[This is a guest post by Kartik Kalra.]
This week, the Supreme Court delivered its judgement in Janhit Abhiyan v. Union of India, holding constitutional the 103rd (Amendment) Act, 2019, which provides for reservations in public employment and education for economically weaker sections (“EWS”) of citizens other than Scheduled Castes (“SC”), Tribes (“ST”) and Other Backward Classes (“OBC”). In adjudicating the constitutionality of the Amendment, it split the question into three parts: the constitutionality of the use of solely economic criteria to reserve seats under Articles 15 and 16; the constitutionality of the exclusion of SC, ST and OBC candidates from the category of EWS; and, the constitutionality of the breach of the 50% ceiling on reservations. This piece concerns the first question: the constitutionality of a solely economic basis to determine the eligible class for reservations.
The Court’s engagement with this issue becomes significant due to the prior, enduring consistency of its jurisprudence with respect to the necessity of a homogenous, precisely delineated class constituting the eligible recipient of reservations under Articles 15 and 16. Indra Sawhney v. Union of India explicitly held that “a backward class cannot be determined only and exclusively with reference to economic criterion” (¶799), and Janki Prasad Parimoo v. State of JK considered income levels to be incapable in constituting a homogenous class eligible for reservations (¶36). Dayaram Verma v. State of Gujarat struck down income-based reservations on the same grounds (¶29.3).
In this piece, I argue that the Court’s deviation from well-established doctrine on the necessity of a homogenous, precisely delineated class as the recipient of reservation benefits, while motivated by a desire to avoid constitutional stultification, ignores a host of creative ways out of the basic structure dilemma. I evaluate the reasoning used by the Court in circumventing the application of prior doctrine on Articles 15/16, and argue that Maheshwari J. succumbs to the basic structure dilemma while Pardiwala J. misrepresents Chitralekha v. Union of India in order to demonstrate its convergence with the 103rd Amendment. I also propose that the Court could have employed a constrained declaration of constitutionality like it did in M. Nagaraj v. Union of India, prescribing the mode of determination of the EWS class eligible for reservations based on its reasoning in Chitralekha. In addition, I also propose sub-classification of the EWS category within the Chitralekha criteria to be an additional weapon in the Court’s arsenal, one that it could have invoked without succumbing to the basic structure dilemma.
EWS and the Application of Doctrine under Article 16
In addressing the question of prior doctrine forbidding solely income-based reservations due to the absence of an identifiable class, Maheshwari J. proposes that Articles 15(5) and 16(5) create a new class altogether, one that isn’t subject to that doctrine in the first place:
72. …The said decisions cannot be read to mean that if any class or section other than those covered by Articles 15(4) and/or 15(5) and/or 16(4) is suffering from disadvantage only due to economic conditions, the State can never take affirmative action qua that class or section.
On the other hand, Pardiwala J. uses Chitralekha to hold that there is no real deviation from doctrine:
70. …Indra Sawhney (supra) holds that the Chitralekha (supra) propounded occupation-cum-means test can be a basis of social backwardness even for the purposes of Article 16(4). Article 15(6)(b) Explanation defining EWS could be said to be fully compliant with this norm.
Trivedi J. agrees with Maheshwari J., indicating her belief that EWS is a separate class altogether, one to which prior doctrine on Articles 15/16 has no application.
Given the exclusion of the classes under Articles 15/16(4), it is true that EWS will be a separate class. The question, however, is how this class will be identified, which strikes at a deeper question that Maheshwari J. assumes the answer to: the applicability of prior doctrine to assess reservations for any class under Article 16, which broadly mandates the existence of a homogenous class of recipients. In case this doctrine is inapplicable, then Articles 15(6) and 16(6) can function as a self-contained code with radically different philosophies and principles as compared to the rest of the equality code. If not, however, then a deviation from prior doctrine must be examined in greater depth.
State of Kerala v. NM Thomas approved Subba Rao J.’s dissent in T. Devadasan v. Union of India, holding that Article 16(4) is an emphatic restatement of the equality principle under Article 16(1) instead of being a deviation therefrom (¶78). All doctrine of Article 16 was imported to Article 15 in Ashoka Kumar Thakur v. Union of India (¶116), and Nagaraj held the entirety of Article 16 (and by implication, Article 15) to be restatements of the basic equality principle under Article 14 (¶102). There is, therefore, a degree of coherence across Articles 14, 15 and 16, and any tinkering must bear fidelity to what the equality code stands for and the means through which it is effectuated. While an author has previously argued here that dismantling caste-based structures of oppression and a reparative mode of reservations is a part of the basic structure, I do not take this burden. Even if one doesn’t subscribe to the reparative underpinnings of reservations (as none of the judges in the majority do), the mode of determining the classes eligible for reservations that has entrenched itself in Articles 15 and 16, and by implication in Article 14, cannot be ignored.
Case law on Articles 15 and 16, even including M.R. Balaji v. Union of India, has insisted on the necessity of a precise and similarly situated class that can receive reservations. Chitralekha held an occupation-cum-income based test to be accurate enough to determine backwardness (¶15); State of Andhra Pradesh v. P Sagar held the meaning of “class” under Article 16(4) to signify homogeneity, discounting the possibility of labelling a caste per se backward and necessitating the production of educational and economic criteria that would signify its backwardness (¶7); and T Murlidhar Rao v. State of Andhra Pradesh struck down the inclusion of the Muslim community in Andhra Pradesh within the OBC list because of its dissatisfaction with the accuracy of the report documenting their socio-economic status, which included both well-off and disadvantaged Muslims (¶73). More recently, Dayaram Verma reaffirmed the criteria in P. Sagar, holding that homogeneity is the sine qua non for determining the class eligible for reservations, and which cannot be achieved using economic criteria alone:
29.3 As the economically weaker sections among unreserved category cannot constitute as homogeneous group for the purpose of reservation and such reservation will not withstand to the scrutiny of twin test under Article 14 of the Constitution of India. Further, the economic criteria being fluctuating issue, the same cannot be the basis for any classification for the purpose of affirmative action for admission to educational institutions and while filling up the posts in the services under the State.
Case law, therefore, maintains that a class receiving reservations must be homogenous and precisely delineated. While Articles 15/16(4) use the term “backward” to define the eligibility of the class receiving reservations, Articles 15/16(5) use the term “economically weak”. For the former, the Court has repeatedly asserted the necessity of all constituents of the class being similarly backward. For the latter, therefore, the Court had the option of asserting the application of prior Article 16 doctrine necessitating a homogenous EWS class, one that is broadly similar in a set of indicators signifying economic capabilities. Does the EWS class possess a similarity in variables constituting economic weakness (other than income), such as, inter alia, occupational mobility, job security, blue v. white collar work, and the availability of health insurance? This is where the Court could have stepped in, mandating that the EWS class be determined in accordance with well-established doctrine that necessitates a homogenous class of recipients. This would not require the invocation of the basic structure test at all, for it would merely be an exegesis of the term “economically weak” under Articles 15/16(5) in the manner that Sawhney did for the term “backward” under Articles 15/16(4). The Court could have undertaken a similar exercise here, holding the Amendment constitutional while limiting its contours to a well-defined class.
Lastly, it must be noted that while cases like Sawhney, T. Murlidhar Rao and Dayaram Verma concerned classes under Articles 15/16(4), N.M. Thomas and Janki Prasad Parimoo concerned affirmative action under both Articles 16(1) and 16(4). Thus, the principle of class homogeneity is applicable to Articles 15/16 as a whole instead of being confined to Articles 15/16(4).
On this basis, I submit that the principle of identification of a homogenous, well-defined class is well-entrenched in the equality code, and any additions must operate within this principle. There are means of reaching this end without going down the basic structure path, other than the above discussion on an exegesis of the word “economically weak”. I explore this in the penultimate section.
Chitralekha, Article 15(6) and Blurring the Underpinnings of the Creamy Layer
As opposed to Maheshwari J., Pardiwala J. holds that reservations for the EWS category are also in consonance with the principles of identification under Article 16, and uses Chitralekha to further his contention. Chitralekha concerned the constitutionality of a reservation scheme that did not take caste as a factor affecting eligibility, instead evaluating income level and occupation type. The Court upheld the use of this criteria on the basis that the impugned scheme was effective in defining a class that was wholly backward and homogenous. Holding that the reservation scheme did not become unconstitutional solely because caste wasn’t accorded centrality, the Court said the following:
15. Two principles stand out prominently…namely, (i) the caste of a group of citizens may be a relevant circumstance in ascertaining their social backwardness; and (ii) though it is a relevant factor to determine the social backwardness of a class of citizens, it cannot be the sole or dominant test in that behalf.
Sawhney was also quite clear in its reading of Chitralekha, holding that as long as a wholly backward class was identified, insisting on the use of caste as a determining criterion was futile:
800. In Chitralekha this court held that such an identification [based on income level and occupation type] is permissible. We see no reason to differ with the said view inasmuch as this is but another method to find socially backward classes…[The] identification of backward classes can be done with reference to castes along with other occupational groups, communities and classes. We did not say that that is the only permissible method. Indeed, there may be some groups or classes in whose case caste may not be relevant to all. For example, agricultural labourers, rickshaw-pullers/drivers, street-hawkers etc. may well qualify for being designated as Backward Classes.
The overarching principle in Chitralekha, therefore, was that a backward class can be defined using non-caste criteria insofar as that other criteria are capable of constituting a homogenous class. It is true that backwardness may exist along axes other than caste, but this backwardness must constitute a precisely delineated homogenous group in the manner it did in Chitralekha. In Articles 15(6) and 16(6) and the Explanation, however, economic disadvantage is considered to be the sole assessor of backwardness. This, instead of being a derivative of the principles in Chitralekha, stands in direct contrast with them, for it constitutes no homogenous “economically weak” class as shown in the above section. The class is presently defined according to a notification issued by the Ministry of Social Justice and Empowerment, which uses solely income and property as class characteristics. This class can hardly be called homogenous, for it comprises of classes having differing occupational mobilities, job security, social safety etc.
These differences may be across lines of caste, religion, sex, income levels or any other factors, but it remains that the EWS category as defined presently does not constitute a homogenous class eligible for reservations.
Pardiwala J. disagrees. He makes the following further argument to establish the ability of purely economic criteria to constitute a homogenous class:
70. If economic advance can be accepted to negate certain social disadvantages for the OBCs [Creamy Layer concept] the converse would be equally relevant… Indra Sawhney (supra) holds that the Chitralekha (supra) propounded occupation-cum-means test can be a basis of social backwardness even for the purposes of Article 16(4). Article 15(6)(b) Explanation defining EWS could be said to be fully compliant with this norm.
If social disadvantage can be undermined by economic advancement, then what stops social advancement from being undermined by economic disadvantage? The purpose of the creamy layer, as is well-established in post-Sawhney doctrine, is to find a class that is truly backward after eliminating its constituents that show better socio-economic indicators. It must be noted that Sawhney itself considered economic advancement per se to be incapable of remedying social advantage, prescribing only stark forms of economic advantage that place one in a socially powerful position to sever them from the backward class they were hitherto a part of:
792. In a backward class under clause (4) of Article 16, if the connecting link is the social backwardness, it should broadly be the same in a given class. If some of the members are far too advanced socially (which in the context, necessarily means economically and, may also mean educationally) the connecting thread between them and the remaining class snaps. They would be misfits in the class. After excluding them alone, would the class be a compact class. The basis of exclusion should not merely be economic, unless, of course, the economic advancement is so high that it necessarily means social advancement.
Situation may, however, be different, if he rises so high economically as to become — say a factory owner himself. In such a situation, his social status also rises. He himself would be in a position to provide employment to others. In such a case, his income is merely a measure of his social status.
Further:
At the same time, it must be recognised that there are certain positions, the occupants of which can be treated as socially advanced without any further enquiry. For example, if a member of a designated backward class becomes a member of IAS or IPS or any other All India Service, his status is society (social status) rises; he is no longer socially disadvantaged. His children get full opportunity to realise their potential. They are in no way handicapped in the race of life. His salary is also such that he is above want.
It is, therefore, economic criteria strictly corresponding to social advantage – and not economic advancement per se – that snaps one’s link with the backward class and makes one a part of a “creamy layer”. This principle has been reaffirmed at many occasions, most prominently in Nagaraj and Ashoka Kumar Thakur. In an Article 15/16(5)-equivalent of this reasoning, the economic disadvantage being faced by a group must be so severe as to snap its link with the socially advanced class it was hitherto a part of. The continuation of this argument to reach its logical conclusion was another opportunity where scope for judicial intervention lay without invoking the basic structure test. The Court could have linked economic disadvantage and social advancement, prescribing a mode of determination that necessitates economic disadvantage so severe that social advancement is rendered meaningless. This would, perhaps, do true justice to the principle being derived from Chitralkeha, for all examples of non-caste classes that Sawhney derives using it are situations where economic disadvantage has the potential of undermining social advancement: “agricultural labourers, rickshaw-pullers/drivers, street-hawkers etc.” (¶800).
Alternatives to the Basic Structure Dilemma
The Court faced an obvious discomfort in striking down the 103rd Amendment, given its concern with possible stultification in holding caste-based reservations to be the only possible constitutional project and forbidding any other form of reservations under this Constitution [¶74.2.1 (Maheshwari), ¶98 (Pardiwala), ¶6 (Trivedi)]. While sufficient evidence is available to show that caste is presently the primary axis of disadvantage, it might not always be. This determination of changing axes of disadvantage could be reasonably considered outside judicial review, and the Court may choose to give deference to reservation policies addressing new axes of disadvantage. Simultaneously, it would desire to accord some significance to well-established doctrine in the equality code, requiring that additions thereto bear some familiarity to the basic principles on determining backwardness. This is the dilemma of the basic structure, where a declaration of unconstitutionality risks stultification, but a simple approval leaves much in the Parliament’s domain.
The same dilemma was faced by the Court in M. Nagaraj v. Union of India. Sawhney held that reservations can only be provided in matters of appointments, but Parliament added Article 16(4A) to also provide them in matters of promotions. Union of India v. Virpal Singh Chauhan propounded a catch-up rule for reservations in promotions, but Parliament further amended the Constitution to replace the catch-up rule with that of consequential seniority. When faced with a constitutional challenge to this set of amendments, the Court approved them, but with specific conditions on the use of this approval:
122. We reiterate that the ceiling limit of 50%, the concept of creamy layer and the compelling reasons, namely, backwardness, inadequacy of representation and overall administrative efficiency are all constitutional requirements without which the structure of equality of opportunity in Article 16 would collapse.
Given the extremely high threshold of the basic structure test and the corresponding judicial deference it introduces, the Court could have used the path in Nagaraj to uphold the amendments while also prescribing a set of constraints within which it must be exercised. As I have shown above, a class receiving reservations must constitute a discernible homogenous class. The Court could have recognized this, constraining the use of Articles 15/16(5) to reserve seats only for those economically weaker classes that satisfy the occupation-cum-income test in Chitralekha, necessitating the existence of a precise, homogenous economically weak class. Alternatively, the Court could have escaped the dilemma by effectuating the reasoning used for the creamy layer, mandating the existence of such a class whose economic disadvantage is so severe that it snaps its link with the socially advanced class it was hitherto a part of. Nagaraj paved the way to navigate the basic structure dilemma, and this case presented a ripe opportunity to reiterate it.
Until now, I have shown three possible ways out of the dilemma: an exegesis of the term “economically weak” under Articles 15/16(6), using the framework in Nagaraj to confine benefits only for classes that meet Chitralekha’s criteria, and necessitating the existence of a class so economically disadvantaged that its social advancement is rendered meaningless.
There is, however, one more direction in which the Court could have gone: mandating (or at least encouraging) the use of the principle of sub-categorisation within the EWS class. While the principle of sub-categorisation was approved in Sawhney for the OBCs, its application to the SCs is under contestation due to conflicting judgements in E.V. Chinnaiah v. State of Andhra Pradesh and State of Punjab v. Davinder Singh, both of which navigate the question of relative homogeneity of the SC group differently. Jeevan Reddy J. explained the underpinnings of sub-categorisation in Sawhney as follows:
802. To give an illustration, take two occupational groups viz., goldsmiths and vaddes (traditional stonecutters in Andhra Pradesh) both included within Other Backward Classes. None can deny that goldsmiths are far less backward than vaddes. If both of them are grouped together and reservation provided, the inevitable result would be that goldsmiths would take away all the reserved posts leaving none for vaddes. In such a situation, a State may think it advisable to make a categorisation even among other backward classes so as to ensure that the more backward among the backward classes obtain the benefits intended for them.
The OBCs, according to the judgement in Sawhney, may constitute a class whose homogeneity is under doubt. There could be some components that are better off, while some facing significant disadvantage. This applies equally well to the EWS class, for it includes persons with vastly diverging socioeconomic variables. The Court could have considered this another mode of constraining the use of EWS reservations, concentrating its use towards the worst-off sections of the EWS class. Simultaneously, this would also effectuate reservations for those persons whose economic disadvantage is so severe that its link with their social advancement has snapped.
Conclusion
As I have demonstrated in this piece, the majority judgement avoids meaningful engagement with the question of the application of prior doctrine under Article 16 that necessitates the existence of a homogenous recipient class. Maheshwari J. denies the application of past doctrine, and Pardiwala J. misleadingly conflates EWS reservations with the doctrine laid down in Chitralekha. While a large segment of the Court’s unsatisfactory reasoning has direct roots in the requirements posed by a basic structure challenge, this case presented a ripe opportunity for a creative solution accommodative of prior doctrine on reservations. The Court did not recognize the possibility of escaping this dilemma, which could have been achieved in at least four ways that I have demonstrated in this piece. Ultimately, it succumbed to the deference introduced by the basic structure test, validating a hitherto alien axis of assessing disadvantage. This doesn’t mean that all is lost: the criteria used to delineate the EWS class is always subject to judicial review even when the 103rd Amendment is held constitutional. The Court may use these principles then, pulling EWS reservations closer to the philosophy of the equality code.