We need well-designed and nationally-accepted tools to measure social inequality

If one of India’s poorest states could pull off a caste count, what prevents other states or the Union of India from conducting a similar exercise? There are three key reasons behind the reluctance to conduct a nation-wide caste census: fear of legitimizing caste, vested interests, and concerns related to data integrity.

One of the motivations behind the imperial censuses of the 19th century was to determine the relative weights of different social groups in the native population. Nationalists saw such inquiries as part of British efforts to sow divisiveness within the country. The censuses in British India also faced criticism for miscounting caste.

All caste details, barring those related to Scheduled Castes (SCs) and Scheduled Tribes (STs), were left out of census questions after India won independence. In the early years of nation-building, caste was perhaps seen as an outdated social construct that must be eradicated. Official acknowledgement of caste groups risked legitimizing a social evil. The first backward classes commission led by Kaka Kalelkar in the 1950s failed to arrive at a consensus on caste-based quotas because of such concerns.

Despite attempts to banish it, caste endures in 21st century India. Just listen to any election campaign speech, and you will find numerous references to castes and subcastes. And don’t blame just the politicians or poor voters. If India’s middle class didn’t care about caste, you wouldn’t find so many references to it in the matrimonial sections of leading dailies. Regardless of whether the next census records caste or not, questions about caste will continue to be raised.

The other reason behind resistance to a caste-based enumeration lies in vested interests. Forward castes fear that the ambit of caste quotas may be enlarged. Upper segments of Other Backward Classes (OBCs) fear that socio-economic data on subcastes will expose the wide chasm between them and lower OBCs. Some forward caste groups, which want to be included in the list of OBCs, fear that such data might undermine their claims of backwardness.

The third key concern relates to data integrity. Census officials have been reluctant to include caste details because they fear that it might compromise census operations. What if powerful sub-castes exaggerate their numbers to demand a greater share of government largesse? What if census enumerators (typically local government employees) are co-opted by such groups? Given the history of failed censuses in countries with sharp ethnic fault lines, such concerns aren’t baseless. Huge variations in caste groupings across space and time make it a challenge to enumerate jatis accurately.

In a comprehensive caste count, demands to club different caste names in one group or split one name into separate entities could overwhelm the census office, the former census official B.K. Roy Burman had warned in a 1998 Economic and Political Weekly article. He argued that the census office should conduct field studies to capture caste details. The 1961 census had followed this approach. In addition to the statutory census data, non-statutory surveys of villages and weddings were conducted to collect caste details, wrote Burman, a noted anthropologist who headed this initiative. The census office roped in independent scholars to conduct village-level case studies that documented changing profiles of caste relations and caste mobility in rural India.

The public discourse on the caste census has focused largely on its electoral implications. The mechanics of conducting such a census has not received the attention it deserves. It is worth noting that the 2011 socioeconomic caste census failed to produce jati-wise details because of lack of adequate preparatory work. The next caste census will have a better fate only when the issues involved in enumerating caste details are discussed threadbare before such an exercise is launched.

In line with Burman’s suggestion, it may be fruitful to combine an extensive census with intensive field studies. Detailed case studies and representative state-wise surveys of caste groups can uncover facets of the caste system that a short census questionnaire is unlikely to reveal. Such studies can also help validate the data generated in the full census count. Given the wide variation in social dynamics across India, state statistical officers and independent scholars should be involved in this exercise.

Such an exercise will be easier to undertake if census operations are brought within the ambit of the national statistics office. As this column has argued earlier (‘Raising the development quotient of the census’, 5 June 2023), a disconnect between census operations and surveys in India limits the potential of both. A well-designed caste census that is rolled out after adequate consultations and pilots can deliver results that are trusted by all sections of society, and are useful for all tiers of government.

This is the second of a three-part series on measuring inequality.