top of page
Image by Annie Spratt




The Fault in Our Class: A Caution on Constitutional Attitudes Towards Economic Weakness

- Lalit Panda[1]

For many critics of the 103rd Amendment to the Constitution, one key premise for their opposition is that there is no place under our constitutional scheme for reservations solely on the basis of economic criteria (for example, see here, here and in further discussion below). Indeed, the apparent political motivations behind the Amendment and the possibility that it heralds an assault on the traditional basis of reservations for subordinated groups, give us adequate reasons to mistrust it and put it to rigorous legal and evidential scrutiny. In this article, however, I argue that many of the arguments challenging the idea of economic reservations actually have implications that might be unpalatable even to those making such arguments, particularly if they value a robust and inclusive conception of substantive equality.

In what follows, I first sketch out, for purposes of visualisation, an ideal beneficiary of reservations for economically weaker sections (EWS) under the Amendment. Next, I evaluate two forms of arguments that have been made against reservations for this ideal beneficiary, and find that these arguments rely on formalism and efficiency in strikingly objectionable ways that run counter to the spirit of substantive equality. I then argue that the challenge on the basis of caste discrimination was correctly identified by the dissent in Janhit Abhiyan v. Union of India as the most offensive aspect of the Amendment. I suggest, however, that it could still have been upheld after placing strictures to ensure that its benefits extend exclusively to persons like the ideal beneficiary. For lack of space, I do not touch upon the challenge to the Amendment on the basis of the 50% ceiling, though I remark on how a relaxation of that ceiling for existing reservations (non-EWS reservations) might affect the caste discrimination challenge to EWS reservations.

A Sketch of the Ideal Beneficiary

Consider a person X who belongs to a social group that does not and cannot benefit from any existing form of reservations. X suffers from poverty that is so grave that the proportion of persons like X who are granted access to education and stable jobs is the same or even lower than some Scheduled Castes (SCs), Scheduled Tribes (STs) and Other Backward Classes (OBCs). This means that X belongs to a section of the unreserved category of citizens that is as statistically under-represented (in the classes of persons selected for educational institutions and public employment) as existing beneficiaries of reservations generally, or perhaps as under-represented as some specific class of existing beneficiaries (e.g., the best-represented, median-represented or worst-represented class amongst existing beneficiaries). I make no claim here as to what family income-level and asset ownership defines this section of the unreserved category (the current proposal seems inadequate), nor as to what proportion/percentage this section constitutes of the relevant population of citizens. Indeed, neither does the 103rd Amendment.

Unless critics claim that poverty cannot possibly limit access to education and jobs to an extent comparable to the limitations suffered by existing beneficiaries of reservations, I think it reasonable to believe that some section of the unreserved category as defined above does exist, though this section might be miniscule and might suffer from poverty of the most extreme form. Arguably, this question may not be relevant in Basic Structure review, and we need only ask whether such a section of the unreserved category can exist. This depends on whether one views Basic Structure review as requiring adjudication of constitutional principles only and not the socio-economic evidence underlying the principles. So then, where does X stand in our Constitution’s egalitarian scheme? What basic principle of equality would completely disentitle X from a constitutional guarantee to opportunities that they would otherwise not have? Possible answers to this are explored below.

Formalism and Efficiency

I will now consider two types of arguments that have been made to differentiate the form of disadvantage faced by the ideal beneficiary of economic reservations and argue that these arguments have unsavoury implications, particularly for earnest supporters of substantive equality.

a) Backwardness and Under-Representation

Perhaps the most prominent argument made against economic reservations, as recounted in Maheshwari J.’s opinion in Janhit Abhiyan, is that it is contrary to a conception of “social justice” dedicated to people at the bottom of a “social hierarchy” who are “socially oppressed” or are from groups suffering “social stigma”.[2] In this view, the 103rd Amendment is argued to empower privileged sections of society who are neither socially and educationally backward nor inadequately represented. These two criteria are important “guardrails” whose observance ensures that reservations remain respectful of the Basic Structure of the Constitution.[3] Appealing to discussions in the Constituent Assembly, lawyers for the petitioners stressed on the significance and care with which a specific form of backwardness was identified by the framers as the sole form of disadvantage worthy of reservations. For example, it has been argued that the “purpose” of reservations is to level the playing field for backward classes,[4] that they are “meant to” provide representation to under-represented and subordinated communities (and not individuals), and that they are provided in public employment “because the framers wanted the backward classes to share State power”.[5] What is more, economic criteria on its own has been consistently held by courts to not meet existing constitutional requirements and is argued to be “obviously unsustainable from the point of view of equality of opportunity”. These views are pithily summed up in a slogan: reservations are not a form of “poverty alleviation”.[6]

As arguments by certain lawyers in the case and the opinion of Maheshwari, J. illustrate, a first-order response is that under-representation and backwardness are conditions imposed under Articles 15(4) and (5) and 16(4), that past Supreme Court judgments on economic criteria apply only to those provisions, that the 103rd Amendment has removed the basis of those judgments, and that reservation cannot be “itself reserved” for its existing beneficiaries on the basis of the “existing text” of those provisions.[7] While this response answers the purely textual challenge to economic reservations, there is more it leaves unaddressed. As Jacobsohn points out in his article on the Janhit Abhiyan judgment on this website: “Constitutional identity will to a large extent reflect how the essentials of constitutionalism combine and interact with the attributes of a constitution that are expressive of unique histories and circumstances.” Similarly, Kalra argues that prior jurisprudence on classification in reservations law is relevant in navigating the EWS Amendment’s basic structure dilemma, “where a declaration of unconstitutionality risks stultification, but a simple approval leaves much in the Parliament’s domain”. An Amendment does not erase the jurisprudence evolved under the original text when aspects of that jurisprudence form part of a constitution’s unique unamendable identity.

Which aspects of traditional reservations law should be considered basic to our Constitution’s identity? Treating “backwardness” as an unamendable precondition for reservations would ironically produce a formalistic vision of substantive equality. Under this view, only certain kinds of unfair disadvantages are worthy of constitutional concern. Behind the repeated incantation that reservations are for “social” justice is a decidedly odd reluctance to understand economic justice as a form of social justice, as if purely economic forces cannot underpin powerful social structures of their own. Economic inequality can itself be unjust and can additionally produce unjust social hierarchies between economic classes and stereotypical attitudes about the poor. Appeals to a fixed “purpose” of reservations or what they are “meant” to do similarly privilege existing legal forms with little consideration of the underlying substantive principles that these forms serve. Surendranath recognises this (in making a different case) when he suggests that reservations can have different objectives depending on the site of their implementation (e.g., primary education, higher education, public employment, representative bodies etc.) and the beneficiaries (e.g., SCs/STs, OBCs, domicile groups, religious minorities, women, transgender people etc.)

The proper inquiry in relation with the EWS reservations is not whether it is a legitimate mode of poverty alleviation (this has served as a ‘straw man’ for critics) but rather whether it can provide direct access to education and jobs to those who are unfairly denied these opportunities due to poverty. Other forms of support are not substitutes for this access. Subsidies and scholarships don’t equate to college seats; we are concerned here with the failure to meet selection criteria and not the inability to pay fees. And economic reservations in public employment need not be about giving the poor a share in State power; it could instead provide employment that is more stable than what is provided under an employment guarantee scheme. Under this view, it might be necessary to limit economic reservations to non-executive, entry-level jobs (and not provide reservations in promotions). By “under-representation” we need not mean inadequate representation of interests or voices in positions of power and decision-making. There is also the question of statistical under-representation of a section of citizens as evidence of inequality in opportunities. The way reservations have been envisaged under the Constitution till today can never be a complete answer as to how they can permissibly be envisaged tomorrow. Exclusive reliance on the intention of the framers would turn the Basic Structure doctrine into a charter for originalism in our treatment of the amending power. Something more than an appeal to the past is needed.

b) Discrimination

We can search then for a principled distinction outside of the existing phraseology of reservations and instead within the broader principles of the right against discrimination. Akin to the clean distinction drawn by some petitioners between social and economic justice, is the clean distinction drawn by others between discrimination and “deprivation”.[8] Under this view, the application of reservations is limited to individuals who have suffered discrimination and those suffering from poverty “who were never subjected to any discrimination, whether historically or otherwise”[9] are disentitled to such protective measures.

Admittedly, it is still contentious to consider socioeconomic status as a ground of discrimination, but there are a variety of reasons why it deserves to be considered one or akin to one. A Report of the UN Special Rapporteur on extreme poverty and human rights makes the point well:

Discrimination is part of the daily experience of people in poverty. It restricts access to employment, education, housing or social services. It may result in certain social goods or programmes not reaching people in poverty owing to discriminatory treatment by officials, employers or landlords, or to the fear of maltreatment. It discourages people who experience poverty from applying for a job, or from claiming certain benefits: it is thus a major source of non-take-up of rights. Discrimination may also lead people in poverty to lower their aspirations regarding what they can achieve, either for themselves or for their children, leading to a reduced investment in education. It explains in part why people in poverty are disproportionately represented in the criminal justice system, as judges may be biased against them or base their sentencing on anti-poor stereotypes.

Scholars in India too have considered the possibility of “economic class” or “socio-economic disadvantage” being a ground of discrimination.[10] How is this possible when it hasn’t been treated as one traditionally? On this, it is worth noting that we do not necessarily have to recognise such a trait as a ground of discrimination deserving the full suite of protections against discrimination. There may be difficulties in doing so in many areas of governance that do not arise in relation with affirmative action.

A general statement often made regarding grounds of discrimination is that they are either immutable (unchangeable) traits or traits that represent a fundamental choice of some sort.[11] Traits like race cannot be changed and traits like religion should not be changed forcibly. But a closer look at what makes up the basis of this binary identification tool reveals that it is undergirded by the value of personal autonomy. Grounds of discrimination not only include traits that are strictly immutable but also traits that are effectively immutable, in the sense that we do not have effective or meaningful control over them.[12] Some things only change with considerable difficulty or great personal cost. This is comparable to the idea that it would be abhorrent to penalize persons for refusing to change fundamental choices regarding religion, marriage, pregnancy etc. And finally, some traits that attract discrimination, like certain forms of disease and disability, may be completely curable or may become curable in future. But this possibility does not reduce the gravity of the disadvantage faced before the cure is administered.[13]

Viewed in this light, we can immediately see the unfairness of suggesting that protective relief for poverty should be denied because it is a changeable status. It suggests that the poor have effective control over their poverty, but simply aren’t willing to climb out of it. Heavy emphasis on the “transience” of poverty[14] or insistence that the poor are not a homogenous class[15] is a step away from suggesting that the poor choose to be poor. Indeed, this was precisely the argument made by one counsel in suggesting that economic reservations could incentivise irresponsible wealth management, careless spending and gambling. By relying on the economic theory of “moral hazards” for rejecting economic reservations, the argument leaned heavily on a particularly hazardous form of efficiency-based reason: if the poor are simply given certain advantages that they would otherwise be unable to access because of their poverty, such direct grant of access can hamper the economic system of incentives that is meant to reward effort with profits and laxity with loss. Apart from this specific function of wealth, poverty is normatively irrelevant to the distribution of social goods.

Indeed, the broad effect of all of these arguments is to turn the beneficent gaze of equality away from the cruelties of capitalism and fixedly in the direction of the cruelties of casteism. In sum, these are arguments that make it easier to check the steady expansion of equality protections and exclude certain forms of suffering from the benefit of our full-blooded concern. This may not have been the intention of those making the arguments, but it is a foreseeable effect. This is simply the burden one bears in operating at the level of abstraction that Basic Structure review demands.

Targets versus Priorities

I have tried to outline above why providing reservations on the basis of economic criteria is justified under the principles of substantive equality, and why resistance against this idea compels us to rely on unpalatable forms of formalism and efficiency. We should be more cautious about these outcomes, especially as legal scholars and practitioners committed to the idea of a Basic Structure.

However, what if concern for the poor is merely an excuse for a government intent on disbursing direct benefits to its vote bank? This fear is fuelled by the fact that a measure aimed at the poor excludes the very classes that have higher proportions of poor persons (the existing beneficiaries of reservations). As the distinction drawn in making this exclusion is based in caste, the dissenting opinion in Janhit Abhiyan correctly identifies the caste discrimination challenge as the more serious charge against the 103rd Amendment.

At the outset, some of the arguments challenging the exclusion appear to be inconsistent with unobjectionable features of our existing reservations policies. For instance, it was argued that because reservations were created to “put an end to monopoly of certain classes”, it could not be used to create a “perpetual monopoly” for those very same classes,[16] and similarly that reservations in favour of forward caste persons is impermissible in itself.[17] Arguments like these seem to suggest that reservations are in principle impermissible for members of advantaged groups defined on the basis of non-discrimination ground (here, caste). But this is not the case, as we do appreciate the possibility of other axes of disadvantage (apart from caste), such as in the need for horizontal reservations that benefit forward caste women and forward caste persons with disabilities. It would be a fair objection to point out that the EWS reservations have not been made horizontal, but such an objection could easily be answered by sub-classifying SCs, STs and OBCs on the basis of economic criteria as well. EWS under the 103rd Amendment would then only be an economic sub-classification of the unreserved category.

As it happens, the majority opinions in Janhit Abhiyan justified the caste-based exclusion by relying on faulty and outmoded doctrinal tools. I will only touch on one of these tools here. The broad justification for the exclusion rested on the understanding that the existing beneficiaries already had reservations of their own and that including them in the EWS reservations would prevent us from effectively targeting a new class of beneficiaries.[18] This justification rests on reasoning that is rooted in the reasonable classification test, where the target group is defined in terms of a chosen goal. The goal of previous reservations was to address the disadvantages of SCs, STs and OBCs. The goal of the 103rd Amendment is to address the disadvantages of the poor amongst the previously unreserved. The reasonable classification test relies exclusively on the priorities implicit in the “object of the legislation” and, crucially, that object can itself be defined in exclusionary ways.[19] The object of EWS reservations can be defined in terms of the poverty within certain castes instead of poverty generally. In justifying this definition of the objective, the prior existence of reservations to address the multi-dimensional disadvantages of other castes is a relevant factor.

Reasonable classification is simply the wrong tool to apply, not least because prejudiced objectives can be disguised as legitimate ones and discrimination can occur in the effects of a law even if there is no hostility in the aim. What then would be the right tool? The primary problem with the EWS reservations, as I see it, is not that it fails to make a principled distinction between different kinds of disadvantages. I don’t think there is any principled distinction between purely economic and multi-dimensional social disadvantages. Unfair disadvantage is unfair disadvantage. Rather, the problem is that the distinction lies in the amounts of disadvantage. One counsel for the petitioners even said as much (“[I]ndeed forward class must have faced some discrimination, but the intensity of discrimination is not enough to justify reservation.”[20]) Imagine if the quantum of reservations provided to SCs, STs and OBCs is allowed to breach the 50% ceiling and brought to a much larger quantum of affirmative action that would constitute a full and effective legal remedy for the disadvantages of those classes. If this eventuality occurs, can we then think of reservations for the forward caste poor? Similarly, consider how a hypothetical India may come to exist in the future where the fissures of economic class are wider than the fissures of caste. Is such an India likely? I don’t know. Is it possible? Most certainly yes. And as long as it is possible, an objection to economic reservations is based in the specifics of our current socio-economic condition and not in any universal or even broadly generalisable principles.

The flaw in the 103rd Amendment then is that in selecting legitimate targets for protective discrimination, it articulates a wholly illegitimate scheme of priorities. Or, at least, it opens the possibility of such illegitimate prioritisation. If a poor forward caste individual suffers barriers to accessing education and jobs, but these barriers are much lower than those suffered by existing beneficiaries of reservations, then providing reservations to such an individual would demonstrate some rather misplaced priorities. The adequate redressal of disadvantages caused by caste is still a distant dream and while that resolution awaits, the welfare state must be very careful about what new battles it picks, else it may fritter away its limited capacity to fight inequality. The trouble, however, is that within the scheme of existing caste-based reservations, constitutional doctrine has only ever treated reservations as a discretionary power of the government, instead of a mandatory remedy for the violation of the right to equality. The entire process of prioritisation that goes into selecting one beneficiary class for reservations and rejecting the other (which may be as much or even more deserving) has thus been left to the executive. This means that a somewhat badly off class can gain the benefits of affirmative action even while the worst-off class languishes (or isn’t adequately provided for), and there is nothing that courts have had to say about this. This, in my view, is what has culminated in the 103rd Amendment.

So how does the failure to prioritise correctly translate into a question of constitutionality? Consider how we would assess the constitutionality of a hypothetical measure under which the proportion of reservations available to OBCs was expanded by taking up some proportion of the reservations available to SCs and STs. The measure would provide for more effective affirmative action for OBCs, but to the detriment of the affirmative action available to SCs and STs. What would our Constitution have to say about such a move? This hypothetical situation echoes the question of sub-classification in existing reservations, in relation with which it has been argued that treating the beneficiary groups of reservations as homogenous operates to exclude relatively marginalised sections within the beneficiary groups. The current scheme for EWS reservations explicitly excludes SCs, STs and OBCs from general category seats and jobs that they could otherwise have availed through open competition and the abovementioned hypothetical example on OBC reservations explicitly excludes SCs and STs from seats that were previously reserved for them. In contrast, the failure to sub-classify in reservations excludes marginalised sections within the beneficiary group in effect. But what all of these situations have in common is that they are all questions about the relative levels of disadvantage of different disadvantaged groups. It may well be that the effective redress of disadvantages faced by OBCs requires that we provide them with more reservations (i.e., they have a right to more reservations), but surely such provision should not be to the detriment of a group facing even greater disadvantages?

Now, whether we think that the rights to equal opportunity of different disadvantaged groups are in conflict and need to be balanced (by weighing and comparing the disadvantages) or whether we think that they are in harmony and just need to be defined appropriately (one group’s right ends where the other’s begins), I think any court engaging in such a balancing or definition exercise will find that the prioritisation of worse-off groups is very much an inevitable feature of our constitutional scheme, built into our understanding of discrimination and equality. If the Janhit Abhiyan bench had compared a) the discriminatory effect of excluding SCs, STs and OBCs from seats and jobs, with b) the enforcement of substantive equality through reservations for the unreserved poor, it could have discovered the grave possibility of a violation of the right to equality and the right against discrimination: a possibility that would be realised depending on the extent of economic weakness that the government eventually selects. Providing reservations for certain minimal levels of economic weakness would, in my view, violate the basic constitutional principle of prioritisation implicit in the right to equality. A less-discriminatory measure (reservations for extreme poverty) would be more likely to pass muster (even if targeted only at the unreserved category). Foreseeing the realisation of this possibility, the bench could have agreed on the need for some new “guardrails” analogous to the old ones.

I have already articulated what these conditions could be in different sections above. First, the degree of economic weakness selected should be of adequate intensity or gravity, demonstrably comparable to the disadvantages faced by existing reservations beneficiaries. Second, the proportion of reservations provided should not be ten percent by default but has to be shown to be appropriate to the proportion of population meeting the first condition. Third, only suitable public goods should be reserved for the EWS (employment reservations may, for example, be restricted to entry-level, non-executive jobs and not provided in promotions).

Let me be clear: the ideas presented above do not form part of India’s existing jurisprudence on reservations law or even our equality law. I have proposed them, however, because the alternatives are not acceptable routes for the future evolution of equality in India and some greater effort of imagination is needed.

[1] Lalit Panda is a Senior Resident Fellow in the constitutional law team of the Vidhi Centre for Legal Policy. Views are personal. The author would like to thank Dr. Bastian Steuwer for his insightful comments on an earlier draft. [2] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 7.1 and 9 (Maheshwari, J.) [3] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 8 (Maheshwari, J.) [4] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 7.5 and 8.2 (Maheshwari, J.) [5] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 9 (Maheshwari, J.) [6] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 15.3 and 16.1 (Maheshwari, J.) [7] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 73 and 74.1.2 (Maheshwari, J.) [8] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 10.3, 13, 15 and 18 (Maheshwari, J.) [9] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 7.2 and 14 (Maheshwari, J.) [10] Gautam Bhatia, The Transformative Constitution (HarperCollins 2019), p.71 [11] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 18 (Maheshwari, J.). One may also see this statement in one of the opinions in Navtej Singh Johar v. Union of India, (2018) 10 SCC 1, paras 638.1-639.4 (Malhotra, J.). [12] Tarunabh Khaitan, A Theory of Discrimination Law (Oxford University Press 2015), pp.56-60. [13] This is the broad thrust of Sunstein’s argument in his hypothetical of a race-altering drug. See Cass R. Sunstein, “The Anticaste Principle” 92(8) Michigan Law Review 2410, 2443 (1994). [14] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 7.6, 8.4, and 13 (Maheshwari, J.) [15] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 10.4 and 14 (Maheshwari, J.) [16] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 7.5 (Maheshwari, J.) [17] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 12 (Maheshwari, J.) [18] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, paras 78-83 (Maheshwari, J.), paras 20-21 (Trivedi, J.) [19] Miron v. Trudel, [1995] 2 SCR 418, 489 (Supreme Court of Canada) (McLachlin, J.) (“Having defined the functional values underlying the legislation in terms of the alleged discriminatory ground, it follows of necessity that the basis of the distinction is relevant to the legislative aim. This illustrates the aridity of relying on the formal test of logical relevance as proof of non-discrimination …”) [20] Janhit Abhiyan v. Union of India, W.P.(C) 55/2019, para 13 (Maheshwari, J.)


Recent Posts

See All




bottom of page