To CAP or not to CAP: The Bombay High Court on Equality and Access to Education

In an interesting judgment delivered yesterday (Yash Pramesh Rana vs State of Maharashtra), a Full Bench of the Bombay High Court struck down Government Resolution [“GR”] dated 27.2.2013. This Government Resolution had restricted the application of a fee-reimbursement scheme only to those SC/ST/OBC students who had taken college admission through the government-run Common Admissions Procedure [“CAP”].

The facts were straightforward. To enter an engineering college in the state of Maharashtra, a student had to undertake the Common Entrance Test [“CET”]. On the basis of ranks obtained in the CET, students could then participate in the CAP, and gain admission into any of the colleges that were part of the CAP. However, not all colleges – including some minority colleges (the case itself concerned a Gujarati-language linguistic minority college) – were part of the CAP. Certain colleges had their own admissions process, that was approved by the Pravesh Niyantran Committee. The impugned GR – as indicated above – provided for a fee-reimbursement scheme to SC/ST/OBC students, but limited it only to the former category (i.e., those who took part in the CAP).

In a judgment authored by Dama Seshadari Naidu J., the Bombay High Court found that the impugned G.R. was entirely arbitrary, and violated Article 14 of the Constitution. The judgment is noteworthy, because it was decided almost entirely on the basis of a textbook application of burdens and evidentiary standards under Article 14. The Court observed that as the impugned G.R. created a classification, and disadvantaged one set of people (the category of students that was not granted fee reimbursement), a prima facie case of discrimination was made out. This, then, shifted the burden of justification onto the State. The State essentially produced two arguments: first, that extending the free-reimbursement scheme to all SC/ST/OBC students would be financially prohibitive, and secondly, that students who had gone through the CAP and those who had not constituted two separate “classes”, as the CAP was a transparent, well-documented, well-regulated, and non-discriminatory process of allocation.

On the first count, the Court held that mere financial difficulties, without something more, could not be a ground for discriminatory treatment. In other words, in a class of similarly situated people, the State could not refuse to one set of people a benefit that it was granting to another, on the basis that it did not have the financial capacity. This is self-evidently correct and logical. On the second count, the Court held that the State had failed to bring any evidence on record to show that the non-CAP process was any less rigorous and transparent than the CAP process, in any sense that justified withholding of identical benefits. Indeed, the Pravesh Niyantran Committee was also run by the government. Consequently, as the State had produced no evidence to justify its claim, the impugned G.R. was arbitrary and unconstitutional. As the Court correctly noted, following the US Supreme Court, the presumption of constitutionality would not stretch so far as to imagine the existence of an “undisclosed and unknown reason for subjecting certain individuals or corporations to hostile and discriminatory legislation.”

It is also interesting to note that Naidu J. framed the dispute within the backdrop of historical inequalities concerning access to education in India (going back to the 1850s), and the use of affirmative action as tool of corrective justice. This was relevant to the case, as one of the arguments raised by the State was that fee-reimbursement was simply a benefit it was conferring upon certain students; as there was no antecedent right to claim fee-reimbursement, a person who had been deprived of it had no locus to move the Court. Now at one level, of course, the Court correctly answered this by stating that any State action – including “largesse” – had to conform to constitutional principles. However, the Court also noted that – within the backdrop of structural inequality in India – fee-reimbursement for SC/ST/OBC students had to be understood as “a facet of affirmative action.” This immediately took it from the domain of largesse/benefits and into the domain of constitutional obligation, thus making it even more incumbent upon the State to frame a non-discriminatory policy of access.

Now, an interesting corollary of the Court’s observation is that if indeed fee-reimbursement is a form of affirmative action, then – as a non-reservation based form of affirmative action – it falls within Article 16(1) of the Constitution (guarantee of equality of opportunity). This raises a host of fascinating questions for the future, including whether specific claims of fee-reimbursement can be made against the State by socially disadvantaged communities (as 16(1) is framed as a right), the fact that such schemes can go beyond SC/ST/OBC communities (as 16(1) affirmative action measures are not limited to 16(4) beneficiaries), and so on. Of course, none of these questions were before the Court; however, it will be interesting to see whether future judgments will carry forward the logic of fee-reimbursement being a form of affirmative action, and what that might mean in practical terms.

Guest Post – Article 16 and the Concept of a “Power plus Duty” – II

(In Part II of this series, Karan Lahiri continues his argument about the interpretation of Article 16 of the Constitution)

In the last post, I introduced the concept of “power coupled with duty”. In this part, I talk about a batch of writ petitions pending before the Supreme Court, emerging from the State of U.P. (the lead petition being W.P. © 273/2015 – Sewa Lal & Ors. v State of UP), where the issue of “power coupled with duty” is likely to be looked at once again. This second blog post gives some background on how these petitions came to be filed, because the factual context actually demonstrates why reading Article 16 as a hybrid provision, containing both a power and a duty, is the most just and equitable solution. At the end of this part, I look at how the current reading of Article 16 creates an asymmetry problem, and I then provide a brief breakdown of what I believe to be the correct reading of Article 16 as a hybrid power-plus-duty provision.

To place this in context, it may be remembered that in Indra Sawhney, it was held that Article 16(4), as it then stood, was not wide enough to bring within its fold reservation in matters of promotion. However, the Court declared that this would not affect promotions that had already been made and, in fact, granted the additional protection that where reservations have already been provided for in Central or State Services, the same position may continue for a further period of five years. Thereafter, Article 16(4A) was inserted vide the Constitution (Seventy-seventh) Amendment Act, 1995, which did provide for reservations in promotions. To recapitulate, this provision presently reads:

(4A) Nothing in this article shall prevent the State from making any provision for reservation in matters of promotion, with consequential seniority, to any class or classes of posts in the services under the State in favour of the Scheduled Castes and the Scheduled Tribes which, in the opinion of the State, are not adequately represented in the services under the State.”

It was after this that the Supreme Court’s judgment in M. Nagaraj was rendered, in which the Court, while upholding the constitutional validity of the amendment, also set out certain conditions precedent for the State to implement ameliorative measures under Articles 16(4A), as also 16(4B) (which deals with carrying forward of vacancies reserved for backward classes/SCs/STs). It was observed –

“…There are numerous petitions pending in this Court in which reservations made under State enactments have been challenged as excessive. The extent of reservation has to be decided on the facts of each case… Therefore, in each case the Court has got to be satisfied that the State has exercised its opinion in making reservations in promotions for SCs and STs and for which the State concerned will have to place before the Court the requisite quantifiable data in each case and satisfy the Court that such reservations became necessary on account of inadequacy of representation of SCs/STs in a particular class or classes of posts without affecting general efficiency of service as mandated under Article 335 of the Constitution.

However, in this case, as stated above, the main issue concerns the “extent of reservation”. In this regard the State concerned will have to show in each case the existence of the compelling reasons, namely, backwardness, inadequacy of representation and overall administrative efficiency before making provision for reservation. As stated above, the impugned provision is an enabling provision. The State is not bound to make reservation for SCs/STs in matters of promotions. However, if they wish to exercise their discretion and make such provision, the State has to collect quantifiable data showing backwardness of the class and inadequacy of representation of that class in public employment in addition to compliance with Article 335…” [Emphasis supplied]


Therefore, M. Nagaraj laid down three requirements: –

  1. General efficiency of the service must not be compromised. This is traceable to Article 335 of the Constitution.
  2. The State must come to the conclusion that the groups in question are not adequately represented in the service, based on quantifiable data. This flows from Article 16(4A) itself.
  3. The State, based on quantifiable data, must come to the conclusion that the class benefitting from the ameliorative measure is “backward”.

While this is a topic to which an entirely separate blog post can be devoted, I wish to point out at this juncture that this third requirement does not flow from the text of Article 16(4A), and this constitutes a major error in M. Nagaraj. We need to remember that Nagaraj was dealing with Article 16(4A), which speaks of “Scheduled Castes and Scheduled Tribes”, and not “backward classes”. Unlike reservations under Article 16(4), which categorically mentions ‘backward class’ (the relevant part of which reads “Nothing in this article shall prevent the State from making any provision for the reservation of appointments or posts in favour of any backward class of citizens), the State did not have to undertake a separate exercise to determine “backwardness” for the purposes of Article 16(4A), because Article 16(4A) allowed for reservation in promotions only for SCs/STs (Nothing in this article shall prevent the State from making any provision… in favour of the Scheduled Castes and the Scheduled Tribes). As per both N.M. Thomas and Indra Sawhney, SCs/STs are presumed to be backward classes and, in E.V. Chinnaiah v. State of A.P., (2005) 1 SCC 394, they have correctly been identified as “not merely backward but the backwardmost.”

 This error in Nagaraj was compounded in U.P. Power Corporation Ltd. v. Rajesh Kumar & Ors., (2012) 7 SCC 1. In this case, the Court was examining the constitutionality of Section 3(7) of the U.P. Public Services (Reservation for Scheduled Castes, Scheduled Tribes and Other Backward Classes) Act, 1994, and Rule 8-A of the U.P. Government Servants Seniority (Third Amendment) Rules, 2007. Section 3(7) preserved pre-existing Government Orders in U.P. that provided for reservation in promotion. Rule 8-A provided for consequential seniority.

As it was found that no quantifiable data had been collected by the State of U.P. for effectuating reservation in promotion and making the rule of consequential seniority applicable, the Court held: –

“86. We are of the firm view that a fresh exercise in the light of the judgment of the Constitution Bench in M. Nagaraj is a categorical imperative. The stand that the constitutional amendments have facilitated the reservation in promotion with consequential seniority and have given the stamp of approval to the Act and the Rules cannot withstand close scrutiny inasmuch as the Constitution Bench has clearly opined that Articles 16(4-A) and 16(4-B) are enabling provisions and the State can make provisions for the same on certain basis or foundation. The conditions precedent have not been satisfied. No exercise has been undertaken. What has been argued with vehemence is that it is not necessary as the concept of reservation in promotion was already in vogue. We are unable to accept the said submission, for when the provisions of the Constitution are treated valid with certain conditions or riders, it becomes incumbent on the part of the State to appreciate and apply the test so that its amendments can be tested and withstand the scrutiny on parameters laid down therein.

 … In the ultimate analysis, we conclude and hold that Section 3(7) of the 1994 Act and Rule 8-A of the 2007 Rules are ultra vires as they run counter to the dictum in M. Nagaraj. Any promotion that has been given on the dictum of Indra Sawhney and without the aid or assistance of Section 3(7) and Rule 8-A shall remain undisturbed.” [Emphasis supplied]

The result of this judgment was, essentially, that all SC/ST candidates in the State services in U.P. would be subjected to reversion, i.e. they would be demoted to the posts they held as on 15.11.1997 (when the five year period of protection provided by Indra Sawhney ended), as the provisions on the basis of which they had been promoted had been effectively struck down with retrospective effect (reaching all the way back to 1997).

The nature of the “exercise” contemplated is set out in Paragraph 81, where Justice Misra summarized the ratio of M. Nagaraj: –

“(v) The State has to form its opinion on the quantifiable data regarding adequacy of representation. Clause (4-A) of Article 16 is an enabling provision. It gives freedom to the State to provide for reservation in matters of promotion. Clause (4-A) of Article 16 applies only to SCs and STs. The said clause is carved out of Article 16(4-A). Therefore, clause (4-A) will be governed by the two compelling reasons—“backwardness” and “inadequacy of representation”, as mentioned in Article 16(4). If the said two reasons do not exist, then the enabling provision cannot be enforced.[Emphasis supplied]

There are two serious errors in this judgment. The first is a doctrinal error, which has already been pointed out, namely that no separate exercise to determine “backwardness” is required for SCs/STs, as the backwardness is presumed in the case of Schedule Castes and Scheduled Tribes, as per previous judgments of the Court.

The second error is based on equity, and is useful for understanding why Article 16 should be treated as a hybrid containing both a power and a duty. If, indeed, quantifiable data is to be collected for applying Articles 16(4A) and 16(4B), then, surely, even if the Court’s ruling is not prospectively applied (to protect those promoted prior in time due to the State’s errors), quantifiable data demonstrating that SCs/STs are adequately represented in various echelons of the State services would be a prerequisite for reversion. Assuming that SCs/STs are overrepresented in the higher echelons of the State services because of the State’s failure to collect data, the resultant reversion, in the absence of data, would probably lead to a situation where SCs/STs are underrepresented in the higher posts. The Court, however, failed to clarify this and, instead, in Paragraph 87, only stated that those promotions protected by the Indra Sawhney judgment would remain undisturbed.

Inevitably, contempt petitions came to be filed, as the Govt. of U.P. failed to either seek clarifications of the Supreme Court’s judgment, or take steps to collect “quantifiable data”. On 05.07.2013, the Supreme Court issued notice in the lead petition [Contempt Petition (C) No. 214/2013]. Instead of working to collect the data, the Govt. of U.P. instead took hasty steps to revert all those in the State services to the posts they held as on 15.11.1997. As recorded in the Court’s order of 13.10.2015 in Contempt Petition (C) No. 214/2013, 15,226 persons have been reverted as a result, all without the State having done anything to compile the data.

In this background, a number of SC/ST personnel, who have been reverted, have filed Writ Petitions before the Supreme Court, and some of these petition contain a positive averment to the effect that there is a “power coupled with a duty” in Article 16 to level the playing field.

Let us now consider the arguments. I contend that the problem with the way Article 16 [particularly 16(4), (4A) and (4B)] is read currently is that it is essentially asymmetrical. For ameliorative measures under these provisions, quantifiable data on backwardness and inadequacy of representation is a prerequisite, contributing to the formation of an opinion by the State. On the other hand, for reversion, no such data needs to be collected, and no opinion needs to be formed. Similarly, the current law on Article 16 allows the State to completely ignore the question of whether backward classes, SCs and STs are adequately represented in public services, by simply omitting to collect data, which is required for the formation of an “opinion” under Article 16(4), (4A) and (4B). Therefore, the State can essentially omit the entire process of forming an “opinion”, by refusing to collect the necessary data. There is a hurdle created to pull up backward groups, but none for pulling them down, or for ignoring them entirely. This, I believe, is inconsistent with the equality code of our Constitution, and Article 16 itself contains no textual basis for such asymmetry. It is this asymmetry problem that can be remedied if the Supreme Court recognizes the fact that Article 16 contains within it an enabling power coupled with a positive duty.

While I will elaborate on this in later posts, this is how I conceive of the hybrid powers and duties contained in Article 16: –

  1. There is a positive duty on the State to periodically collect the data required for the formation of an opinion under Articles 16(4), 16(4A) and 16(4B).
  2. Once the data is collected, it is the duty of the State to apply its mind on backwardness, adequacy of representation and efficiency (i.e. the Nagaraj troika).
  3. In applying its mind, the State has discretion in the line-drawing exercise involved in deciding which group is backward, what degree of representation is adequate, and what is the optimum balance in the interest of efficiency.
  4. Based on the above, the State has the discretion to decide if any ameliorative measures are required. As discussed in Indra Sawhney, this need not necessarily be through reservation (or reservation in promotion), but through lesser concessions as well. Therefore, the State’s discretion would also extend to deciding the nature of ameliorative measure.

Therefore, I argue for a reading of Article 16 where the State has discretion in deciding whether ameliorative measure are required for backward classes/SCs/STs, and the extent and nature of of those measures, but, prior to that, the State would bear the positive responsibility of collecting data (as per Nagaraj) and then forming an “opinion”, whatever that opinion may be. In other words, the choice that is eliminated is complete inaction on the part of the State to apply its mind to the issue of adequacy of representation of backward groups in the State services, whether that inaction is a symptom of policy paralysis or political expediency.

(The author is a Supreme Court advocate)

Guest Post: Does Article 16 Impose a “Power Coupled with a Duty” upon the State? – I

(In an upcoming series of guest posts, Karan Lahiri argues that, properly interpreted, Article 16 of the Constitution not only permits, but actually casts an affirmative duty upon the State to ensure the adequate representation of backward classes of citizens in public employment.)

As this blog already contains an excellent analysis on the topic of reservation, in the ten-part series titled “Reservation, Equality and the Constitution” (in which I would particularly encourage the reader to look at Parts III, IV, V and VII), I think I can skip any preliminaries on the contours of Article 16, and jump right to the question at hand – Does Article 16 of the Constitution of India confer on the State a power coupled with a duty to ensure that backward classes of citizens are adequately represented in public services? Further, is there a similar power coupled with a duty to provide for reservations in matters of promotion for those belonging to Scheduled Castes and Scheduled Tribes which are not adequately represented in the public services under the State?

In “Reservations, Equality and the Constitution – VI: The Middle Way”, it was correctly mentioned that, in Ajit Singh (II), the Supreme Court of India concluded that “Arts. 16(4) and 16(4A) are enabling provisions, and do not confer fundamental rights” and, further, that “the Court also rejected the argument that these Articles conferred a power coupled with a duty, overruling the cases of Jagannathan and Kuldeep Singh.”

What I will attempt to argue, in a series of posts, is outlined below: –

  1. I will first explain the concept of “power coupled with duty” (which has its roots in English administrative law principles).
  2. Based on a set of cases pending before the Supreme Court at the moment, I will attempt to explain why answering the questions posed in the first paragraph in the affirmative leads to the most just and equitable solution, in keeping with the principle of equality as it exists in the Constitution of India.
  3. Finally, I will demonstrate why Ajit Singh (II) is incorrectly decided, and how it conflated the issue of “power coupled with duty” with the issue of whether there is a fundamental right to reservation. In this part, I will also try to demonstrate that reading Article 16 as containing a power coupled with a duty is not only the most just and equitable solution to the interpretive issue at hand, but is also the most consistent with the other strands of precedent that exist regarding the interpretation of Article 16.

(At this point, it is important to clarify that I am NOT arguing that the State has a positive duty to specifically provide for reservations. The scheme of Article 16 (which I believe to be the proper interpretation), in respect of where the State’s duty ends and its discretion begins, will be set out in subsequent posts..)

Degrees of Discretion – Understanding the concept of “power coupled with duty”

 To understand the interpretive question posed in the first paragraph of this post, we must look at the scheme of Article 16 once again: –

“16. Equality of opportunity in matters of public employment.—

(1)There shall be equality of opportunity for all citizens in matters relating to employment or appointment to any office under the State.

(4) Nothing in this article shall prevent the State from making any provision for the reservation of appointments or posts in favour of any backward class of citizens which, in the opinion of the State, is not adequately represented in the services under the State.

(4A) Nothing in this article shall prevent the State from making any provision for reservation in matters of promotion, with consequential seniority, to any class or classes of posts in the services under the State in favour of the Scheduled Castes and the Scheduled Tribes which, in the opinion of the State, are not adequately represented in the services under the State.

(4B) Nothing in this article shall prevent the State from considering any unfilled vacancies of a year which are reserved for being filled up in that year in accordance with any provision for reservation made under clause (4) or clause (4A) as a separate class of vacancies to be filled up in any succeeding year or years and such class of vacancies shall not be considered together with the vacancies of the year in which they are being filled up for determining the ceiling of fifty per cent reservation on total number of vacancies of that year.


Both Article 16(4) and 16(4A) are framed as – Nothing in this article shall prevent the State from making any provision for [X], if in the opinion of the State, [Y] exists. This is, of course, a broad approximation, made only to simplify the language of the provision for the purpose of interpretation. The interpretive issue is whether the framing of these provisions, which appear to be enabling or permissive, can ever be read as conferring not merely a discretionary power but a concomitant duty on the State. In this first part, I will not be looking at the interpretation of Article 16 itself. Instead, I will try to explain the concept of “power coupled with duty”, and explain how this hybridity is often located in provisions containing permissive language, like the language contained in Article 16(4) and (4A).

One of the best expositions on the subject of discretion is contained in “De Smith’s Judicial Review”. On the subject of discretionary power, De Smith states: –

“The concept of discretion in its legal context implies power to make a choice between alternative courses of action or inaction. If only one course can lawfully be adopted, the decision taken is not the exercise of a discretion but the performance of a duty. To say that somebody has a discretion presupposes that there is no unique legal answer to a problem. There may, however, be a number of answers that are wrong in law. And even in cases where the power is discretionary, circumstances can exist which mean the discretion can only be exercised in one way. There are degrees of discretion, varying the scope for manoeuvre afforded to the decision maker.[Emphasis supplied]

 What flows from this is that even though discretion may be conferred on a State authority, the law may limit the choices that may be made in the exercise of that discretion.

However, when a provision is couched in permissive language, can it be read in a manner that imposes a duty? More specifically, can permissive language, at the very least, suggest that while various courses of action are open to a particular State authority to remedy a particular problem, the State authority can never ignore the problem itself and refrain from at least applying itself to the problem and exercising its discretion? The position in English law, which we have adopted, is summarized in De Smith’s Judicial Review as under: –

“Sometimes, the question before a court is whether words which apparently confer a discretion (words such as “may”, or “it shall be lawful if”) are instead to be interpreted as imposing a duty. The word “may” has over the years, primarily been construed as permissive, not imperative. However, exceptionally, it was construed as imposing a duty to act, and even a duty to act in one particular manner.”

Interestingly, one of the authorities cited in De Smith’s treatise is Julius v. Bishop of Oxford, (1880) 5 App. Cas. 214, which the Supreme Court of India has repeatedly relied on for the proposition that permissive language may often imply a duty to act. The most famous of these, perhaps, is Ambica Quarry Works v. State of Gujarat, (1987) SCC 213, where Justice Sabyasachi Mukharji observed: –

“Often when a public authority is vested with power, the expression “may” has been construed as “shall” because power if the conditions for the exercise are fulfilled is coupled with duty… “May” is a permissive or enabling expression but there are cases in which for various reasons as soon as the person who is within the statute is entrusted with the power, it becomes his duty to exercise it. As early as 1880 the Privy Council in Julius v. Lord Bishop of Oxford explained the position. Earl Cairns, Lord Chancellor speaking for the judicial committee observed dealing with the expression “it shall be lawful” that these words confer a faculty or power and they do not of themselves do more than confer a faculty or power. But the Lord Chancellor explained there may be something in the nature of the thing empowered to be done, something in the object for which it is to be done, something in the conditions under which it is to be done, something in the title of the person or persons for whose benefit the power is to be exercised, which may couple the power with a duty, and make it the duty of the person in whom the power is reposed, to exercise that power when called upon to do so. Whether the power is one coupled with a duty must depend upon the facts and circumstances of each case and must be so decided by the courts in each case. Lord Blackburn observed in the said decision that enabling words were always compulsory where the words were to effectuate a legal right.[Emphasis supplied]

A lesser known, yet interesting, examination of “power coupled with duty” is contained in Madhav Rao Jivaji Rao Scindia v. Union of India, (1971) 1 SCC 85, commonly referred to as the Privy Purses case. In this case, the subject matter of challenge was a Presidential Order withdrawing recognition to Rulers of Princely States. One of the questions that arose is whether Article 366(22) of the Constitution, which invested the President with the power to recognize Rulers, also contained within it a power to withdraw such recognition. It was argued by the Petitioner that the President’s power was coupled with a duty to maintain the constitutionally recognized institution of “Rulership” and, therefore, there could be no power to arbitrarily withdraw recognition and extinguish a constitutionally recognized concept. Justice J.C. Shah’s majority judgment, in upholding this contention, explained the concept of “power coupled with duty” as follows: –

“117. There are many analogous provisions in the Constitution which confer upon the President a power coupled with a duty. We may refer to two such provisions. The President has under Articles 341 and 342 to specify Scheduled Castes and Scheduled Tribes and he has done so. Specification so made carries for the members of the Scheduled Castes and Scheduled Tribes certain special benefits e.g. reservation of seats in the House of the People, and in the State Legislative Assemblies by Articles 330 and 332, and of the numerous provisions made in Schedules V and VI. It may be noticed that Scheduled Castes and Scheduled Tribes are specially defined for the purposes of the Constitution by Articles 366(24) and 366(25). If power to declare certain classes of citizens as belonging to Scheduled Castes and Scheduled Tribes includes power to withdraw declaration without substituting a fresh declaration, the President will be destroying the constitutional scheme. The power to specify may carry with it the power to withdraw specification, but it is coupled with a duty to specify in a manner which makes the constitutional provisions operative.[Emphasis supplied]

Justice Hegde drew a similar analogy: –

 “… In my opinion Article 366(22) imposes a duty on the President and for that purpose has conferred on him certain powers. In other words the power conferred on the President under that provision is one coupled with duty. There are similar powers conferred on the President under the Constitution. Under Chapter XVI of the Constitution certain special provisions were made for the benefit of the Scheduled Castes and certain Scheduled Tribes. Seats were reserved for them both in the Parliament as well as in the State Assemblies. Certain other benefits were also secured to them in the matter of appointments to services and posts in connection with the affairs of the Union or of a State. But the Constitution did not specify which castes were Scheduled Castes and which Tribes were Scheduled Tribes. Under Articles 341(1) and 342(1) of the Constitution, the President was given power to specify the castes which he considered to be Scheduled Castes and the Tribes which he considered to be Scheduled Tribes. Though both the Articles say the President “may” specify the castes which he considers as Scheduled and Tribes which he considers Scheduled, it is clear that a constitutional duty was imposed on him to specify which castes were Scheduled Castes and which tribes were Scheduled Tribes for the purpose of the Constitution. The word “may” in those clauses must be read as “must” because if he had failed or declined to specify the castes and tribes, Articles 330, 332, 334, 335, 338 and 340 would have become inoperative and the constitutional guarantees given to the Scheduled Castes and Scheduled Tribes would have become meaningless.[Emphasis supplied]

The reason I favour this explanation of “power coupled with duty” is not because of its explanation of the legal principle itself. Indeed, the interpretive concept is explained better in Ambica Quarry, and also in State v. I.K. Nangia, (1980) 1 SCC 258. However, the Privy Purses case appeals to me because it crystallizes the principle that, in a constitutional context, constitutional provisions which, at first blush, appear to confer discretion, ought to be read as conferring a power coupled with a duty, when the authorities failure or refusal to act would nullify a portion of the constitutional scheme, however small that portion might be. This is wider than the iteration of this concept in Ambica Quarry, which only speaks of a duty being implied if a legal right is to be effectuated.

Let’s call this principle, apparent in the Privy Purses case, the rule of constitutional consistency — a constitutional provision conferring power/discretion on a State authority, couched in permissive language, is to be treated as a provision containing a power coupled with a duty, if the failure or conscious omission on the part of such authority to act would nullify the effect of another/other constitutional provision, or render nugatory a constitutional principle emerging from a mosaic of constitutional provisions.

In the next essay, I will substantiate this argument in the specific context of Article 16. By discussing a set of pending writ presently petitions before the Supreme Court, I will show that there is an asymmetry problem in the way that Article 16 as it is currently read, such that there is a hurdle for ameliorative measures to be put in place, but there is nothing to stop the State from entirely ignoring the under-representation of backward classes in the State service, or even from reverting them to lower posts for no better reason than political expediency.

(Karan Lahiri is an advocate practicing at the Supreme Court.)

The Supreme Court on Reservation Benefits for the Reconverted

In an interesting judgment delivered last month, a two-judge bench of the Supreme Court held that, subject to certain conditions, an erstwhile Christian who “reconverted” to Hinduism was eligible for Scheduled Caste reservation benefits. On March 12, the Indian Express carried a highly critical appraisal of the judgment. For reasons that I will shortly explain, I do not agree with the article. On the contrary, I feel that in terms of the limited question before it, the case was correctly decided – and furthermore, it also contains the germs of a new, progressive jurisprudence dealing with conversions and affirmative action.

The appellant, K.P. Manu, had a great-grandfather who belonged to the Hindu Pulaya community (which is a Scheduled Caste). His son – i.e., the appellant’s grandfather – converted to Christianity. The appellant’s father was also a christian. But at the age of 24, the appellant “reconverted” to Hinduism. He applied for a caste certificate from his community and from the local tehsildar, both of whom issued it to him. This was challenged before a Scrutiny Committee, which recorded a finding that the caste certificate was incorrectly issued, on the ground that the appellant’s father and grandfather had been christians, and that even after his “reconversion”, the appellant had married a Christian woman.

On the basis of the Scrutiny Committee’s report, the State Government directed the appellant’s employer to remove him from service and recover Rs. 15 lakhs from him. This was challenged before the High Court, which affirmed the Committee’s findings. Consequently, the appellant approached the Supreme Court.

Therefore, the Supreme Court was faced with a very narrow question: is a reconvert from Christianity entitled to reservation benefits if his original/ancestral caste is a Scheduled Caste? To answer the question, the Court analysed a number of its previous judgments dealing with the issue of conversions and affirmative action. According to the established jurisprudence of the Supreme Court, “ordinarily”, conversion from Hinduism would mean an expulsion from one’s caste. This was not a hard-and-fast rule though, since admittedly, the caste system been partially incorporated into other religions as well. Consequently, the question would be decided on a case-to-case basis, depending upon the “structure of the caste, and its rules and regulations.”

Because, ultimately, the structure of the caste, and its rules and regulations, were all internal matters, the Court held that a “reconvert” from Christianity would be treated as belonging to his original caste if “the members of the caste accept him.” (paragraph 13) The basic principle, as explained by a prior Constitution Bench decision, was that on conversion from Hinduism, one’s original caste was “eclipsed”, and if “during his/her lifetime the person is reconverted to the original religion the eclipse disappears and the caste automatically revives.”

The State argued that the Court’s prior decisions were limited to reconversion either within the same generation, or – at most – situations where one’s parents converted to Christianity, and one converted back. It contended that the “eclipse theory” could not be extended beyond a generation. In paragraph 33, the Court rejected this contention, holding that “as we understand the authority it does not lay down that it only would apply to the parents and exclude the grandparents.” The Court observe – in my opinion, correctly – that caste stigmas tended to persist through generations (and even upon conversions) – and that therefore, “eclipse-and-revival” was not limited to a single generation. Therefore, it held:

“In our considered opinion, three things that need to be established by a person who claims to be a beneficiary of the caste certificate are (i) there must be absolutely clear cut proof that he belongs to the caste that has been recognised by the Constitution (Scheduled Castes) Order, 1950; (ii) there has been reconversion to the original religion to which the parents and earlier generations had belonged; and (iii) there has to be evidence establishing the acceptance by the community. Each aspect according to us is very significant, and if one is not substantiated, the recognition would not be possible.” (para 34)

All three, of course, were empirical questions.

In my opinion, it is difficult to find fault with either the reasoning or the conclusion. On the specific issue of reconversion and the benefits of reservation, the Court was bound by previous three-judge and Constitution Bench decisions. The innovation in this case is the holding that eclipse-and-revival does not cease if there is a generational gap between conversion and reconversion. I think that the Court is correct to ground this in the fact that social and economic disabilities – which is the entire raison d’etre for reservations – do not disappear with conversion, even across generations.

The Indian Express article has two broad criticisms of the judgment. First, it argues that this constitutes an incentive to “ghar wapsi“. That may or may not be so, but as we have just discussed, that question was not res integra, since it has already been established that reservation benefits can be availed of in cases of reconversion. In any event, ought the Court have held that one is not entitled to reservation benefits on reconversion? That, to me, seems to be as morally fraught a position as the incentive-to-ghar-wapsi.

Secondly, the Express piece argues that in holding that a person loses reservation benefits upon conversion to Christianiaty (or Islam), the judgment enshrines a “majoritarian view” of religion. Since the Constitution treats Buddhism and Sikhism as falling within the category of “Hinduism”, it discriminates against non-Indic religions by creating a situation where a person will not lose reservation benefits if she converts to Buddhism/Sikhism, but will if she converts to Islam/Christianity. The piece also argues that it is incorrect to assume that social and economic disabilities disappear on conversion. The focal point of its criticism is contained in the following quote that it excerpts:

“The court observes: “Once such a person ceases to be a Hindu and becomes a Christian, the social and economic disabilities arising because of Hindu religion cease, and hence it is no longer necessary to give him protection…””

At this point, it is crucial to note that this was not an observation of the Court. It is a quote from the previous judgment of Arumugam vs Rajgopal, a previous three-judge bench that the present Court was bound by. Now there is no doubt that this is a very unfortunate legal position, and as a matter of simple fact, patently untrue. Caste-based social and economic difficulties are far too deeply entrenched to be ameliorated by a formal conversion. So, the Indian Express piece is correct to note that:

“Caste consciousness is part of our everyday lives; its universality renders it normal. However much we may delude ourselves, we are all scarred by it. Dalits, irrespective of their religious beliefs, are oppressed victims of deep prejudice and the irrational, wicked belief in Dalit inferiority. To make the argument that the caste system is not recognised by Christianity and Islam is to ignore the realities of the country.”

That said, however, the question of losing caste – and thereby, reservation benefits – upon conversion was not before the Court, and it could not have decided upon it. Furthermore, being a two judge bench, it was entirely bound by precedent, and could not have changed this position even if it wanted to.

That said, a close reading of the judgement reveals that the Court was acutely aware of the problem, and to the extent possible, attempted to craft the beginnings of an alternative jurisprudence. From paragraphs 22 to 31, it quoted B.R. Ambedkar, James Massey, Archbishop George Zur, the Mandal Commission, a Church of the South India Commission, the Chinnappa Commission Report, and a political scientist, all for the proposition that caste-disabilities do not disappear upon conversion. Summing up these observations, it observed in paragraph 32:

“We have referred to the aforesaid materials and the observations singularly for the purpose that there has been detailed study to indicate the Scheduled Castes persons belonging to Hindu religion, who had embraced Christianity with some kind of hope or aspiration, have remained socially, educationally and economically backward.”

The Court then used this fact to reject the State’s contention that the eclipse-and-revival theory was limited to a single generation. If the caste stigma was never lost, even upon conversion, then why ought there be a denial of the reservation entitlement upon reconversion.

While this seems to be in line with Arumugam’s Case, there is a crucial conceptual shift. In Arumugam’s Case, because the Court first held that reservation benefits are lost upon conversion because the social/economic disabilities disappear, its eclipse-and-revival conclusion must be based upon the corollary that on reconverting, one falls right back into those very disabilities that one had escaped in the first place by converting. In the present case, however, the Court extends the eclipse-and-revival theory to multiple generations on the very different reasoning that caste-based disabilities are not lost upon converting and, indeed, persist throughout generations. This is actually precisely in line with the Indian Express critique!

But if caste disabilities are not, as it turns out, lost upon converting, then the inevitable conclusion must be that reservation benefits should not be lost upon conversion. This is because, as the Court as held multiple times, the basis of reservations is social and economic backwardness, and “caste” is simply a convenient (but by no means only) marker of measuring such backwardness. Holding that reservation benefits remain even after conversion, because the ultimate justification (backwardness) is not altered, would be well in line with the Court’s core affirmative action jurisprudence and philosophy.

In K.P. Manu’s Case, the Court has recognised these realities, and its reasoning on eclipse-and-revival directly undermines the established jurisprudence on loss-of-benefits-upon-reservation. It represents the possibility of an alternative jurisprudence that is much more sensitive to the lived realities of caste, disabilities and discrimination.

Reservations, Equality and the Constitution – IV: Indra Sawhney v. Union of India – Background and Preliminaries

In the last post, we saw how State of Kerala v. N.M. Thomas marked a shift in the conception of equality that the Constitution is committed to. It did so by abandoning the existing individualistic, colour-blind equality model that presumptively proscribed all classifications on prohibited bases (sex, race, caste etc.), whether for benign purposes or otherwise, unless explicitly permitted. We also noticed, however, that the seven-judge N.M. Thomas bench split in a manner that left it entirely unclear what the new model was. What four judges (a majority) did agree on was that Article 16(4) was not an exception to, but a facet of Article 16(1) – that is, the justification for affirmative action could be found within the conception of equality of opportunity under 16(1), and by extension, within the broader conception of equality under Article 14. 

There are, however, at least two distinct ways in which Article 16(1) can be interpreted, and no clear consensus emerged in N.M. Thomas on that point. Both interpretations are sensitive to group inequalities, but in different ways. Under the first interpretation (to fix ideas, let it remain the “group subordination principle”), the point of affirmative action is to achieve equality between groups. In other words, the site of constitutional action is the group, and the right (to equality) that is being vindicated is that of the group’s. The reason for this approach is that historically, injustice has been meted out to groups (women, dalits etc.), and thus, it makes sense for remedial action to be focused on groups.

The second interpretation is also sensitive to the realities of historical discrimination, but even so, it insists that ultimately, the right to equality is an individual right. However, since historically, the individual has been discriminated against in virtue of her membership of a particular group (e.g., as a woman, or as a Dalit), it might be necessary to take group identity into account, in order to achieve truly effective ameliorative measures that take us towards an eventual position  of genuine equality for all persons.

Notice that both these approaches support affirmative action, based on a historical understanding of the way in which discrimination has worked in our society, but for radically different reasons. The first – group-subordination approach – is geared towards preserving group identities by equalising groups qua groups. The second approach aims at eventual elimination of group identity, accepting group-based classifications as a necessary set of evils on the road to a society in which all individuals are brought to a position of genuine equality. The debate is not merely academic because, as we shall see, it deeply affects the models of reservation and affirmative action that are justified under Articles 15 and 16.

While N.M. Thomas left this basic philosophical point unresolved, this – and many other questions – came to a head in Indra Sawhney v. Union of Indiadecided by a nine-judge bench in 1992. Indra Sawhney saw a plurality opinion, written by Jeevan Reddy J. for himself and three other brother judges, and four separate concurrences. Indra Sawhney, therefore, has no majority opinion, and so, special care must be taken in determining what propositions, precisely, the case stands for.

Before diving into the case, however, in light of the sheer number and complexity of the issues raised, a brief historic background is necessary.

Reservation in India has a long history. Article 16(4) permits the government to make reservations in the matter of employment for backward. The first question which arises, therefore, is “what constitutes a backward class?” To answer this question, the Kalelkar Commission was set up in 1953. The commission identified six markers to identify a backward class, these being: 1) Traditional occupation and profession, 2) Literacy, 3) Population, 4) Distribution and concentration, 5) Social position in the caste hierarchy and 6) Representation in the Govt. service, or the industrial sphere.

The findings of the commission were criticized by a Central Government memorandum in 1956. Subsequently, in 1961, the Central Government agreed to let each State Government draft its own list of backward classes. This was carried out by individual State Governments.

In 1979, by a Presidential Order under Article 340 of the Constitution, the Mandal Commission was set up to identify backward classes. After an exhaustive survey, the Mandal Commission came up with eleven indicators of “social and educational backwardness,” which, in turn, were grouped into three broad headings – social, economic and political. Social indicators included castes/classes considered backward by others, caste/classes depending upon manual labour for their livelihood, castes/classes with low average ages of marriage and castes/classes with a low proportion of female workforce. Educational criteria included percentages of school attendance, dropouts and matriculation. Economic criteria included value of family assets, number of families living in kucha houses, distance of sources of drinking water, and households having taken consumption loans. Social indicators were given three points each, educational indicators two points, and economic indicators one point. The total points, therefore, amounted to 22. All castes/classes that had a score of above eleven, were deemed to be “socially and educationally backward.” The commission then used some further criteria to identify “other backward classes,” and also from the non-Hindu section of the population. In the final analysis, the Mandal Commission found that the percentage of OBCs in India was 52%. In the view of the Supreme Court decision (Balaji) limiting reservation to less than 50%, the Commission recommended a 27% reservation for OBCs in addition to the existing 22.5% reservation for SCs and STs. Subsequently, there were two Government Memoranda (an executive order) which enforced these recommendations, and provided that Backward Classes would mean “those castes and communities” that were common to the Mandal Commission report, and various State Government lists. The Mandal Commission report proved to be bitterly controversial, stoking tensions all over the country, and eventually, it was these memoranda that were challenged. A nine-judge bench was constituted, and Indra Sawhney came to Court.

One major sticking point was the use of caste to identify a 16(4) backward class, and this itself had had a troubled history. In Balaji, the Court had held that caste cannot be the sole, determining factor for identification of backward class. Subsequently, though, in P. Rajendran v. State of Madras, the list of “other backward classes” simply included a list of castes. The Court held that although a selection based purely on caste would fall foul of Art. 15(1), if it could be shown that the castes which were selected were, in fact, socially and educationally backward, 15(4) would apply, and the selection would be saved from invalidity. In a certain sense, therefore, P. Rajendran can be seen to have watered down the judgment in Balaji to an extent. This was followed in a number of cases such as Triloki Nath, Periakurappum v. State of Tamil Nadu and State of Andhra Pradesh v. Balram. However, in State of Uttar Pradesh v. Pradip Tandon, Ray C.J. came up with a very different interpretation: he argued that caste or religion could not be made a basis for identification under Art. 15(4), as this would stultify the operation of Art. 15(1). This judgment, of course, seemed to go against the stare decisis evolved in Rajendran and beyond – and N.M. Thomas then added yet another twist to the tale by holding  that the word “caste”, as a proscribed basis under 16(2), was not the same “caste” as the “caste” in “Scheduled Caste” – the latter being a purely administrative category. The Indra Sawhney Court, faced with these precedents, was called upon to resolve the role of caste and its relationship with class in the context of affirmative action, once and for all.

Before we delve into the core of Justice Reddy’s plurality opinion, one last point remains to be noted. In his opinion, Justice Reddy cites – and discussed in some detail – four American cases on affirmative action: DeFunis v. OdegaardRegents of the University of California v. Allan BakkeFullilov v. Philip M. Klutznick and Metro Broadcasting Inc. v. Federal Communications Commission.

In Defunis, a percentages of seats in the University of Washington Law School were reserved for racial minorities. Although by the time the case went up to the Supreme Court, Defunis, who had filed the case, was in his final term, and therefore the question had become academic, Douglas J. still wrote an extensive judgment. The crux of his argument was that the Equal Protection Clause does not require Law Schools to admit students purely on the basis of testing results and undergraduate grades; on the other hand, it is equally permissible to view a candidate’s prior record, including the barriers he had to overcome to arrive at where he is. Taking the example of a Black student who had pulled himself out of a ghetto to arrive at Law School, and the son of a wealthy alumnus who got into Harvard, Douglas J. said that although, for obvious reasons, the Harvard man might achieve better grades throughout Law School, in the long run, the Black student might have a more successful career because of the perseverance and undoubted talent that had seen him even reach Law School. This logic was rejected in Regents of the University of California v. Allan Bakke, however, where again, the issue was regarding racial reservations in a University. By a narrow margin (5:4), the Supreme Court held racial quotas to be unconstitutional, and applied judicial “strict-scrutiny” to race-based classifications, holding it justified on the narrow ground only, of achieving a diverse student body.

The other two cases were business-related. In Fullilov v. Philip M. Klutznick, by an Act it was required that at least 10% of federal funds granted for local projects must be utilized in procuring supplies from business owned by minority groups. A challenge to this Act failed, the Court holding that the legacy of unequal treatment now justified differentiation and positive discrimination. Similarly, in Metro Broadcasting Inc. v. Federal Communications Commission, minority groups were given certain preferences in the matter of distribution of licenses. The Court held that benign race conscious measures are constitutionally permissible even if they are not designed to compensate victims of past governmental or societal discrimination so long as they serve important governmental objectives and are substantially related to achievement of those objectives.

What is particularly remarkable about the Court’s survey of American law was the one judgment it omitted to cite, decided three years before, in 1989: City of Richmond v. Croson, which was in direct opposition to Fullilov and Metro Broadcasting (the difference tracks complicated issues of American federalism that our irrelevant for our purposes here). Croson prohibited race-based classifications aimed at ameliorating past societal inequalities, holding that any such remedial action would have to be strictly limited to intentional and specific discrimination. 

In other words, by citing four cases out of which three were largely sympathetic to affirmative action, but out of which one was a moot judgment, and omitting the hostile case of Croson, the Court painted a skewed picture of American law, one that was more sympathetic to group-based philosophies of affirmative action than actual American law. Perhaps it did so to set the stage for its own group-based discussion, that we shall come to in the next post.

Reservations, Equality and the Constitution – III: State of Kerala v N.M. Thomas and the Transformation of Equality

At the beginning of 1976, the meaning of the 14-15-16 equality code seemed to be reasonably well-settled. Twenty-six judges over an equal number of years had consistently affirmed that affirmative action was constitutional by virtue of the existence of Articles 15(4) and 16(4), both of which operated as exceptions to the equality and non-discrimination provisions of 15(1) and 16(1). Not just the outcomes, but judicial reasoning as well, as we have seen, relied upon a colour-blind vision of equality, one that considered any kind of classification on the bases of prohibited categories (race, religion, sex etc.) as presumptively unconstitutional. This idea, in turn, was grounded upon the belief that in distributing a benefit or burden, government must treat individuals as  individuals, and not as members of groups.

The lone dissenting voice had been Justice Subba Rao’s, in Devadasan. And in 1976, a seven-judge bench, in State of Kerala v. N.M. Thomas, adopted that view in full, and in doing so, entirely repudiated existing precedent. As a transformative decision, N.M. Thomas stands alongside Kesavananda Bharati and Menaka Gandhi in our constitutional history: it not only changed the constitutional understanding of reservations, but in doing so, transformed the deep foundations of the idea of equality that the Constitution commits us to.

In N.M. Thomas, a governmental order granted provisional promotions to members of Scheduled Castes and Scheduled Tribes who did not have the requisite qualifications to be eligible for such promotion, along with a two year grace period for them to gain such qualifications. This was challenged. The key contention of the aggrieved parties was that the classification was clearly void under Articles 16(1) and (2), and not covered by Article 16(4). Under existing precedent, this ought to have been a simple case. If Article 16(4) did not apply, then special provisions for SCs and STs clearly did violate Articles 16(1) and (2).

Not so, held the Court. In Paragraph 31, the Chief Justice Ray, writing the judgment of the Court, held:

“The rule of differentiation is enacting laws differentiating between different persons or things in different circumstances. The circumstances which govern one set of persons or objects may not necessarily be the same as those governing another set of persons or objects so that the question of unequal treatment does not really arise between persons governed by different conditions and different sets of circumstances.


“The rule of equality within Articles 14 and 16(1) will not be violated by a rule which will ensure equality of representation in the services for unrepresented classes after satisfying the basic needs of efficiency of administration.” (Para 37)

The Court thus holds that Article 16(1)’s conception of equality itself includes remedial action to ensure due representation for hitherto excluded classes. No longer, then, is equal treatment to be accorded to individuals; the Court adopts the group-subordination idea of equality, one that locates the site of historical discrimination as the group, and seeks to remedy that by targeting groups.

That said, the bar of 16(2) remained, which expressly listed caste as a prohibited basis of classification. The Court’s answer was to accept that 16(2) would instantaneously void such classification (Paragraph 37), but then to perform a sleight of hand (one that would be repeated subsequently) by holding that Scheduled Castes were not the same as castes (under 16(2)), and could fairly be called “backward classes” within the meaning of 16(4) (Para 43) This raises a troubling issue for those situations where this sleight of hand will not be available to the Court. What, for instance, can the Court say if Government wishes to make special provisions in employment for women? 16(2) expressly prohibits classification on the basis of sex, and purely on the logic of N.M. Thomas, 16(2) continues to operate as an absolute bar.

In any event, the fundamental shift in N.M. Thomas is this: 16(4) no longer operates as an exception to 16(1), allowing the government limited powers to do what it would otherwise be prohibited from doing by virtue of 16(1), but now exists as “one of the methods of achieving equality embodied in Article 16(1).” (Paragraph 46) But this necessarily involves a shift in the nature of equality itself, because until now, 16(1) and 16(4) were considered to embody differing visions of equality: 16(1) was about individuals, and 16(4) – textually – about remedial measures targeted at historically discriminated groups. And now it is that latter logic that governs both 16(1) and 16(4).

On what basis does the Court deal with precedent in such cavalier fashion? The majority does not say, and it is in Justice Mathew’s concurring judgment that we find an articulate defence of the new principle of equality. Drawing upon sources as diverse as Brandeis and Laski, Justice Mathew conceded that Article 16(1) was about equality for individuals – equality of opportunity. But equality of opportunity, he argued, meant that in distributing a benefit or burden, the State must set those criteria for selection that “people from all sections of the society have an equal chance of satisfying them.” (Paragraph 84) To sharpen the idea, he drew upon Bernard Williams’ famous example: in a certain society, the most prestigious role is that of the warrior, which requires great physical strength. Hitherto, recruitment for this role has been – formally – solely from the wealthy/propertied classes. A reform movement succeeds in removing the formal barrier in favour of equal competition. Nothing changes, however, because the rest of the population is so undernourished by reasons of poverty, that they do not pass the physical requirements of the recruitment test. In other words, the criteria for selection has been changed from wealth to strength, but it so happens that – for obvious reasons – the poor also happen to be weak.

For Justice Mathew, this is not equality of opportunity. According to him: “To give X and Y equality of opportunity involves regarding their conditions, where curable, as themselves part of what is done to X and Y and not part of X and Y themselves.  Their identity for this purpose does not include their curable environment, which is itself unequal and a contributor of inequality.” (Paragraph 87)

This is a crucial philosophical point (and the connection to Amartya Sen will be obvious to everyone). Much turns upon what exactly “where curable” and “curable environment” mean, but broadly, Justice Mathew’s point is that equality of opportunity makes no sense without taking into account the structural conditions into which people are born and grow up in, and which define, limit or otherwise profoundly affect the formulation and achievement of their goals.

Justice Mathew then reconciled his argument that Article 16(1) was about individual equality with upholding group classification: any classification, he argued, must group together individuals sharing certain characteristics. Categorization into Scheduled Castes/Tribes, then, was no more than a convenient method for identifying individuals who did suffer from those structural conditions that required affirmative action. (Paragraph 108) Crucially, for Justice Mathew, Article 16(1) is not about group equality. It is about effective individual equality, which is to be achieved by using the method of group-identification. The obvious problem with this is the two-pronged evil of over-inclusiveness and under-inclusiveness. As soon as you abandon separate analyses of each individual case for group-categorization, it is inevitable that certain members of your chosen group will not be “backward” (over-inclusiveness), and certain members who do not fall within the group will be “backward” (under-inclusiveness). This makes the definition of the “group” absolutely critical, and as we shall see in subsequent cases, it is this issue that would emerge as a political and constitutional battleground. In N.M. Thomas, however, Justice Matthew was oddly unconcerned with this very real problem.

Justice Krishna Iyer, in his concurring opinion reiterated the shared vision of equality embodied in Articles 16(1) and 16(4), and added a further ground for the argument by invoking Article 46, of the Directive Principles. While the Directive Principles are, of course, unenforceable, Justice Iyer argued that the interpretation of Articles 16(1) and 16(2) must be such that gives effect to Article 46. In this way, by invoking the Directive Principles as background, structuring values, helping us select which conception of equality Article 16 embodied (colour-blind, or group-subordination, or something else), Justice Iyer provided the constitutional grounding to Justice Mathew’s philosophical vision of equality.

Justice Fazl Ali, in his concurrence, moved away from Justice Mathew’s individual-centric notion in categoric terms. While adopting the same philosophy of equal opportunity, he held:

“Equality of opportunity would naturally mean a fair opportunity not only to one section or the other but to all sections by removing the handicaps if a particular section of the society suffers from the same.” (Paragraph 193)

This, as we have seen, is the classic statement of the group-subordination theory of equality. He then adopted Justice Iyer’s argument about the interpretation of Articles 14 and 16 being determined by the Directive Principles (Paragraph 200). Note, however, that Article 46 categorically refers to weaker “sections” of society. If, therefore, the interpretation of Article 16(1) is grounded in Article 46, then it cannot but be taken to be embodying the principle of group-subordination.

The last concurring judgment was Justice Beg’s, but he only concurred in the judgment, siding with the majority on the ground that the present case was covered by Article 16(4). On the reasoning, he disagreed, sticking to the past interpretation of 16(4) remaining an exception to 16(1).

Justices Gupta and Khanna dissented. Justice Gupta held that Scheduled Castes were, indeed, “castes” within the meaning of 16(2); and that any event, the manner of distinction was not one that was permitted by 16(1). In a detailed dissent, Justice Khanna reaffirmed that 16(4) was an exception, and the legitimating ground for reservations for SCs and STs was found in, and limited to, 16(4). In particular, he held:

“There is no scope for spelling out such preferential treatment from the language of Clause (1) of Article 16 because the language of that clause does not warrant any preference to any citizen against another citizen.” (Paragraph 57)

Which, as we can see, if framed in explicitly individualistic language. Hammering the point home, he observed in the very next paragraph:

Equality of opportunity in matters of promotion must mean equality between members of the same class of employees and not equality between members of separate, independent classes.” (Paragraph 58)

In other words, an express rejection of the group subordination principle. And to make it even clearer, he went on to hold:

What Clause (1) of Article 16 ensures is equality of opportunity for all citizens as individuals in matters relating to employment or appointment to any office under the State.” (Paragraph 60)

N.M. Thomas, thus, leaves us with a deeply divided Court, one in which each of the seven judges wrote his own opinion, and came to deeply divergent conclusions on a number of issues. Let us try to sum up:

(1) A majority of five judges concurred in the judgment, and two dissented

(2) A majority of four judges held that 16(4) is not an exception, but an “emphatic restatement” of 16(1) – and thus, 16(1) itself permits reservations and preferential treatment

(3) Two judges – Fazl Ali and Krishna Iyer JJ – by invoking Article 46, specifically adopted the group-subordination principle of equality in their interpretation of Article 16(1)

(4) Four judges (or five, depending on how one interprets Ray CJ’s judgment, which seems to adopt both principles) – that is, Mathew and Beg JJ.’s majority opinions, and Khanna and Gupta JJ.’s dissents – while coming to different conclusions, nonetheless maintained the individual-centric view of Article 16(1)

In essence, therefore, while N.M. Thomas changes the idea of of equality under 16(1), a bare majority upholds the deep philosophy of individual equality that forms the ultimate philosophical basis of it. This, essentially, makes Justice Mathew’s opinion controlling. The point of the Article 16 scheme – after Thomas, therefore – is not to achieve group equality qua groups, but to achieve individual equality, and the use of groups is a convenient mechanism to achieve the end goal of individual equality. This, indeed, comports well with the founders’ vision (as we have seen in previous posts) of an end-goal of a society in which class and other such markers become entirely irrelevant. What Justice Mathew understands is that to achieve a colour-blind society, you might need to take colour into account on the way, in order to ameliorate the continuing negative effects of structural inequalities; the arguments are not novel. Feminists have regularly argued that the end-goal of a society in which gender is irrelevant can only be achieved by taking gender into account on the way, and bringing women to a position where formally making gender irrelevant really means actually making gender irrelevant; and the same with race. What is crucial to note is that none of these arguments lose their ultimate goal of emancipating the individual. Classification is a means to an end, not an end in itself.

The distinction is important, and not just simply to understand what kind of equality our Constitution commits us to, and whether it is a vision that we find inspiring and worth believing in. It is also important because – as we shall see – debates about identifying beneficiary groups (the Mandal Commission and beyond), the bitter fight over the “creamy layer”, and ultimately, how far we are willing to go with reservations (for instance, over-inclusiveness and under-inclusiveness aren’t issues at all if your goal is objective is group equality) – depend upon whether our goal is to make groups equal to each other (and thus, reinforce group identity), or to make individuals equal to each other (and thus, ultimately, dissolve group identities). After N.M. Thomas, the latter view had an edge. In subsequent posts, we shall examine how it would fare in the fraught and divided coming years.

Reservations, Equality and the Constitution – II: The Early Cases

In our first post on reservations, we argued that the text and structure of the Constitution suggests that the vision of equality embodied in Articles 14, 15 and 16 is colour-blind: that is, it views as a violation of equality any governmental attempt to classify individuals on the basis of certain prohibited categories, like caste, race, sex etc., even if such classification is ostensibly for a benign purpose, like remedying past inequalities. It is important to understand that formal equality and colour-blindness are different things. Formal equality merely calls for treating like cases alike, on pre-determined criteria of likeness. Colour-blindness, on the other hand, is a deep substantive vision of the ideal society, motivated by a strongly individualistic bent that insists on considering individuals only as individuals, when they are subject to any government-mandated distribution of benefits and burdens.

The Supreme Court’s early cases emphatically affirm the colour-blind view. State of Madras v. Champakam Dorairajan is the first, decided by a 7-Judge bench in 1951. The Madras government’s policy, which stipulated admission to medical and engineering colleges in a proportion, based upon caste and religion, was challenged, inter alia, under Article 15(1) [recall that 15(4) did not exist at the time]. The State made an argument that reservations in educational institutions were justified under Article 46, part of the Directive Principles of State policy, which required the State to “promote with special care the educational and economic interests of the weaker sections of the people.” The Court rejected the argument on two grounds: first, obviously, that the Directive Principles were not enforceable.  Secondly, however, it argued that if reservations could be justified under Article 46, this would make 16(4) redundant. It logically follows, then, that Article 16(4), which allows for reservations under the broader Article 16 scheme of equality of opportunity, is an exception to Article 16(1), and that 16(1) itself does not contemplate reservations in its guarantee of the equality of opportunity – because if it did, finding another source for the government’s reservation-making power, in Article 46, would not make Article 16(4) redundant. In other words, the redundancy argument works only if we assume that Article 16(4) is the source of the government’s power to make reservations, and from that it follows that Article 16(1) cannot be. This, precisely, is the ideal of colour-blindness that we discussed in the last post: even for remedial purposes, equality under the colour-blind theory does not permit classification on prohibited bases, and such classification can be justified only by carving out a specific constitutional exception (16(4)). Since Article 15 had no parallel 15(4), the Court struck down the Madras Government’s policy.

Directly in response to the Supreme Court’s judgment, the Constitution was amended to insert Article 15(4). Article 15(4) reads:

Nothing in this article… shall prevent the State from making any special provision for the advancement of any socially and educationally backward classes of citizens or for the Scheduled Castes and the Scheduled Tribes.

Apart from a few differences that are not relevant at this stage, this is framed much like Article 16(4). And much like Article 16(4), the language – and in this case, the specific history, Article 15(4) suggests that it is an exception to the broad equality provision under 15(1), to allow the government to do something it could not otherwise do – classify individuals on the basis of sex, race, caste etc., for certain specific purposes (the advancement of socially and educationally backward classes, i.e., remedial affirmative action).

The logic of Champakam Dorairajan was at work in the next Supreme Court case on point, The General Manager, Southern Railway v. Rangachari. Among the many questions at issue in Rangachari, the question also was whether the Article 16 scheme allowed for reservation in promotions. Although the Court held that it did, it also observed that Article 16(4) – restricted as it was to “appointments“, did not cover the full spectrum of employment possibilities that were covered by Articles 16(1) and (2). In particular, the Court held that “in matters relating to salary, increment, gratuity, pension and age of superannuation there can be no exception even in favour of backward classes.” (Paragraph 22) We have already explained above why this approach is consistent with the colour-blind vision.

Specifically, the practical effects of adopting a colour-blind interpretation of Articles 15 and 16, and treating 15(4) and 16(4) as exceptions, were felt in M.R. Balaji v. State of Mysore, a Constitution Bench decision from 1963. In that case, the State of Mysore , fixed 68% reservation for “socially and educationally backward classes” as required by Art. 15(4). These backward classes were identified on the basis of caste. The policy was struck down by the Court on two grounds, both of which underscore the character of Article 15(4) and its relationship with 15(1). First, the Court rejected the sole use of caste as a determinant of social and educational backwardness, on three grounds: first, that class does not equate to caste, and an identification based on caste to choose a class may not always be logical; second, it may result in perpetuating the vice of caste, which defeats the entire purpose of the provision; and third, there exist communities such as the Muslims and Christians, which do not recognize the caste system. The second reason is particularly crucial, because it assumes that the purpose of the Article 15 scheme is to achieve a society in which caste is legally and sociologically irrelevant (this is consonant with the drafting history, and Ambedkar’s idea that the reservation provisions would sunset in ten years). That, precisely, is a colour-blind society. There is, of course, a distinction between the ideal of a colour-blind society, and a colour-blind Constitution. The latter would not permit any classification unless specifically authorised by a provision such as 15(4) or 16(4), whereas the latter would, conceivably, permit classification in a narrow area where such classification was a necessary step on the road to eventual colour-blindness (a number of feminist theorists, for instance, argue that while their end-goal is to make sex irrelevant, that goal can only be achieved by present-day remedial action based on sex, to bring women up to a point where colour-blind legislative programs do not handicap them because of underlying structural inequalities).

Colour-blindness was also at play in the Court’s rejection of the 68% quantum. Since Article 15(4) was an exception, the Court held, the quantum of reservations permissible under it could not exceed 50% – because, logically then, the exception would swallow up the rule, the rule here being the colour-blind equality code under Article 15(1).

The logic in Balaji led the Court to hold unconstitutional the “carry forward” rule in its next case, T. Devadasan v. Union of India. The carry-forward rule holds that unfilled vacancies that have been set apart for reserved-category candidates are to be “carried forward” to the next year in addition to the already existing reservations, thus increasing the quantum or reservations in the succeeding year by the amount unfulfilled presently (to a limit of two years, in the case at hand). Following Balaji, the Court held that insofar as, because of the carry-forward rule, the reserve vacancies in any one year grew to exceed 50%, the rule was unconstitutional. In addition, the Court framed Article 16(1) in explicitly individualistic terms:

“… the guarantee contained in Art. 16(1) is for ensuring equality of opportunity for all citizens relating to employment, and to appointment to any office under the State… the guarantee is to each individual citizen and, therefore, every citizen who is seeking employment or appointment to an office under the state is entitled to be afforded an opportunity for seeking such employment or appointment whenever it is intended to be filled.” (Paragraph 22)

A closer analysis reveals that the carry-forward rule itself is implicitly framed in group-communitarian terms. The goal of carry-forward is to ensure that at a given time t, the work-force has a certain specified number of persons belonging to a particular community. The procedure of selection, and the classifications involved in it, are designed to achieve the end-goal of adequate group representation. As such, therefore, the Court’s striking it down in Devadasan was both unsurprising and doctrinally consistent.

Justice Subba Rao dissented. And his dissent was not just to the narrow holding of unconstitutionality in Devadasan, but a radical challenge to the entire conception of equality that we have been discussing thus far, and that the Court had (thus far) adopted as given over the course of fifteen years. Justice Subba Rao argued:

“Article 16 is an instance of the application of the general rule with special reference to opportunity of appointments under the State. It says that there shall be equality of opportunity for all citizens in matters relating to employment or appointment to any office under the State. If it stood alone all the backward communities would go to the wall in a society of uneven basic social structure; the said rule of equality would remain only an utopian conception unless a practical content was given to it… that is why the makers of the Constitution introduced clause (4) in Art. 16. The expression “nothing in this article” is a legislative device to express its intention in a most emphatic way that the power conferred thereunder is not limited in an what by the main provision but falls outside it. It has not really carved out an exception but has preserved a power untrammelled by the other provisions of the Article.” (Para 35)

Each of the underlined portions reveal the depth of Justice Subba Rao’s challenge to the dominant vision. In making 16 the specific application of a general rule of equality, he links 14, 16(1) and 16(4) – that is, reservations, equality of opportunity, and equality more broadly, are no longer at odds, but part of an overall coherent scheme. Thus, 16(4) transitions from being an exception to the rule in 16(1) and (2) to being an “emphatic” restatement of it. And thus, logically, the legitimising constitutional source of affirmative action is no longer 16(4), but 16(1). Or, in other words, the very concept of equality that the Constitution is committed to includes within it the guarantee of affirmative action. The shift is completed in Justice Subba Rao’s reference to the “uneven basic social structure“, which has rendered certain “communities” backwards. This is the classic statement of the group-subordination principle of equality that we discussed in the last post: status quo is structurally constructed to disadvantage communities, and insofar as individuals are part of those communities, they suffer structural discrimination. Remedial action aimed at communities, therefore, is not an exception to the general rule of equality, but serves to fulfill equality’s deep purposes.

Justice Subba Rao’s dissent had no immediate impact. In C.A. Rajendran v. Union of India, the Court continued on the straight and narrow path laid before it. We need not go into the complex facts of the case here; suffice it to say, broadly, that it was the absence of reservations that was challenged in C.A. Rajendran, on the ground that, far from being an exception to 16(1), 16(4) was itself a constitutional guarantee and a fundamental right. The Court roundly rejected this contention, holding that 16(4) was merely an enabling provision. In so doing, the Court mentioned the divergence between the majority and dissent in Devadasan, but refused to definitively reject Justice Subba Rao’s views. Implicitly, however, it certainly did so, because if 16(4) was, like Justice Subba Rao said, an emphatic restatement of 16(1), and if 16(1) included within its conception of equality the ideal of affirmative action, then reservations would indeed be a matter of right, because 16(1) makes it clear that “there shall be equality of opportunity.” This, unambiguously, is the language of right.

Thus, in the first twenty years of the Court’s equality-and-reservations jurisprudence, with the exception of a single lone dissent from Justice Subba Rao, a series of (large-ish) benches affirmed a colour-blind vision of equality in the 14-15-16 code. Justice Subba Rao’s opinion, however, was soon to play a transformative role in the Court’s jurisprudence, and it is that we will turn to in the next post.



Indian Medical Association v UoI’s structural reading of Article 15: Did the Court go too far?

Readers will recall that we have defended two readings of Article 15 that would militate against a textual-originalist approach to the Constitution: the Delhi High Court’s reading of “sexual orientation” into “sex” (here), and a hypothetical problem of inter-sectionality (here). We justified both readings on two grounds: first, a structural reading of the Article 15 in the broader context of Part III reveals that the Constitution subscribes to something we called the non-discrimination principle, i.e., “government cannot discriminate on the basis of essential aspects of any individual’s private and public identity (i.e., the series of labels used by the society to identify her and differentiate her from others) that she is born into and is powerless to choose or change.” Both sexual orientation and inter-sectional positions fall within this category. And secondly, both readings are grounded in an intelligible interpretation of the text of Article 15.

In our previous posts, we focused primarily on the non-discrimination principle, but here I would like to pause briefly at the second argument. As we have observed before, it is crucial to remember that structure flows from text. It is not – as Mathew J. and Justice Holmes both observed in related contexts – a brooding omnipresence in the sky; it is firmly tethered to text, to history, to constitutional and political tradition, and to precedent. The non-discrimination principle is derived from a close reading of the specific categories listed in Articles 15, 16, 17 et al, raised to a level of abstraction that unites all of them without doing violence to any, and possesses the scope for bringing in analogous categories – that, again – it must at least be possible to support through text. A contrary position runs the grave risk of opening the floodgates to judicial legerdemain, where judges infuse into the constitutional text values that are their own, and not the Constitution’s.

Keeping this in mind, let us examine the Supreme Court’s opinion in its recent 2011 case, Indian Medical Association v. Union of India. Running into 160 pages of the judicial reports, IMA v. UoI is a long and complex case, dealing with a host of issues ranging from affirmative action, minority rights, the proper interpretation of Article 19(1)(g), the compatibility of Article 15(5) with the basic structure, and the role of the directive principles (and we shall have occasion to discuss this case at length when we discuss affirmative action). For the moment, let me focus, however, on one particular holding of the Court, that can be found in paragraphs 112 – 114. Briefly, the Court held that a private, non-minority higher educational institution that admits students only on the basis of their scores in an entrance test is in violation of Article 15(2). Here’s how, in three steps:

Article 15(2) states, in relevant part:

No citizen shall, on grounds only of religion, race, caste, sex, place of birth or any of them, be subject to any disability, liability, restriction or condition with regard to… access to shops, public restaurants, hotels and palaces of public entertainment...”

Step One: Referring to certain comments by Ambedkar in the Constituent Assembly Debates, the Court read the word “shop” in a manner broad enough to include educational institutions (paying for and receiving a service). This brought educational institutions within the ambit of 15(2). For the purposes of argument, let us accept this as correct.

Step Two: The Court observed that entrance tests operate so as to discriminate against students who hailed from socially and educationally backward backgrounds. For the purposes of argument, let us accept this as correct as well.

Step Three: The Court then noted:

“There are two potential interpretations of the use of the word ‘only’ in clause (2) of Article 15. One could be an interpretation that suggests that the particular private establishment not discriminate on the basis of enumerated grounds and not be worried about the consequences. Another interpretation could be that the private establishment not just refrain from the particular form of overt discrimination but also ensure that the consequences of rules of access to such private establishments do not contribute to the perpetration of the unwarranted social disadvantages associated with the functioning of the social, cultural and economic order.

Et voila! What Articles 15(1) and (2) prohibited was not just invidious discrimination on enumerated grounds, but also the consequences of such discrimination (in large part, social and educational backwardness); thus, any action that contributed to maintaining such social and educational backwardness fell foul of the Article 15 prohibitions.

Now, it is submitted that the word “only” can mean many things (and indeed, we plead guilty to having played fast and loose with the phrase “any of them” earlier on this blog); but in no circumstances, twist, stretch, bend or contort the word as you may, can it bear the stress the Court has put on it here, in the context of the sentence in which it has been used. The Merriam-Webster dictionary, for instance (and I invoke a dictionary with all due caution) lists four possible uses of the word “only”, the first two of which are applicable but do not support the Court’s conclusion, and the last two are not germane to the context.

Yet be that as it may, there is a more serious objection to be made. Let us agree that the object of Article 15 is to prevent the further entrenchment of exiting inequalities by prohibiting forms of discrimination. But if that was all that was at stake, the Constitution could simply have said that. Imagine a hypothetical Article 15: “all forms of discrimination that serve to perpetuate existing inequalities and entrench social and educational backwardness are hereby prohibited.” But that isn’t what the text says, and that isn’t all there is to it, because the Constitution also specifies the manner in which that object is to be achieved by laying out specific, proscribed categories of non-discrimination: caste, religion, place of birth, sex etc. In other words, the Constitution specifies the goal through the path.

And nor is this devoid of logic. As we have discussed before, what unites the categories that we find in 15, 17, 18 et al – caste, sex, untouchability and so on – is not only are they forms of private identity that one is born with and is more or less powerless to change, but that they are an essential part of one’s public identity – that is, the labels that our society imposes to classify and categorise its subjects. It is these societal labels, by virtue of being societal labels, that have been historically responsible for the worst and most invidious form of discrimination. And we can now see why the Court’s incorporation of social and educational backwardness into Article 15 is structurally flawed – not only is it controversial whether social and educational backwardness fits within the “born into and powerless to change” category, but more importantly, it is at best a fluid and amorphous set that in no way tracks the societally-imposed clarity and rigidity of sex or caste or race.

Now we may not agree with this logic. I do not myself; Article 15 misses a trick, I think, by failing to include “class”, a fundamental driving force of discrimination over the last two centuries; and it ignores, I believe, the arbitrary role of money in blocking access to goods and services. Recent scholarship has also cast doubt over the rigidity of categories such as race and sex. But that is besides the point: the story we are telling is the story of our Constitution, and not that of our individual convictions and values – and the most plausible story that takes into account text and structure is one that the Court, it is submitted with respect, does not keep faith with.

As a final structural point, we may also note that where the Constitution wishes to provide for socially and educationally backward classes, it has done so expressly (Articles 15(4), 15(5), 16(4)). Indeed, the First Amendment inserted Article 15(4) into the Constitution precisely in order to overcome the Champakam Dorairajan decision, where the Court struck down legislation on grounds of a 15(1) violation, rejecting an argument (then based on the DPSPs) that it promoted the interests of socially and educationally backward classes. By now reading that very same category into 15(2) (and therefore, by extension, 15(1)), the Court repudiates both its own past holding, as well as the legislature’s constitutional response to it, and thus makes a break with history that it fails to justify.

We part ways, therefore, with the Court, on its interpretation of Article 15(2); nonetheless, it is gratifying to note that the Court engaged in explicitly structural analysis (saying so directly at one point), which we have defended before as the best method of constitutional exegesis. There will, of course, be disagreements over outcome; but it is far more important that the Court identifies and follows a distinct, defensible and attractive analytical method when it adjudicates cases of a constitutional nature. Hopefully, in that respect, IMA v. UoI will serve as precedent to build upon.