constitution assigns identification to lists notified through the locked procedure of Articles 341, 342 and 342A and to commissions appointed under Article 340, not to the census, which is why the absence of a published caste count since 1931 does not paralyse the scheme and why a census caste count, if published, would inform but not perform identification. The Mandal Commission’s work was identification built on an enumeration base two generations old, supplemented by field inquiry through its eleven indicators, and the examinable criticism of that work, that its social indicators were impressionistic and its economic data derivative, is a criticism of identification method, not of the percentage the method produced.

Within promotion, two conditions are persistently merged and must be kept apart. The creamy layer condition excludes individuals: it asks whether this candidate, by family income and status, belongs to the advanced segment of the class. The data conditions of the Nagaraj framework address the class: whether the class is inadequately represented in the cadre and, until Jarnail Singh removed it, whether quantifiable material showed its backwardness. An answer that applies the creamy layer as a class-level test, or that treats inadequacy of representation as a means test on individuals, has crossed the two, and the crossing is visible to any marker who knows the promotion line. The distinction also explains the sequence’s logic: the individual exclusion survived every pruning because it disciplines distribution within the class, while the class-level data burden on backwardness fell because the lists had already performed that determination.

A commission of inquiry is not a standing commission. The Mandal and Rohini Commissions were appointed under Article 340 for a defined task, reported, and dissolved; the three national commissions are permanent constitutional bodies with continuing inquiry jurisdiction over the administration of safeguards. Candidates merge the two kinds because both are called commissions and both produce reports, but their constitutional logic differs: the Article 340 body exists because the ordinary machinery is thought insufficient to generate the knowledge a new policy needs, while the standing bodies exist to police safeguards that already exist. Questions on institutional design are answered correctly only when the kind is identified first.

The civil rights statute and the atrocities statute are not interchangeable charges, and their boundary is a tested distinction of a practical kind. The Act of 1955 punishes the enforcement of untouchability disabilities against any person, and it does not require the victim to belong to a Scheduled Caste or a Scheduled Tribe, because Article 17’s prohibition is general. The Act of 1989 requires the victim’s membership of one of the two groups and an offender who is not a member of them, and in return it carries the graver offences, the special courts and the relief machinery. The same street conduct can therefore fall under either statute or both, and the choice of charge is a decision with procedural consequences, which is why the pillar articles treat the boundary between the two as a subject in itself and why an answer that speaks of atrocity law without distinguishing the instruments has merged what the law keeps apart.

Finally, the two protections of the tribal cluster are not one protection. The Fifth Schedule restrains the ordinary law’s operation in Scheduled Areas through gubernatorial power and protects land through regulation of transfer and moneylending; it does not create rights in the inhabitants enforceable in the manner of the forest rights statute, which vests defined individual and community rights and prescribes the authority that determines them. The panchayats extension statute adds a third mechanism, democratic rather than protective or vesting, by making the Gram Sabha competent in specified subjects. An answer on tribal governance that runs the three together as tribal protection has missed that the Constitution and Parliament chose three different instruments for three different failures: administrative excess, historical dispossession, and political exclusion.

How the questions are framed: the four demand types

Questions in this field arrive in four recurring shapes, and recognising the shape is the fastest mark available to a prepared candidate, because each shape has its own authorities and its own failure mode.

The doctrinal demand asks what the law permits: whether a ceiling may be crossed, whether a category may be constructed by exclusion, whether a state may grade within a list. Its authorities are the text and the holdings, its structure is the three-layer separation in its pure form, and its failure mode is the importation of enforcement material, which cannot answer a permission question. The identification demand asks who belongs and who decides: its authorities are the list provisions, the commission reports and the identification holdings from Indra Sawhney to Davinder Singh, and its failure mode is the treatment of population shares as entitlements, the move from a group’s estimated share of population to a claimed share of reservation that no decision supports as a rule. The enforcement demand asks why a statute’s outcomes diverge from its text: its authorities are the statute’s machinery and the attributed reporting, its strongest form assigns the divergence among drafting, institutional design and administrative capacity, and its failure mode is the assertion of figures without source or the conclusion that a badly enforced statute is a badly conceived one, which does not follow. The institutional demand asks whether a body’s design fits its task: its authorities are the composition, appointment, powers and reporting provisions of Articles 338 to 340 and the statutes beneath them, and its failure mode is the recital of a body’s functions without an account of its limits, since in this field every institution’s limits are the analytical content.

The verbs carry the same information in compressed form. A demand to discuss requires the dimensions of a question to be set out without a forced verdict; a demand to examine requires the parts to be tested against the record; a demand to examine critically, or critically examine, requires both cases at full strength and a verdict the candidate owns; a demand to comment requires a defended position in fewer words. In this field the critical form is the commonest mains shape, and it is the one the anxious candidate most often flattens into summary. The scaffold in this article is built for the critical form: provisions and holdings first, the strongest opposing cases second, the empirical dispute attributed third, and the conclusion taken on the tractable layer last.

A fifth shape, the comparative demand, cuts across the other four and has grown with the list litigation. It asks the candidate to compare positions across the federation: a state list against the central list after the One Hundred and Fifth Amendment, a Sixth Schedule council against a Fifth Schedule Gram Sabha, one state’s sub-classification design against another’s. Its authorities are the same provisions read at two levels, and its discipline is symmetry: the same criteria applied to both sides of the comparison, with the constitutional allocation of the power stated first, since most comparative questions in this field are decided by the allocation before the merits are reached. The Maratha sequence is its model instance, a state statute, a centralising amendment, a judgment reading the amendment, and a corrective amendment, in which the comparison between the state’s position in 2018 and its position in 2021 is the whole answer.

One framing habit completes the picture. The paper in this field prefers the live institutional question to the historical one: not what the Mandal Commission recommended, which is a prelims fact, but whether identification should rest with commissions, legislatures or courts; not what the atrocities statute prohibits, but whether its procedure should carry the safeguards ordinary criminal procedure carries; not whether reservation is justified in the abstract, but whether a specified design feature, the ceiling, the exclusion, the grading, survives the tests the Court has built. A candidate revising for the live form of each question, with the history held as the evidence for it, is revising in the direction the paper moves.

The empirical layer: what to do with figures

Every question family in this field has an empirical dispute attached to it, and the dispute is part of the answer, but the figures are the trap. Representation percentages, conviction rates under the atrocities statute, counts of manual scavengers, rejection rates under the forest rights statute: each of these numbers exists in several published versions, compiled on different definitions for different periods by different authorities, and a candidate who asserts one without its source and period has handed the examiner an error. The discipline the enforcement articles in this cluster practise, and the one to reproduce in the examination hall, is to reference the category of evidence and attribute it rather than assert the figure: to write that conviction outcomes under the atrocities statute form a small fraction of disposals in the National Crime Records Bureau’s annual reporting, that pendency is high and acquittals are driven substantially by witness hostility and investigation quality on the record of the ministry’s own annual reports, and that the figures vary by state and year. That sentence is defensible. A bare percentage is a wager.

The pattern of outcomes under the protective statutes is the one empirical subject on which an answer is expected to say something definite, and it can be said definitely at the level of mechanism. The delivered pillar on the atrocities statute, working from the crime records and the ministry’s reporting, attributes the gap between registration and conviction substantially to the investigation stage, to delay and hostility among witnesses in the interval before trial, and to the capacity of the special courts relative to the volume the statute generates; it does not attribute the gap to the statute’s substantive breadth, which is the answer the enforcement question is usually fishing for. Representation data has the same structure. The personnel ministry publishes the representation of the Scheduled Castes, the Scheduled Tribes and the backward classes in central government services as part of its annual reporting, and the examinable use of that material is to identify the level at which representation concentrates and the levels, particularly the higher grades, at which it thins, rather than to quote a percentage whose year and coverage the candidate cannot state. In both subjects the discipline is identical: the mechanism and the direction, attributed; the figure, only where its source and period travel with it.

Three data questions recur often enough to need their own settled handling. The caste data question arises because the state does not publish a caste count of the population: the Mandal Commission’s estimate that the backward classes formed fifty-two per cent of the population was constructed from the census of 1931, the last published caste enumeration, by taking the enumerated non-Hindu share and a residue calculation for Hindu communities, and preparation notes that present fifty-two per cent as a counted fact misstate its status. The caste enumeration conducted alongside the census of 2011 through the socio-economic and caste census collected caste returns that were never released as a caste table, the published output being confined to the socio-economic indicators, so the state’s position is that it holds an unreleased enumeration whose reliability its own custodians have questioned in public debate, and a state government caste survey, published by Bihar, revived the national question without settling it, because a state survey is evidence about one state compiled on one method. The correct examination handling is to identify the absence of an authoritative count as itself the institutional fact, and to connect it to the identification debate: the Constitution commits identification to lists and commissions precisely because enumeration is contested. The second question is the sub-categorisation report: the commission chaired by Justice G. Rohini, appointed under Article 340 in 2017 to examine sub-categorisation of the backward classes, submitted its report on 31 July 2023, and the report has not been published or implemented. It is examinable as a pending policy question, and describing it as implemented policy is an error of the same class as describing an observation as a holding. The third question is beneficiary and identification data inside the lists, where the Davinder Singh majority’s insistence on empirical material for sub-classification makes the quality of a state’s data, rather than its conclusion, the examinable point.

The commissions double as the field’s evidence machinery, and using them as sources is the mark of a prepared answer. The standing commissions’ annual reports, laid before Parliament, contain the administration’s own account of how safeguards operate, and the Article 340 commissions generate the field’s foundational datasets, Mandal’s indicator-based identification in its time and the Rohini Commission’s examination of how the backward class quota’s benefits distribute among the communities on the central list in ours. That second exercise illustrates the data problem in its sharpest form: an inquiry into distribution must work from service and admission records rather than from a population count, because the count does not exist, so its conclusions, when published, will be conclusions about the distribution of benefits among listed communities and not about the communities’ shares of population. An answer that understands what kind of evidence a commission can produce, administrative records for distribution, field inquiry for identification, census material for population, and that asks no source for a kind it cannot supply, handles every data question in the field without asserting a single contested figure.

Why is the caste data question argued rather than answered in the hall?

Because no authoritative caste count has been published since the census of 1931, and the estimate every answer relies on, the Mandal Commission’s fifty-two per cent, is a construction from that census rather than a count. The examinable material is the absence itself: which institution counts, on what method, and what follows for identification when the count is contested.

Writing the answer: separating the three questions

The worked method has four moves, and it applies to any demand in this field. Take the recurring demand type on sub-classification, which asks in various wordings whether dividing the lists is constitutionally sound and socially wise. The first move is to split the demand. The constitutional question is whether Articles 341 and 342, which commit the lists to the President and Parliament, leave room for a state to grade within a list; the judicial question is what has been held about that room; the empirical question is whether the distribution of benefits inside the lists is in fact concentrated, and on what evidence. The second move is to answer the judicial question with holdings only: E. V. Chinnaiah held sub-classification impermissible, Davinder Singh overruled it six to one and permitted sub-classification on empirical material, and four judges added observations on a creamy layer that are not law. The third move is to state the strongest form of each side on the constitutional question: that grading within a list is the state’s instrument for reaching the most disadvantaged inside a group the Constitution treats as one class, against the position in the dissent that the lists are indivisible constitutional units and that internal grading invites political fragmentation of a protective category. The fourth move is to conclude on the tractable question, which is the institutional one: after 2024 the permissibility question is settled and the live question is evidentiary, whether a state’s material can bear the classification. The value disagreement, whether internal grading strengthens or weakens the solidarity the lists protect, is named as a value disagreement and left standing. An answer built this way takes a position, and the position is on ground the examiner can mark.

The same decomposition disciplines the ceiling question, where the constitutional layer is the text of Articles 15 and 16 (which fix no number), the judicial layer runs Balaji to Indra Sawhney to Janhit Abhiyan (a rule of prudence, restated as a governing rule with an extraordinary exception, then held not to be an inflexible basic structure feature), and the empirical layer is the distribution argument about who benefits when the ceiling is raised or held. It disciplines the commissions question, where the constitutional layer is Articles 338, 338A and 338B, the judicial layer is thin, and the real question is institutional: recommendatory powers, reporting to the President, and the distance between inquiry and enforcement. And it disciplines the atrocities question, where candidates most often collapse the layers, offering enforcement failure as if it answered the constitutional question about Section 18A, or the constitutional validity of Section 18A as if it answered the enforcement question. The layers are the marks.

The method transfers across demand types, and a second worked instance fixes it. Take the institutional demand on the commissions, which asks in various wordings whether the constitutional commissions protect the groups they exist for. Split it. The constitutional layer is the text: inquiry powers of a civil court, reports to the President, laying before Parliament with an action-taken memorandum, and no power of direction. The judicial layer is thin and should be declared thin: the commissions’ design has not been substantially reconstructed by litigation, and their one great judicial moment is generative, Indra Sawhney’s direction producing the statutory backward classes commission of 1993. The empirical layer is the reporting record, which the candidate states by attribution: governments’ responses to commission findings are recorded in the memoranda laid with the reports, and the distance between recommendation and action is documented there rather than in commentary. The conclusion then takes itself on the tractable layer, which is design: a body that can investigate anything and direct nothing protects by publicity and by the parliamentary moment its report creates, and whether that is protection enough is the value question, named and left. The candidate who writes that answer has not recited three commissions’ functions, which is the summary the demand was designed to refuse.

Two further habits are worth stating because they are cheap and consistently rewarded. Answer in the vocabulary of the provisions: an answer that says Article 15(5) rather than “the amendment about private colleges” signals command of the spine, and the spine is short enough that there is no excuse for paraphrase. And close every answer on the question the evidence settles. Where the demand is distributive, whether reservation should expand, the evidence cannot settle it, and the strong close converts it: the tractable questions are whether the instrument is constitutionally available, whether its conditions are met on the material, and which institution decides. That conversion, done openly, is analysis rather than evasion, because it tells the examiner exactly which part of the demand is being answered and why the remainder is a matter of democratic choice.

The stance the method requires also has a content rule: a side must be given in the voice of a holder, never as an anonymous tendency. In this field the holders are identifiable and few. The proposition that reservation secures representation of groups excluded from state power, rather than relief of individual poverty, is the reasoning of the Indra Sawhney majority and the analytical frame of Marc Galanter’s study of compensatory discrimination, and it should be cited to them. The proposition that the lists are indivisible and that internal grading dissolves their constitutional protection is Bela Trivedi J.’s dissenting reasoning of 2024, and the proposition that economic criteria cannot define a reservation category without destroying its rationale is the reasoning of the Janhit Abhiyan dissent, each available to be stated at full strength and answered. The empirical claims have holders of the same kind: the ministry’s annual reports for the enforcement record, the crime records bureau for disposal patterns, the commissions for identification and distribution. An answer that writes “critics argue” or “supporters point out” has declined the attribution the examiner is scanning for, and in a field where every serious position has a named institutional author, the decline is visible.

Examples obey the same filing discipline as propositions. A general claim in this field should be grounded in one named instance whose status the candidate knows: a state caste survey for the data question, a state’s sub-classification legislation for the post-2024 process, a district’s exclusive special court for the enforcement machinery, a single commission’s report for the institutional design. The instance does the work that an invented statistic attempts and fails to do, because it is checkable in principle and because it shows the examiner the general proposition operating somewhere real. The rule for choosing instances is the rule for choosing citations: prefer the instance whose facts are settled and whose status is current in the candidate’s own file, and drop any instance whose details have gone uncertain, since an answer needs one example and is damaged by a wrong one out of proportion to the credit a right one earns.

The same discipline governs the two ends of the answer. The opening should begin from the provision or the tension the demand names, since the field’s questions are generated by texts and an opening that quotes the operative clause, the ceiling’s absence from the text of Article 16, the exclusion written into Article 15(6), the locked procedure of Article 341, establishes command before any argument begins. An opening that begins with a general statement about caste in Indian society spends the first impression on Paper One material. The close should do the opposite of the opening’s work: where the opening narrowed to a text, the close widens to the institution, stating which body holds the question after the answer ends, Parliament for the lists and the ceiling’s statutory applications, the Court for permissibility, the commissions for identification and review, the district administration for enforcement. That closing move converts a summary into a judgment about where the field is decided, and it is the sentence examiners remember.

For the preliminary examination the same anchor stock serves a different retrieval. Prelims questions in this field test the spine at the level of the single proposition: which article carries the promotion power, which amendment made a commission constitutional, which statute carries a given year and number, whether a statement about a holding’s status is correct. The mains method of separating layers converts directly, because the prelims distractor is almost always a layer error, a provision paired with the wrong clause, a holding stated as an observation, a statutory body described as constitutional. A candidate whose anchor file records status as part of every entry is protected against the distractor by construction, which is a further reason to build the file in the form the next section prescribes.

The method must also survive the hall’s arithmetic. Mains answers in the general studies papers are written to word limits of the order of one hundred and fifty words for the shorter questions and two hundred and fifty for the longer, against the clock, and a structure that cannot compress will be abandoned under pressure for the summary the candidate swore to avoid. The three-layer separation is a compression technology before it is an analytical one. A two hundred and fifty word answer on the scaffold spends its first sentences on the provision and the holding, which are short because they are propositions; its middle on the two strongest opposing cases, which are short because each is one holder’s argument stated once; its empirical sentence on an attributed dispute, which is short because attribution replaces exposition; and its close on the tractable question, which is one sentence because the separation has already done the work. Candidates who run out of structure at the word limit have usually spent the middle of the answer narrating history the demand did not ask for, and the scaffold’s rows exist to make that error visible in practice before it is made in the hall.

The field’s open edges, stated with their statuses

A study article earns its keep at the edges, where the field is unfinished, because the paper prefers live questions and the live questions are exactly where status errors breed. Four edges should be held with their statuses stated.

Sub-classification has moved from doctrine to administration. After Davinder Singh, the operative questions are the ones the majority’s conditions generate: what material a state assembles, how it grades, and whether its grading survives the challenge that it has altered the list in substance. Telangana’s sub-classification legislation of 2025, enacted after a dedicated one-person commission examined the distribution of benefits inside its Scheduled Caste list, is the working model of the post-judgment process, commission first, legislation second, litigation expected, and the examination tests the process and its conditions rather than any state’s outcome, which remains sub judice in the nature of the subject. The candidate should therefore learn the conditions as the content and treat each state’s exercise as an instance to be filed under them.

The caste count has changed status. The Union government’s decision in 2025 that the forthcoming census will enumerate caste converts the oldest data question in the field from an argument about whether the state should count into an argument about what a count will do when it arrives: whether identification by lists and commissions should yield to enumeration, how a count interacts with the ceiling arithmetic that Mandal’s estimate made possible, and what happens to categories constructed for a data-poor era when data arrives. No count has been published, and the enumeration itself lies in the future, so the examinable content remains the institutional analysis, with the decision recorded as a decision. An answer that writes as though a caste count exists, or as though the demand for one were still the live form of the question, is behind the record in opposite directions.

The Sixth Schedule’s extension is the third edge. Demands to bring further tribal areas within the Schedule, most prominently Ladakh, where the National Commission for Scheduled Tribes recommended inclusion in 2019, test the distinction this article drew between the two territorial models: the demand is precisely for the self-governing model in place of the protective one, and it is made on the ground that land and demographic protection in a sensitive border region require legislative institutions of the community itself rather than gubernatorial safeguards. The question is pending as a policy question, no amendment having been enacted, and it should be answered as a design question, which model fits which failure, rather than as a prediction.

The census decision also reopens the legislative reservation timetable, and the connection should be filed with it. The women’s reservation amendment of 2023 operates only after the delimitation that follows the first census taken after its commencement, and the Article 334 reservation for the Scheduled Castes and the Scheduled Tribes runs to eighty years from the Constitution’s commencement on the strength of the One Hundred and Fourth Amendment, a period the forthcoming census and the delimitation after it will frame politically as well as administratively. A census that enumerates caste therefore arrives in a field where three separate mechanisms, identification, legislative reservation and women’s reservation, are all indexed to census events or census-derived numbers, and the candidate who sees the indexing can write the institutional answer the data question is really asking for: not what the count will show, which nobody knows, but which decisions the count will unlock, which the constitutional text already specifies.

The fourth edge is the backward classes’ internal distribution, which the Rohini report addresses and which publication alone can move. Until the report is public, the examinable content is the question’s structure: the central list’s communities are unequally placed, the benefits of the twenty-seven per cent are distributed by competition inside the list, and the instruments available, sub-categorisation by the Union on the commission’s framework, are policy instruments awaiting a political decision. Each of the four edges, held this way, is an answer waiting for its demand; held loosely, as headlines, each is a source of the status errors the paper is designed to catch.

The scaffold, and the anchor file that maintains it

The findable artifact for this article is the social justice answer scaffold: six question types, each with the provisions to cite, the statutes to cite, the judgments to cite, the empirical dispute to acknowledge without asserting figures, and the closing position the structure supports. A candidate who has internalised the six rows can construct a defensible answer to any demand in this field under time pressure, because the rows are the field’s recurring shapes rather than specimen questions.

Question type	Constitutional provisions to cite	Statutes to cite	Judgments to cite	Empirical dispute to acknowledge	Closing position the structure supports
Reservation and equality	Articles 14, 15(4), 15(5), 15(6), 16(4), 16(6)	Central Educational Institutions Act, Act 5 of 2007	Indra Sawhney; Ashoka Kumar Thakur; Janhit Abhiyan	Who benefits inside each category, and whether benefits concentrate	The permission is settled; the live questions are design and distribution
Identification and the caste data question	Articles 340, 341, 342, 342A, 366	State list legislation under the restored state power	Indra Sawhney on caste as a starting point; Jaishri Laxmanrao Patil; Davinder Singh	No published caste count since 1931; the Mandal estimate is a construction	Identification is committed to lists and commissions because enumeration is contested
Atrocities and enforcement	Articles 17, 21, 46	Protection of Civil Rights Act, Act 22 of 1955; Atrocities Act, Act 33 of 1989, amended 2016 and 2018	Mahajan; the 2019 review; Prathvi Raj Chauhan	Conviction and pendency patterns in official reporting, varying by state and year	Validity and enforcement are separate questions; the statute answers one, administration the other
Tribal self-governance and land	Articles 244, Fifth and Sixth Schedules, 46	Panchayats Extension Act, Act 40 of 1996; Forest Rights Act, Act 2 of 2007	Samatha; the Niyamgiri judgment	Recognition and rejection patterns under the forest rights statute, varying by state	The Gram Sabha is the institutional centre of both statutes; outcomes turn on its working
Sanitation and dignity	Articles 17, 21, 46	Manual Scavengers Act, Act 25 of 2013, and the Act 46 of 1993 it followed	Safai Karamchari Andolan; the 2023 compensation judgment	Counts of workers and of hazardous cleaning deaths rest on surveys and case material of differing scope	The field’s question is rehabilitation and mechanisation, framed as dignity rather than sanitation
Commissions and institutional design	Articles 338, 338A, 338B, 339, 340	The 1993 backward classes commission Act in its historical place	Indra Sawhney for the direction that produced the statutory commission	Whether recommendatory bodies alter administrative behaviour is contested on the reporting record	The design question, powers and reporting lines, is the tractable question; effectiveness follows design

One habit protects the candidate against the field’s continuing movement, because this cluster of law does not stand still between the notification of an examination and its writing. Every new development in the field, a judgment, a commission’s report, a state sub-classification law, a survey’s publication, should be processed through four questions before it is admitted to the anchor file. Which layer does it belong to: does it change what the Constitution permits, what has been held, or what the evidence shows? Which institution produced it, and is that institution competent to settle the point it addresses, as a state survey is competent evidence about its state and incompetent as a national count? What is its status: decided, observed, recommended, published, implemented, or merely announced? And which existing entry does it displace, since a development admitted without displacing the superseded entry leaves the file, and then the answer, asserting both. The candidate who processes the 2024 judgment through those questions files it as a judicial-layer holding on sub-classification with observations attached, displacing Chinnaiah; the candidate who files it as the creamy layer’s extension to the Scheduled Castes has failed the third question, and the failure will appear in the answer. The method is slower than collecting headlines and faster than relearning the field, which is the exchange revision time should make.

The scaffold is only usable if the material behind it is held in a form the candidate can revise from, and that is a filing problem before it is a memory problem. The working method is a personal anchor file: one entry per provision with its one-sentence function, one entry per judgment with its holding in one sentence and its observations in a separate sentence, one entry per statute with its number, year and examination function, and a page for each pending item with its status and the date of that status. Keeping holdings and observations in separate sentences inside the file is what prevents the commonest error in this field from reaching the answer sheet. The companion routed to this article is VaultBook, whose encrypted single-file workspace, built for keeping source material private, suits an anchor file that a candidate maintains across a long preparation, and the discipline the file enforces matters more than the tool: an entry that cannot state its own status, decided, observed, recommended or pending, is not ready to be cited.

The scaffold converts into a revision cycle without further apparatus. On a first pass the candidate builds the anchor file from the spine and the judgment set, one entry per provision, case and statute, each entry carrying its status. On a second pass the candidate writes one answer per scaffold row under time, using only the file, and marks the places where the file failed to supply a needed proposition; those failures, and not a fresh reading of the subject, define the third pass. On the final pass the candidate rehearses the error list, the statuses of the ceiling, of the creamy layer across its three contexts, of the Rohini report, of the Sixth Schedule’s reach, of the women’s reservation amendment’s commencement, until stating each status is reflexive. For the researcher and the journalist the same material serves as a map of ownership: every live question in this field belongs to a named institution at a named stage, the lists to Parliament and the President, permissibility to the Court, identification to the commissions, enforcement to the district, and the map’s value outside the examination hall is that it prevents the commonest professional error in the field, asking a body to answer for a stage it does not own.

The verdict on this field, for the candidate deciding how much of a preparation season it deserves, is that it is the highest-yield doctrinal area in Paper Two relative to its size. The spine is a page of provisions. The judgment set is nine decisions with one-sentence propositions. The statute set is six instruments with one function each. What separates the strong answer from the opinion is not more material but the separation of the three questions every demand contains, and that is a skill, which means it can be practised on the scaffold until it is automatic. Revise the spine until it is recitable, revise the holdings until their status is part of the sentence, and let the values in this contested field be argued by the commissions, the courts and Parliament, whose arguments, stated at full strength and attributed, are the answer. That discipline is the examination form of this series’ standing thesis. The constitutional text is the text, short enough to be learned entire; the amendments, the judgments and the implementation record are the record, too large to be learned and therefore to be organised; and the candidate’s craft, like the citizen’s understanding, consists in never mistaking one for the other, never reading the record into the text or the text as a substitute for the record. Social justice is the field where that mistake is most tempting, because the text is noble and the record is contested, and it is therefore the field where the series’ method, and the examination’s marking, reward the separation most.

Frequently Asked Questions

Q: How is social justice legislation tested in the civil services examination?

Social justice legislation is tested mainly in General Studies Paper Two, under the Constitution head for significant provisions and amendments and under the welfare head, whose syllabus language covers the mechanisms, laws, institutions and bodies constituted for vulnerable sections. The same material appears in Paper One through caste in Indian society and in the preliminary examination as polity fact. The demands are analytical rather than factual in the mains: candidates are asked to evaluate ceilings, identification, institutional design and enforcement, which is why a provision recalled without its judicial position and its empirical dispute earns only part of the available credit on any of these surfaces.

Q: Which constitutional provisions on social justice must an aspirant know?

An aspirant must know Article 14 as the background norm; Article 15 with clauses (3), (4), (5) and (6) covering women and children, backward classes, private educational institutions and economically weaker sections; Article 16 with clauses (4), (4A), (4B) and (6) covering reservation, promotion and carried-forward vacancies; Article 17 abolishing untouchability; Article 46 directing protection of the weaker sections; and Article 335 on efficiency. The list provisions, Articles 341, 342 and 342A, the commission provisions, Articles 338, 338A and 338B, the inquiry power in Article 340, and the Fifth and Sixth Schedules complete the set from which nearly every question in the field is generated.

Q: Which reservation judgments matter most for the examination?

Indra Sawhney matters most, as the source of the creamy layer rule, the fifty per cent ceiling in its operative form and the bar on promotion reservation. Balaji supplies the ceiling’s origin and the rule that caste cannot be the sole criterion. Nagaraj and Jarnail Singh govern promotion, the first imposing conditions and the second withdrawing the backwardness data requirement while keeping the creamy layer. Ashoka Kumar Thakur covers admissions in central institutions. Janhit Abhiyan governs the economically weaker sections quota and the ceiling’s flexibility, and Davinder Singh governs sub-classification within the Scheduled Caste and Scheduled Tribe lists, with its creamy layer observations kept distinct from its holding.

Q: How do you write about reservation without taking a side?

By separating the three questions a reservation demand contains and answering each on its own authorities. State the constitutional position from the text, the judicial position from holdings stated with their status, and the empirical dispute as a dispute, attributed to the reporting and the commissions rather than asserted. Give each side of the value disagreement in its strongest form and attribute it. Then conclude on the tractable question, usually the constitutional or institutional one, and name the remaining disagreement as a disagreement about values. An answer built this way is not neutral filler; it takes every position the evidence supports and declines only the position no evidence can settle.

Q: Which social justice commissions are most often confused?

The three national commissions are most often confused: for the Scheduled Castes under Article 338, for the Scheduled Tribes under Article 338A, and for the Backward Classes under Article 338B. The confusion has specific points. The backward classes commission was statutory from 1993 and became constitutional only in 2018, long after the other two. Candidates also confuse Article 340, the power to appoint a commission of inquiry into backward class conditions under which the Mandal Commission sat, with Article 339, which concerns the administration of the Scheduled Areas and carries the ten-year language. All three standing commissions are recommendatory bodies reporting to the President, with civil court powers for inquiry.

Q: What is the commonest error in social justice answers?

The commonest error is misstating the status of a proposition: describing an observation in a judgment as its holding, a commission’s recommendation as law, a pending report as implemented policy, or the fifty per cent ceiling as absolute. The sharpest instance is the claim that the 2024 sub-classification judgment applied the creamy layer to the Scheduled Castes and the Scheduled Tribes; four judges expressed that view as an observation for the future, and the holding permits sub-classification on empirical material. Examiners probe for status errors because they distinguish a candidate who has read the record from one who has read summaries of it. Every proposition in an answer should carry its status inside the sentence.

Q: How much data should a social justice answer use?

Very little as asserted figures, and a great deal as attributed evidence. The figures in this field, representation shares, conviction rates, beneficiary counts, survey totals, are compiled on differing definitions for differing periods by differing authorities, and an unattributed number is an error waiting to be marked. The strong practice is to reference the category and the source: to note that outcomes under the atrocities statute form a small fraction of disposals in the crime records bureau’s reporting, or that no caste count has been published since 1931, rather than to wager on a percentage. Where a figure is genuinely settled, such as a statutory share or a compensation amount fixed by a judgment, state it with its source and period.

Q: Which social justice topics overlap with federalism?

Identification is the largest overlap: the One Hundred and Second Amendment centralised the specification of backward class lists in Article 342A, the Maratha judgment read it as withdrawing state power, and the One Hundred and Fifth Amendment restored the states’ lists, so any question on identification is also a question on the distribution of power. Tribal governance overlaps through the Fifth and Sixth Schedules and the panchayats extension statute, which condition state legislation in Scheduled Areas. Enforcement overlaps because the protective statutes are central legislation administered by state police and state courts. A candidate should expect the federalism angle inside social justice questions and prepare the list and Schedule material for both papers.

Q: Which constitutional amendments matter most in social justice answers?

Eight amendments carry most of the field. The First inserted Article 15(4). The Seventy-seventh enabled reservation in promotion, the Eighty-first took carried-forward vacancies outside the annual ceiling, the Eighty-second permitted relaxation of qualifying standards under Article 335, and the Eighty-fifth added consequential seniority. The Ninety-third inserted Article 15(5), taking reservation into private educational institutions. The One Hundred and Second made the backward classes commission constitutional and centralised its list in Article 342A, the One Hundred and Third created the economically weaker sections quota, and the One Hundred and Fifth restored the states’ power over backward class lists. A candidate should attach each amendment to the judgment or problem it answered.

Q: What is the difference between Articles 15(4) and 15(5)?

Article 15(4) is the original special-provision clause, inserted by the First Amendment, permitting provision for the advancement of socially and educationally backward classes, the Scheduled Castes and the Scheduled Tribes across the state’s field of action. Article 15(5), inserted by the Ninety-third Amendment, is narrower in subject and wider in reach: it concerns admission to educational institutions and extends the power to private institutions, aided or unaided, while excluding minority institutions under Article 30(1). The distinction matters because clause (5) is the textual basis on which reservation entered the unaided private college, and because the central educational institutions statute of 2006 was enacted in reliance on it and upheld in Ashoka Kumar Thakur.

Q: How should a candidate state the fifty per cent ceiling in an answer?

As a rule with a history and an exception, not as an absolute. Balaji stated that reservation should ordinarily remain below fifty per cent. Indra Sawhney restated fifty per cent as the governing rule, breachable only in extraordinary circumstances, and the Maratha judgment refused to find such circumstances in a large and politically organised community’s claim. Janhit Abhiyan then held that the ceiling is not an inflexible feature of the basic structure, since the economically weaker sections quota operates above it. The accurate sentence is that the ceiling governs caste-based reservation as a rule of constitutional prudence with a narrow exception, and that its status differs by the category to which it is applied.

Q: What does the creamy layer mean, and to which categories does it apply?

The creamy layer is the exclusion of the socially advanced members of a reserved category from the benefit of reservation, introduced for the backward classes by Indra Sawhney as a condition of the quota’s validity and applied to admissions in central institutions in Ashoka Kumar Thakur. For the backward classes it is law, administered through income and status criteria revised by executive order. For the Scheduled Castes and the Scheduled Tribes it applies in promotion, where Jarnail Singh retained it as a condition, but it is not a general condition of their reservation: the suggestion by four judges in Davinder Singh that it should extend there was an observation, not a holding, and answers must keep that distinction.

Q: Which statutes should a candidate cite in an answer on atrocities against the Scheduled Castes?

Two statutes, kept distinct. The Protection of Civil Rights Act, Act 22 of 1955, is the penal statute under Article 17, punishing the practice of untouchability; renamed and given its operative penalty scale by the amendment of 1976, it is the older and more general instrument. The Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, Act 33 of 1989, is the protective statute addressed to offences by persons outside the two groups against members of them, with special courts and, after the amendment Act 1 of 2016, an expanded offence list and rights for victims and witnesses. Answers lose credit when the two are merged into a single unnamed atrocities law.

Q: What should a candidate know about the Forest Rights Act for the examination?

Three things: its identity, its machinery and its companion case. The Act is Act 2 of 2007 under a 2006 short title, in force from 31 December 2007, and it recognises individual rights to forest land under occupation and community rights, including over minor forest produce, for forest dwelling Scheduled Tribes and other traditional forest dwellers. The Gram Sabha initiates the determination of rights, which makes the statute an institutional question as much as a rights question. The companion authority is the Niyamgiri judgment of 2013, in which the Gram Sabhas’ decisions on religious and cultural rights were treated as determinative, and the earlier Samatha judgment on alienation of tribal land supplies the historical line.

Q: Why does the manual scavenging statute belong in answers on dignity rather than sanitation?

Because the statute and the courts frame the subject as a violation of dignity and equality, not as a defect of municipal engineering. The Prohibition of Employment as Manual Scavengers and their Rehabilitation Act, Act 25 of 2013, prohibits manual scavenging and the hazardous cleaning of sewers and septic tanks, and pairs prohibition with survey, rehabilitation and alternative employment, which its 1993 predecessor lacked. The Supreme Court’s compensation jurisprudence carries the framing: ten lakh rupees for sewer deaths since 1993 in Safai Karamchari Andolan in 2014, raised to thirty lakh rupees in the judgment of 2023. An answer that treats the subject as sanitation management misses the Articles 17 and 21 ground on which the law stands.

Q: How should a candidate use the Mandal Commission in an answer?

As the identification authority, with its method stated rather than merely its recommendation. The Commission, appointed under Article 340 and reporting in 1980, identified backward classes through eleven indicators grouped under social, educational and economic heads and weighted three, two and one, and estimated the backward classes at fifty-two per cent of the population by construction from the census of 1931. Its recommendation of twenty-seven per cent reservation in central services was implemented by the office memorandum of 13 August 1990 and is the figure Indra Sawhney upheld. The examinable precisions are the estimate’s status as a construction, the intersection with state lists in the first phase of implementation, and the 1991 memorandum’s economic preference, whose fate the judgment divided.

Q: What is the Rohini Commission, and how should its status be stated?

The Rohini Commission is the commission chaired by Justice G. Rohini, appointed under Article 340 in 2017, to examine the sub-categorisation of the backward classes, meaning the distribution of the twenty-seven per cent backward class quota among the communities on the central list. It submitted its report on 31 July 2023. The report has not been published, and no sub-categorisation of the central backward class list has been implemented on its basis. The correct examination handling is to present sub-categorisation of the backward classes as a pending policy question with a completed but undisclosed report behind it, distinct from sub-classification within the Scheduled Castes and the Scheduled Tribes, which Davinder Singh decided as a judicial question.

Q: How did reservation in promotion change between Indra Sawhney and Jarnail Singh?

Indra Sawhney held that reservation under Article 16(4) operates at initial appointment and not in promotion. Parliament answered with the Seventy-seventh Amendment inserting Article 16(4A) for the Scheduled Castes and the Scheduled Tribes, followed by the amendments on carried-forward vacancies, qualifying standards and consequential seniority. Nagaraj in 2006 upheld that architecture on conditions: quantifiable demonstration of backwardness and of inadequacy of representation, consistency with efficiency under Article 335, and exclusion of the creamy layer. Jarnail Singh in 2018 withdrew the requirement of quantifiable data on backwardness for the two groups, holding it contrary to Indra Sawhney, while retaining inadequacy of representation, efficiency and the creamy layer as operative conditions.

Q: How does the EWS judgment change the way reservation should be discussed in answers?

Janhit Abhiyan, which upheld the One Hundred and Third Amendment by three judges to two in 2022, changes the discussion in three ways. Reservation can no longer be described as exclusively caste-based relief, since an economic criterion was upheld as a permissible basis. The fifty per cent ceiling can no longer be described as an absolute limit, since the majority held it is not an inflexible part of the basic structure and the new quota operates in addition to existing reservation. And the sharpest examinable point is the ground of division: the dissent objected to the exclusion of the Scheduled Castes, the Scheduled Tribes and the backward classes from the new category, so answers should present the case as a dispute about exclusion, not about poverty.

Q: Which facts in this field are worth memorising, and which are not?

Worth memorising are the items from which questions are generated: the clause numbers of Articles 15 and 16 and what each does; the numbers of the eight amendments that structure the field; the Act numbers and years of the six statutes; the composition and dates of the leading benches where they are settled, such as nine judges in Indra Sawhney and seven in Davinder Singh; and the settled figures fixed by text or judgment, including the admission shares in the central institutions statute and the compensation amounts in the sewer death cases. Not worth memorising are moving administrative statistics, representation percentages and disposal rates, which should be handled by attribution to their reporting source and period instead.