Kalelkar Commission explained: how India first tried to define who is backward

The Constitution allowed the State to make special provision for backward classes and never said what a backward class was. That left two questions open, and independent India spent 40 years answering neither cleanly.

The first is identification: who counts. The second is quantity: how much is too much. They were argued in different rooms, by different people, and they only met in 1979.

The chairman who asked for his own report to be thrown away

The Government of India appointed a Backward Class Commission under Article 340 on 29 January 1953,¹ and it became known by its chairman's name.¹ It reported on 30 March 1955, listing 2,399 castes.¹

Then the thing that makes this the strangest document in the sequence:

The Chairman of the commission, Kaka Kalelkar, however, had second thoughts after signing the report.

In his covering letter he pressed for the report's rejection. The man who ran the inquiry, having signed its findings, asked the President not to act on them.

A Memorandum of action was placed before Parliament on 3 September 1956.¹ The report was considered over the next half-decade and rejected in 1961.¹

A machine with no definition inside it

Article 340(1) empowers the President to appoint a Commission to investigate the conditions of socially and educationally backward classes.² Article 340(2) requires that Commission to investigate and present a report. Article 340(3) requires the President to lay the report before Parliament, together with a memorandum explaining the action taken on it.³

Read the machine carefully. It says who may appoint, what they must investigate, and what must be laid before Parliament. It does not say what socially and educationally backward means.

Article 16(4) has the same hole. It provides that nothing in the article prevents the State from reserving appointments or posts in favour of any backward class of citizens not adequately represented in the services.¹ Backward class again, undefined again.

Ambedkar was asked in the Constituent Assembly what the words meant, and his answer was not a test:

A backward community is a community which is backward in the opinion of the Government.

That sentence is the whole of the next 40 years. If backwardness is whatever a government says it is, then every list is a political act and every list can be challenged, and both of those things happened.

The lists were already being made

The Constitution did not invent the category. It inherited it.

A Government of Bombay Finance Department Resolution of 5 February 1925 defined Backward Classes as all communities other than a listed set.¹ A 1930 Bombay State Committee then produced three lists at once: 36 Depressed Classes with an approximate 1921 population of 1.475 million, 24 Aboriginal and Hill Tribes at 1.323 million, and 95 Other Backward Classes at 1.041 million.¹

In 1935 an order of the Government of the Maharaja of Travancore recognised separate communities for purposes of public service.¹ The Madras Provincial and Subordinate Service Rules of 1942 defined backward classes as the communities mentioned in a schedule.¹

Four administrations, four schedules, and none of them a definition. A list is not a test. It is the output of a test nobody wrote down.

83 posts, six columns

The first collision came over jobs. In December 1949 the Madras Public Service Commission invited applications for 83 posts of District Munsifs,¹ to be filled on a fixed communal ratio.

The 83 vacancies were distributed in the following manner: Harijans-19, Muslims-5, Christians-6, Backward Hindus-10, Non-Brahmin (Hindus)-32 and Brahmins-11.

Six columns, adding to 83. Every judicial appointment in the state passed through that grid. The related challenge to the same scheme in education became State of Madras v Champakam Dorairajan, decided under Article 15.¹

Parliament's answer in 1951 was the First Amendment, which wrote Article 15(4) into the Constitution so that special provision for backward classes could not again be struck down as discrimination. What it did not do was define the class it was protecting.

The government's case against caste

In 1961 the Centre tried to close the definition question from the other end. ¹ Government of India it would be better to apply economic tests.

The spelling is the judgment's own. The instruction is a recommendation, not a rule: the States keep the discretion and are told what the Centre would prefer they do with it.

That letter is the fork. An economic test is checkable and individual. A caste test is inherited and collective. From 1961 the two run alongside each other, and no instrument ever chooses.

How much is too much

While identification went unresolved, the courts were building a ceiling nobody had legislated.

The rule arrived in Balaji: a special provision ¹

Read what that is. Not a number in a statute. A judicial estimate, hedged twice, in a judgment. Every argument for the next 30 years is about whether a hedged estimate is a limit.

The states tested it immediately. On 26 July 1958 Mysore issued an order under Article 15(4) ¹

Everyone except one community, at 75 per cent. On 13 July 1972 a further Mysore order reserved 68 per cent of seats in engineering and medical colleges.¹

How 17 per cent became 65

Devadasan showed that a ceiling can be crossed without anyone raising a quota.

In the Central Secretariat Service scheme, reservation was 12.5 per cent for Scheduled Castes and 5 per cent for Scheduled Tribes.¹ Seventeen and a half per cent, well under any ceiling. The problem was the carry-forward rule, which rolls unfilled reserved vacancies into the following year.

Of the 45 appointments made in 1961 under that rule, 29 went to Scheduled Castes and Scheduled Tribes.¹

In other words, the extent of reservation in the third year came to 65 per cent.

A 17.5 per cent quota produced a 65 per cent year. The Constitution Bench declared the carry-forward rule unconstitutional, Subba Rao J dissenting.¹

That is the first time the ceiling is shown to be a question about arithmetic over time rather than about the number written in the order.

Seven judges, seven opinions

Thomas went the other way. In Kerala, of 51 Upper Division Clerk vacancies arising in 1972, 34 were filled by members of the Scheduled Castes and Scheduled Tribes after the State granted serving members an exemption from a qualifying test.¹

The Stats of Kerala carried the matter in appeal to this Court which was allowed by a majority of 5:2. All the Seven Judges wrote separate opinions.

The typing is the judgment's. Seven judges, seven opinions, a 5 to 2 result, and no single ratio that a later court could apply cleanly.

Vasant Kumar repeated the pattern: a Constitution Bench of five, each expressing a different opinion on the caste-versus-economic question, with Chandrachud CJ opining that existing reservations should continue for a further period.¹

Twelve judges across two benches, twelve opinions, and the question still open.

What Kalelkar was afraid of

The identification problem produced a figure that shows why he wanted his own report rejected.

in Tamil Nadu out of 27,05,960 people belonging to Muslim minorities 25,60,195 are included in the backward list which works out to 94.61% of the total Muslim population of the State.

When 94.61 per cent of a community is inside the backward list, the list has stopped discriminating between the backward and the rest of that community. It has become a description of the community.

That is the failure mode Kalelkar signed his name to and then wrote to disown: a test that identifies almost everybody identifies nobody.

Twenty-five to forty-four, struck down at sixty-five

Andhra Pradesh produced the cleanest illustration of the ceiling doing work.

¹

The list survived. The quantum did not. The enhancement had taken total reservation under Article 16(4) in the State to 65 per cent.¹

Note what the court accepted and what it refused. Who is on the list was allowed to stand. How many seats the list may claim was not. Identification and quantity, separated in a single judgment.

Mandal, submitted and shelved

In 1979, by Presidential Order under Article 340, a second Backward Classes Commission was appointed, popularly known as the Mandal Commission.¹ One of its terms of reference was to determine the criteria for defining the socially and educationally backward classes, which is the question left open in 1950.¹

Its report was originally due by 31 December 1979, the deadline was extended, and it was submitted on 31 December 1980.¹

Then it was laid before each House and discussed twice, in 1982 and in 1983.¹ R. Venkataraman, then Minister for Defence and Home Affairs, told the Lok Sabha in 1982 that the debate had cut across party lines, and that a number of people on his own side had supported the Mandal Commission's recommendations.¹

P. C. Sethi, then Home Minister, said in 1983 that although the Commission had been appointed by a predecessor Government, his Government now desired to continue with it and implement its recommendations.¹

Support across party lines in 1982. A stated intention to implement in 1983. And so far as Central services were concerned, no reservations were ever made in favour of Other Backward Classes, though reservations existed in some southern States for years before the Constitution.¹

Praised in two Parliaments and acted on in neither, for 10 years.

The question is still open

Sixty-six years separate the Bombay resolution of 1925 from the moment this report was finally picked up. In that time four administrations wrote lists, a commission's own chairman asked for his report to be binned, the Centre recommended economic tests and left the choice to the States, and the courts built a 50 per cent line out of a sentence that said speaking generally and in a broad way.

What nobody produced was a definition.

That is what was inside the report sitting on the shelf, and what happened when a government finally opened it is the subject of the next piece.

MangoPOV. "Kalelkar Commission explained: how India first tried to define who is backward." 14 September 2026. https://mangopov.in/stories/who-counts-as-backward-kalelkar-to-mandal/

@misc{mangopov-who-counts-as-backward-kalelkar-to-mandal,
  title        = {Kalelkar Commission explained: how India first tried to define who is backward},
  author       = {{MangoPOV}},
  year         = {2026},
  howpublished = {\url{https://mangopov.in/stories/who-counts-as-backward-kalelkar-to-mandal/}},
}