, ,

On September 26th, a Constitution Bench of the Supreme Court delivered judgment in Jarnail Singh v Lacchmi Narain GuptaThe Court was essentially called upon to decide whether the correctness of the previous five-judge bench judgment in M. Nagaraj v Union of India ought to be referred to a seven-judge bench for reconsideration.

Nagaraj was a reservations case (examining a challenge to Articles 16(4A) and (4B) of the Constitution. In particular, two of its findings were under challenge: firstNagaraj had held that as a precursor to granting reservations (including to Scheduled Castes and Scheduled Tribes), the State would have to collect “quantifiable data” demonstrating their backwardness. And secondlyNagaraj had held that the “creamy layer” concept – where certain members of a group were deemed to belong to the “creamy layer” and therefore not entitled to reservations – was also applicable to Scheduled Castes and Scheduled Tribes. The Attorney-General for India argued that both these findings were incorrect, as they were contrary to the holding of the nine-judge bench in Indira Sawhney v Union of India.

Quantifiable Data 

On the first issue, the Constitution Bench – in a unanimous judgment authored by Nariman – held that Nagaraj stood directly contrary to Indra Sawhney, and was therefore incorrect:

Insofar as the State having to show quantifiable data as far as backwardness of the class is concerned, we are afraid that we must reject Shri Shanti Bhushan’s argument. The reference to ―class‖ is to the Scheduled Castes and the Scheduled Tribes, and their inadequacy of representation in public employment. It is clear, therefore, that Nagaraj (supra) has, in unmistakable terms, stated that the State has to collect quantifiable data showing backwardness of the Scheduled Castes and the Scheduled Tribes. We are afraid that this portion of the judgment is directly contrary to the nine-Judge Bench in Indra Sawhney (1) (supra). (paragraph 14)

While this is no doubt correct on law (and indeed, on logic – the Presidential list under Articles 341 and 342 containing the list of Scheduled Castes and Scheduled Tribes is already based on an assessment of marginalisation and vulnerability), it does raise a question of judicial propriety. Can a five-judge bench hold that a coordinate bench wrongly interpreted the law, and is therefore incorrect? Does not propriety require that the subsequent bench refer the case to a seven-judge bench, which can then overrule the (allegedly) incorrect judgment? Not that this is not even a case where it was argued that Nagaraj was per incuriam: rather, what was advanced was the far more modest claim that Nagaraj got Indira Sawhney wrong. But even if Nagaraj got Indira Sawhney hopelessly, irredeemably wrong, that was not for the Constitution Bench in Jarnail Singh to decide.

Creamy Layer

Nariman J. then went on to hold that even though Indira Sawhney had not expressly chosen to apply the creamy layer principle to Scheduled Castes and Scheduled Tribes, it had always been clear that the principle was a facet of constitutional equality. Nariman J. relied upon N.M. Thomas for this principle, relying upon some observations in Krishna Iyer J.’s concurring opinion, to note that:

The whole object of reservation is to see that backward classes of citizens move forward so that they may march hand in hand with other citizens of India on an equal basis. This will not be possible if only the creamy layer within that class bag all the coveted jobs in the public sector and perpetuate themselves, leaving the rest of the class as backward as they always were. This being the case, it is clear that when a Court applies the creamy layer principle to Scheduled Castes and Scheduled Tribes, it does not in any manner tinker with the Presidential List under Articles 341 or 342 of the Constitution of India. The caste or group or sub-group named in the said List continues exactly as before. It is only those persons within that group or sub-group, who have come out of untouchability or backwardness by virtue of belonging to the creamy layer, who are excluded from the benefit of reservation. Even these persons who are contained within the group or sub-group in the Presidential Lists continue to be within those Lists. It is only when it comes to the application of the reservation principle under Articles 14 and 16 that the creamy layer within that sub-group is not given the benefit of such reservation. (paragaraph 15)

Importantly, Nariman J then went on to hold that courts could themselves “exclude” the creamy layer from reservations (paragraph 16).

But for a conclusion of this magnitude, Nariman J.’s reasoning is disappointingly sketchy. There are at least two good reasons why the creamy layer doctrine should not be applied to Scheduled Castes and Scheduled Tribes, neither of which are addressed or acknowledged by the judgment. First, the very concept of a “creamy layer” presupposes that some members of a subordinated group have attained a level of privilege, that they no longer share the characteristics of subordination, and therefore, are no longer a part of the “group” in that limited sense. Now, when the subordination is economic or political, it makes sense to assume that it is at least possible for some people, or groups of people, to “escape” that subordination, so to say. However, that is a far more dubious claim when the subordination is social in character, and is founded on discrimination of a certain kind, where group identity itself is the locus of disadvantage. We don’t have to go too far afield to understand this – there are more than enough examples of Dalits who have broken free of economic marginalisation or penury, but continue to be subjected to the most shocking forms of social discrimination. And indeed, this has historically borne out: right from the early-20th century, the first lists of “Depressed Classes”, drawn up by British commissions, focused on social stigma as the basis of classification (for a detailed discussion, see Marc Galanter’s Competing Inequalities). Recall Ambedkar’s own words about how caste society was based on an “ascending scale of hatred and a downward scale of contempt.” Does the concept of a “creamy layer” make any sense here? Well, if it does, than it requires substantially more detailed justification than the Court provides.

And secondly, even if we are to accept that “creamy layer” makes conceptual sense in the case of SCs and STs, there nonetheless remain justifications for the existing reservation model that the Court does not touch upon. It is argued, for example, that even if the benefits of reservation are “captured” by the elite within a group, even that constitutes positively towards greater social mobility (of the group) in the long term apart from ancillary benefits of greater political salience. I am not saying that these arguments are correct, or persuasive; however, it does behoove the Court to address them if it is going to introduce the creamy layer concept into these two categories.


Jarnail Singh suffers from two problems. On one point, it overrules a coordinate bench without even a finding of per incuriam. And on the second point, it puts its seal on a significant doctrinal shift in reservations jurisprudence, but does not back it up with the depth of analysis that it demands. I suspect, however, that this is not the last that we have heard on the issue – especially from the government’s side!