Scheduled Castes, Creamy Layer and Jarnail Singh Judgment

[Editor’s Note: The Supreme Court’s approach on merit and efficiency of administration in the cases of the reservation is not a correct approach conceptually and philosophically, and such an approach leads to deflection from the values of the constitution and compromise the struggle for constitutional justice. This approach is based on the view that reservations and merit are opposed to each, instead, there is a need to balance the two.]

In the Jarnail Singh v. Lachhmi Narayan Gupta judgment, the Constitutional Bench of the Supreme Court decided two questions of law: one, whether the ‘creamy layer concept’ should apply to Scheduled Caste and Scheduled Tribes, and second, whether there is a need to reconsider the opinion of the five-judge bench in M. Nagaraj v. Union of India which held that ‘the state has to collect quantifiable data showing backwardness of SC/STs’? In this post, the author will analyse the judgment of the Court in Jarnail Singh and discuss the issues which arise from it.

How is it determined that a person belongs to SC/ST Category?

Under Article 341 of the Constitution, it is stated, in clause 1 that, it is the President who shall specify, through a public notification, which ‘castes, races or tribes’ shall deem to be a part of the Scheduled Castes. Further, in clause 2, it is the power of the Parliament to exclude or include any caste, race or tribe from the list of Scheduled Castes. Inclusion or exclusion can be done only through a law made by the Parliament and not otherwise. (See addendum at the end of this post)

Article 335 of the Constitution states that the state shall not be restricted from making reservations in the favour of the Scheduled Castes and Tribes in making appointments to services and posts. The provision also says that the state must ensure the reservations made must be ‘consistently with the maintenance of efficiency of administration’. This brings the controversial argument of merit versus reservations. The Courts have thought of efficiency as a synonym of merit or the marks achieved by a candidate, but this position was philosophically challenged and curated in B.K Pavithra II v. Union of India (the author will analyse the judgment of BK Pavithra II in his next post and discuss more of this issue). Anyhow, it is clear from Article 341 and 335 that only the Parliament, by law, can make changes in Scheduled Castes/tribes list. In the judgment of E.V. Chinnaiah v. State of A.P. (2004), it was held by the five-judge bench of the Supreme Court that the group mentioned under the Presidential List of Scheduled Castes under Article 341 is a ‘homogenous group’ and it cannot be sub-divided by the State governments. But the validity of Chinnaiah is in question after the controversial judgment of State of Punjab v. Davinder Singh which was delivered in March 2020 by another five-judge bench. The Court in Davinder Singh, paragraph 41, held that the list under Article 341/342 is not homogenous and sub-divisions can be made under the list based on backwardness. Both Davinder Singh and Chinnaiah will be reconsidered by the seven-judge bench in future and will determine whether the SC/ST list is homogenous or heterogeneous.

Application of ‘Creamy Layer’ Concept to SC/STs

Justice Reddy in Indira Sawhney v. Union of India held that the creamy layer concept will not apply to the Scheduled Castes and Scheduled Tribes in paragraph 792. The decision and discussion in Indira Sawhney were confined to OBCs and the reservation thereto. The judgment also stated that the reservations would apply to an appointment in the initial stage only, but not in the promotional stages. Hence, there would be no reservation in promotion as per the judgment in Indira Sawhney. Then, the Parliament amended the Constitution and introduced Article 16(4A) and 16(4B) to overrule the promotional aspect of Indira Sawhney. The amended provision stated that nothing shall prevent the state from making reservations in promotion in favour of SC/STs. This was then challenged in M. Nagaraj v. Union of India. The judgment in Nagaraj upheld the constitutional validity of the Amendment but it presented another set of issues like the introduction of the concept of the creamy layer within the scheduled castes and tribes, for instance, the state shall collect quantifiable data to exclude a certain class of people within the scheduled castes/tribes from the benefits of reservation. However, no test was given to determine the quantifiable data.

So what is this creamy layer? The Court in Indira Sawhney permitted sub-classification of OBCs, as backwards and more backwards based on their comparative underdevelopment. For instance, there is a certain threshold, say (hypothetically) if any person has an annual income of more than eight lacs, then he/she will be excluded from the benefits of reservation or say, if a person belonging to OBC category becomes an IAS officer, then he/she will be excluded from the ‘more backward’ category and their heirs will not get the benefit of the reservation. But would that creamy layer concept will apply to Scheduled Castes and Scheduled Tribes? The caste of a person is determined by birth and that cannot be changed, so, even if a person (from SC/ST category) becomes economically and educationally well, then will that ‘tag’ of being an SC/ST as a social stigma will vanish? The author doubts that. This application of creamy layer will make sense if we can assume that “it is possible to escape one’s group identity (through prosperity or other ways of social advancement, for example).” But that assumption can be a dangerous proposition and empirically, not sustainable (see here and here). The reservation is afforded to the most backward communities of SC/ST because there is a history of social oppression by the virtue of their group identity. Hence, the list under Article 341 is considered to be homogenous and it should not be sub-divided as creamy or non-creamy as, it was argued by the author here that, “a group right cannot be redistributed since group discrimination cannot be reinterpreted to individual or sub-group rights”. The discrimination faced by the Scheduled Castes/Tribes is not because of their economic standing or job profile, unlike in the case of OBCs.

But in M. Nagaraj, the Court held that “if they (state) 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.” This allows the state to classify the Scheduled Castes as backward and more backward, inherently creating a sub-classification and a concept of creamy layer. This is where the problem started. Later in 2008, the Supreme Court in Ashoka Kumar Thakur v. Union of India (a reference to the opinion of CJ Balakrishnan) opined that the concept of creamy layer shall not apply to SC/STs and “creamy layer principle is not a principle of equality, rather a principle of identification of OBCs”. But that case again was confined to the matter of reservation for OBCs (as per the concurring opinions), hence this view of Justice Balakrishnan is disregarded by Court in Jarnail Singh.

The judgment in Nagaraj which allow for a sub-classification under SC/STs and the judgment in Chinnaiah which disallows sub-classification under SC/STs conflict with each other. But the Court in Jarnail Singh disregarded the conflict as it held that “Chinnaiah dealt with a completely different problem, apart from dealing with a State statute and not a constitutional amendment, as was dealt with in Nagaraj.” The reasoning in Chinnaiah was whether the state can sub-classify and tinker with the list made under Article 341 and even the Court’s holding in Nagaraj, in essence, dealt with the same thing (the part where the Court opined about ‘quantifiable data’ and creamy layer).

However, the Court in Jarnail Singh held that, fortunately, the proposition in Nagaraj which stated that the State has to collect quantifiable data for SC/STs in contradictory to Indira Sawhney in paragraph 14 and is bad in law. The part about quantifiable data is held to be unconstitutional, but the application of creamy layer principle was still dealt by the Court and in a questionable way. The Court considered the concept of the creamy layer as a matter of equality, not identification (contrary to the opinion of CJ Balakrishnan).

The concept of Creamy layer or the issue of “higher backward groups” among the SCs was dealt with by the Justice Krishna Iyer in State of Kerala v. N.M. Thomas where he held observed that in paragraph 124, reservation “benefits, by and large, are snatched away by the top creamy layer of the ―backward caste or class, thus keeping the weakest among the weak always weak and leaving the fortunate layers to consume the whole cake.” Further Justice Iyer observed that certain group among the SCs always want to wear the tag of “weaker sections”. But lastly, Justice Iyer observes that a lasting solution can come only if there is an “improvement of the social environment, added educational facilities and cross-fertilisation of castes by inter-caste and inter-class marriages sponsored as a massive State programme”. If that is the case, then unless the social group identity of SCs and STs are eradicated, until then there must not be any sub-classification and application of creamy layer because the excluded group, socially and politically, will be in a disadvantaged position. The social change is necessary, but not by tinkering with the scheduled castes list under Article 341 or by excluding a particular group. Economy and education do not annihilate a person from SC category, as there are plenty of cases where the educated group among the SCs is still discriminated and oppressed. Hence, I respectfully disagree with Justice Iyer’s observation on the creamy layer as this is an elitist way of thinking about reservations. Further, in Vasanth Kumar judgment (1985), justice Chinappa O Reddy remarked, “How can it be bad if reserved seats and posts are snatched away by the creamy layer of backward classes if such snatching away of unreserved posts by the top creamy layer of society itself is not bad?” He disagreed with Justice Iyer’s understanding of reservation and observed exclusion of creamy layer concept from SC/STs. As Professor Anurag Bhaskar** and Surendra Kumar argue, “The Indra Sawhney judgment, which was deciding all aspects related to reservations authoritatively, chose not to apply “creamy layer” criteria on the SCs and STs. In that way, Indra Sawhney can be said to have endorsed Justice Chinnappa Reddy’s proposition in KC Vasanth Kumar and not that of Justice Krishna Iyer in NM Thomas.”

The Court in Jarnail Singh, however, relies on the observation in N.M Thomas, but fails to consider the observation of Justice Balakrishnan in Ashoka Kumar Thakur and subsequently rejects it. Does this create confusion as to whether a view of a particular judge is over and above the view of the subsequent judge? This needs to be reconsidered by a larger bench. Furthermore, the Court suggests that exclusion a sub-group from the benefits of reservations will not tinker the list under Article 341, but that argument does not convince the reader as there is a lack of reasoning. In contrast, Justice Hedge in E.V. Chinnaiah held that “therefore, any executive action or legislative enactment which interferes, disturbs, re-arranges, re-groups or re-classifies the various castes found in the Presidential List will be violative of the scheme of the Constitution and will be violative of Article 341 of the Constitution.” Is the Court in Jarnail Singh by allowing re-arrangement or classification of SCs as creamy and non-creamy layers, in the name of harmonious construction of Article 14, 16, 341 and 342, allowing an unconstitutional action, as per Chinnaiah? We never know until a larger bench decides and settle all the connected issues concerning sub-classification and creamy layer of SC/STs.


The Court in Jarnail Singh rightly held that there is no need for quantifiable data to be collected by the state in providing reservations. However, the Court’s observations about creamy layer application to SC/STs is seriously flawed and needs to be revisited. Justice Nariman went on to say that the Courts can apply the creamy layer formula to SC/STs, in paragraph 16, which is a judicial overreach and must be reconsidered by a larger bench. The basis of the list prepared under ‘Scheduled Castes’ is based upon the social stigma that exists in the society, in B.R. Ambedkar’s own words, the caste is based on an “ascending scale of hatred and a downward scale of contempt.” In a society like this, is it even making sense to apply the principle of the creamy layer to a group whose identity is based on discrimination and a “fatal accident” of their birth? The author refuses to accept the proposition. The judgment in Jarnail Singh suffers from a problem that it shifts the jurisprudence on the reservation and does not back it up with ‘sufficient reasoning’. It needs to be reconsidered by a larger bench.  

* Addendum: When a person from any caste (for eg: Jatav, Chamaar, Ahir, Gujjar etc.) is included in the list created by the President under Article 341, then that group attains “a new status by the virtue of Presidential notification”. The word ‘caste’ under Article 16(2) does not include Scheduled Caste, that is why the state can treat SC/STs differentially to ensure substantive equality in fact. To support this argument, reliance is placed on the judgment of NM Thomas v. State of Kerala (paragraph 107).

** I would like to thank professor Anurag Bhaskar for introducing me to his brilliant paper, after the publication of this post, on inconsistency in Nagaraj and Jarnail Singh judgment as both of them, according to him, are ‘per incuriam’. I would suggest readers go through his paper: here.