Census data in India are losing their relevance in the development agenda
India is busy debating the caste census when the regular Census itself has not been conducted owing to the pandemic. It is quite ironic that various elections have been held, and people gathered together at large rallies flouting COVID-19 norms, while the Census has still not been conducted. This is the first time that India has not conducted its decadal Census since the exercise began.
The design of the Census (whenever the exercise is held) can be improved. A digital Census would ensure better quality, coverage and quick results in this digital age. Given this promise on the one hand and the uncertainty in conducting the Census on the other, the demand for including caste enumeration within the Census only adds to the confusion.
First, we must recognise what the Census does. It has lot of potential in policymaking and the exercise is not merely about counting the population. Unfortunately, though, the limited information collected, and the under-utilisation or non-utilisation of Census data, have limited the role of the Census in policymaking.
Its importance is further diminished when numerous large-scale surveys are funded by the various ministries of the Government of India. These surveys are conducted periodically. They allow for a detailed analysis of the socio-economic issues of significance since the raw data are made available in the public domain. Hence, the Census, at best, serves as a framework for designing these surveys.
But the fundamental reason why the Census has lost significance is because the data collected are not disseminated on time, despite the use of technology. The primary reason for this is that the government regulates the release of the numbers based on its calculations of whether or not the Census data have the potential to harm the political agenda. For instance, the data on internal migration collected in the 2011 Census were made available to the public only when the Chief Economic Advisor decided to write a chapter for the Economic Survey 2016-17 in 2017.
This more-than-century-old decadal exercise is a matter of pride and distinction for this country. Unfortunately, its potential is hardly tapped by policymakers. Concerns now are only about counting castes and minorities, which will help political masters serve their own interests. Census data are mainly used by demographers, who have now redefined themselves as data analysts.
That this exercise has been reduced to just a count of the population is a great pity. Census-based information was important at a time when there was no alternative way of gauging the dynamics of population change alongside its varied features like employment, education, etc. While there is no denying the fact that alternative sources of information have enriched our understanding of population dynamics and facilitated focused interventions through programmes and policies, the Census has lost its potential relevance. Information is released late owing to bureaucratic regulations. There is also a lack of interest by the scientific community in a nuanced exploration of the data.
Despite the decadal nature of the data, the inter-Censal and post-Censal information could very well be generated with interpolation and extrapolation. Further, the fundamental demographic attributes around which the Census data are structured offer a lot of scope for interpretation and exploration for understanding future trends as well. The pseudo cohort inspection of the Census data can go a long way in informing us of the changing dynamics of population attributes over time. The fascination and engagement with the Census have been quite limited to two concerns: sex ratio and work participation (female work participation in particular). But the Census data, if explored intelligently and systematically without the limitation of survey-based data sets like biases, errors and representational issues, have much more potential.
The primary axes of disaggregation of Census-based information are residence, age, gender, administrative units, Scheduled Castes and Scheduled Tribes, and religion. Apart from such disaggregation, the Census offers two units of analysis: at the individual level and at the household level. These may appear quite limited, but a lot can be inferred from these attributes of disaggregation. Attributes of disaggregation are simply meant for identification and they are more neutral for intervention purposes. Disaggregated attributes should serve a purpose, i.e., help policymakers make interventions, if any. If the reason behind such a purpose is to gauge selective adversity or failure in entitlements, then ascribed attributes like caste and religion are perhaps less important than objective criteria like adversity or failure itself. In fact, associating caste/religion for identification and intervention generates an environment of patronage. In political terms, this can create clientelism. While there is no disagreement that systematic adversities are generated by one’s caste position, it is not necessary to have the count of the attribute as it is to know the magnitude of adversity and its locational attributes. With a widespread information base through administrative records as well as periodic surveys, it is not difficult to focus on these adversities and alleviate them.
Counting ascribed identities like caste and religion is perhaps less progressive than counting achieved identities or capability attributes like education and profession and other tangible endowments like the ownership of land, house and other consumer durables. Further, associating any adversity with an ascribed identity may at best help focus the intervention but the effort should be on addressing the adversity irrespective of the identity. Injustice or wrongdoings need not necessarily be associated with ascribed attributes. In fact, many make the fallacy of association leading to causation and that leads them to conclude that adversity/discrimination associated with ascribed attributes are largely due to the attributes themselves. Going beyond this association and examining the failure in entitlements and circumstantial differences will perhaps be more effective in thinking of interventions and in addressing concerns. A better example to this effect is blaming certain minority communities for high fertility rates rather than identifying the real reason for the same in terms of socio-economic exclusion.
On the whole, count and characteristics are equally important, but the characteristics that are modifiable hold the key towards change. It is rightly said that what can be counted may not count and what counts is seldom counted.
S. Irudaya Rajan is Chairman, The International Institute of Migration and Development, Kerala. U.S. Mishra is Professor, Centre for Development Studies, Kerala