Breaking down the share of social groups among religions

The Supreme Court on Thursday upheld a Madras high court decision that barred the Tamil Nadu government from providing reservation to those who convert to Islam on the basis of their pre-conversion status. In March 2024, the Tamil Nadu government allowed people converting from Backward Classes, Most Backwards Classes, Denotified Communities, or Scheduled Castes (SCs)to be treated as one of seven Muslim sects that are classified as backward for issuing community certificates . The Madras high court struck this government order down earlier this year, and the Supreme Court has upheld the high court’s decision.

How large is the population affected by these changes? How significant is the share of backward classes in religions other than Hinduism? Are these numbers affected by economic class? Government and private surveys can help in answering these questions.

Religious conversions are extremely rare in India – across religions

There is no official data on religious conversions in India . However, a survey conducted by Pew Research Center, a US-based not-for-profit research organisation, among 29,999 adults in India in the November 2019-March 2020 period shows that conversions are rare. The survey asked respondents if the religion they identify with is different from the one they were raised in. The total proportion of such people was under 2%. To be sure, because the survey’s sample is small and conversions are rare, it is not possible to say if trends in Tamil Nadu in particular are any different.

Chart 1

What is the social break up of religions in India?

The classification of castes into categories such as Scheduled Castes (SCs), Scheduled Tribes (STs), or Other Backward Classes (OBCs) is done officially in India. But there is no administrative count of people who can be certified as such, which is what matters for reservation. However, estimates of these groups – individual states may also have categories such as Most Backward Classes (MBCs) – are available from government surveys, where people self-identify their categories. While the numbers are not exactly the same, two big surveys – 2019-21 National Family Health Survey (NFHS) and the 2023-24 Household Consumption Expenditure Survey (HCES) – agree on broad trends.

One, the share of SCs is negligible among Muslims, lower than average among Christians, and higher than average among Sikhs. To be sure, there is no official recognition of SC status among Christians, even if some people self-identify as such. Similarly, while Buddhists can also identify as SC, the proportion of Buddhists is very small in both surveys to see this meaningfully. Two, the proportion of STs is higher than average among Christians and lower than average among religions other than Hinduism. Three, the proportion of OBCs among Muslims is more than that among Hindus, with the estimate somewhat different among surveys. In any case, changes in rules and laws around who gets counted as a Muslim OBC are not an insignificant change.

Chart 2 and 3

Does economic class affect social group breakup of religions?

This is important because the rich among OBCs – the so-called “creamy layer” -- are excluded from reservations, at least in central government jobs. Removing the richest 10% by monthly per capita spending (MPCE) from the anlaysis shows that the share of OBCs among Hindus and Muslims remains almost the same. To be sure, the share of Muslims among OBCs is less than average in the richest 10% MPCE class and that of Hindus higher than average. However, this does not alter the composition of the rest of the population drastically.

Another thing to keep in mind while reading these numbers is that the “creamy layer” threshold is based on income and not spending. However, government surveys in India so far only ask about spending patterns. The richest 10% MPCE class had a minimum annual per capita spending of ₹99,474 in the 2023-24 HCES. That translates to an annual spending of ₹397,898 for a family of four, well below the ₹8 lakh income threshold set by the central government.