The government?s decision to introduce a caste headcount in the ongoing decennial census once again brings to the fore a contentious issue that has remained unresolved since the colonial times. This is because the number of caste groups and their status in the social hierarchy differs from region to region. Also, different censuses and other studies have themselves thrust out vastly disparate figures about the actual number of castes themselves and their sub-groups.
So, for instance, while the 1901 Census identified 1,646 castes in all, the number rose to 4,147 in the 1931 Census, including 300 castes among Christians and 500 among Muslims. The main problem was that the government?s effort to identify caste itself led to the evolution of new caste labels that aimed at merging diverse castes spread across different geographical regions but with a common occupation. Thus was born the now politically ascendant Yadav category, created by combining Ahirs, Goalas, Gopis, Idaiyans and other castes of milkmen and grazers, pointed out by the 1931 Census Commissioner, JH Hutton. The latest numbers on caste are from The Peoples of India Project, launched by the Anthropological Survey of India, which identified as many as 4,635 different castes in the late-1990s.
If the identification of castes was so problematic then, the estimation of the size of the backward classes was even more complex and remains so even now. This grouping first came into significant use only in 1918 when the princely state of Mysore appointed a committee to investigate the status of backward communities in public service. Then in the mid-1920s, the Bombay government followed up with a definition for the group. Later other regions like the United Provinces, Travancore and Madras prepared lists enumerating these castes.
In Independent India, the Constitution spoke only about backward classes and not about backward castes. There was no major attempt to define backward classes at a national level so much so that TT Krishnamachari, India?s first finance minister, predicted that the term backward classes would be a paradise for lawyers. The general belief was that the grouping would be designated at a local level. And it also reflected an acceptance of the wide divergence of this category across the states and even within them.
Article 340 of the Constitution also mentions that the President is to appoint a backward class commission and this was first done in 1953 under the chairmanship of Kaka Kalelkar and later in 1979 under BP Mandal. While Kalelkar prepared a list of 2,399 backward castes, including 837 classified as most backward; Mandal enumerated 3,743 castes as OBCs and used the 1931 Census figures to estimate the Hindu and non-Hindu OBC population at 52%.
This estimate of the OBC population seems to be excessively high if one looks at the data from the estimates made by the NSSO. For instance, a 1999-2000 survey showed that 39.9% of Hindus and 31.2% of Muslims described themselves as OBCs. However, a later survey in 2004-05 showed that the share of Hindus describing themselves as OBC had gone up to 44.9% and that of Muslims to 40.9%.
But such self-supported identification is not deemed valid in the census procedures. For instance, the enumeration of scheduled caste and scheduled tribes is done strictly on the basis of the caste list supplied to enumerators, which varies across states. So even if a person identifies himself as SC or ST, he would be enumerated as one only if his caste or tribe name figures in the enumerators list.
Preparing state-specific lists of OBC from 4,000-odd castes would be a Herculean process. It can also raise new claims and counter-claims causing conflicts and violence as was seen during the Gujjar agitation in recent years. Fears have also been expressed that such a caste census would provide a stamp of legal-administrative sanction and only further promote caste identities. The only positive impact of a well worked out caste census will perhaps be that it would bring down what are likely to be widely exaggerated figures on the size of the OBC population. The 2001 Census estimates show that STs, SCs, Muslims, Christians and Sikhs account for roughly 42% of the population. That figure by itself suggests that OBCs are not likely to constitute 50% of the total population as is often claimed.
Still, given the many procedural difficulties involved, it is probably best that a caste census is avoided at this point in time.
p.raghavan@expressindia.com