Is the caste Census a useful exercise?

The article explores the challenges of implementing a caste Census and individual caste-based reservations in India.

The demand for a caste Census has become a heated political issue, fuelled by calls from opposition leaders, NGOs, and, more recently, the Rashtriya Swayamsevak Sangh (RSS) also adding itself to the cohort. Proponents argue that such a Census would determine the population sizes of various castes and that these numbers can be used to provide a proportionate share to each caste in government jobs, land, and wealth.

This article discusses how the attempt to collect individual caste data will once again prove to be a futile exercise, and how individual caste-based proportionate reservations is a regressive policy.

The exercise of a caste Census in India dates back to the late 19th century when the first detailed caste Census was conducted in 1871-72. It attempted to collect caste-based information and classify various groups, and was conducted across four major regions — the North-Western Provinces (NWP), the Central Provinces (CP), Bengal, and Madras.

There were several arbitrarily constituted “sets” based on a very superficial understanding of caste. In the NWP, for instance, only four sets were officially recorded — Brahmins, Rajputs, Banias, and “other castes of Hindus”. Meanwhile, in the CP, groups such as “servants and labourers” and “mendicants and devotees” were arbitrarily included under these sets. Some of Bengal’s classifications included beggars, musicians, and cooks, while Madras added “mixed castes” and “outcastes” as distinct categories.

Frustrated with the complexities of understanding caste, W. Chichele Plowden, who prepared the 1881 Census report, termed the whole question of caste ‘confusing’ and hoped that ‘on another occasion no attempt will be made to attempt to obtain information as to the castes and tribes of the population’. However, the same issues persisted in the caste Census of 1931 where 4,147 castes were identified. The officials were surprised to find that caste groups frequently claimed different identities in different regions.

These challenges are not relics of the past but continue to shape the difficulties India faces to conduct a caste Census today. For instance, the Socio-Economic and Caste Census (SECC) of 2011 identified over 46.7 lakh castes/sub-castes with 8.2 crore acknowledged errors. A more recent example is the controversy surrounding the inclusion of ‘hijra’ and ‘kinnar’ as categories in the caste list in the Bihar Census (2022).

Upward caste mobility claim —the reporting of one’s caste by respondents can be influenced by the perceived prestige associated with certain social groups and their position within the varna hierarchy. This is evident from changes in caste claims between the 1921 and 1931 Censuses, where some communities that initially reported belonging to lower positions within the varna system in 1921 later reported themselves as belonging to higher categories in 1931 (see Table 1). Another notable observation from these claims is that different members of the same community, such as Sonar, reported belonging to different social categories —Kshatriya and Rajput in 1921, and Brahmin and Vaishya in 1931, in the same region (see Table 1). These occurrences were noted in colonial Censuses but their implications remain relevant even today.