Counting castes in India has always been about more than numbers - it is about who gets a share of government benefits and who doesn't.
The country's next national census, scheduled for 2027, will - for the first time in nearly a century - count every caste, a social hierarchy that has long outlived kingdoms, empires and ideologies. The move ends decades of political hesitation and follows pressure from opposition parties and at least three states that have already gone ahead with their own surveys.
A 2011 survey - neither run nor verified by census authorities or released by the government - recorded an astonishing 4.6 million caste names.
A full count of castes promises a sharper picture of who truly benefits from affirmative action and who is left behind. Advocates say it could make welfare spending more targeted and help recalibrate quotas in jobs and education with hard evidence.
Yet in a provocative new book, The Caste Con Census, scholar-activist Anand Teltumbde warns that the exercise may harden the deeply discriminatory caste system when the need is to dismantle it.
The argument cuts against the prevailing view that better data will produce fairer policy. For Mr. Teltumbde, castes are too pernicious to be managed for any progressive purpose.
Caste is, at its core, a hierarchy seeking impulse that defies measurement, he writes.
Mr. Teltumbde sees the modern caste census as a colonial echo.
British administrators began counting castes in 1871 as a deliberate response to the post-1857 unity of Indians across caste and religion, turning it into an effective tool of imperial control. They held six caste censuses between 1871 and 1931 - the last full caste enumeration in India.
Each count, Mr. Teltumbde argues, did not merely record caste, but reified and hardened it.
Independent India, in Mr. Teltumbde's reading, preserved the system under the moral banner of social justice, while effectively evading its core obligation of building the capacities of all people, which is a prerequisite for the success of any genuine social justice policy.
The obsession with counting, he says, bureaucratises inequality. By turning caste into a ledger of entitlements and grievances, the census reduces politics to arithmetic - who gets how much - rather than addressing what Mr. Teltumbde calls the architecture of social injustice.
He sees the demand for a caste census as a push for more reservations - a cause driven by an upwardly mobile minority, while the majority slips into deprivation and dependence on state aid. Nearly 800 million Indians, he notes, now rely on free rations.
Political scientist Sudha Pai broadly agrees with Mr. Teltumbde's critique that counting castes can solidify identities and distract from deeper inequalities based on land, education, power and dignity; however, she acknowledges that caste has already been politicized through welfare and electoral strategies, making a caste census inevitable.
A caste census would be useful if the income levels within each caste group are collected. The government could then use the data collected to identify within each caste the needs of the truly needy and offer them the required benefits and opportunities, such as education and jobs for upward mobility, Dr. Pai says.
Yet, scholars warn that counting castes and interpreting the data will be fraught with challenges.
It won't be painless. India has changed tremendously in the century since 1931. Castes that were designated as being poor and vulnerable may have moved out of poverty, some new vulnerabilities may have emerged. So if we are to engage in this exercise honestly, it cannot be done without reshuffling the groups that are eligible for benefits, says Professor Desai.
Mr. Teltumbde remains unconvinced. He argues that endless enumeration cannot remedy a system built on hierarchy. You will be counting all your life and still not solve the caste problem. So what will be the use of that counting? he wonders. I am not against affirmative action, but this is not the way to do it.
