Plug data gaps: State-level statistical surveys could help

During its long journey, the NSSO has faced criticism and controversies, particularly over the past three decades. 

The first of these was around its 1999-00 consumption survey. But the agency weathered that storm by boldly releasing unit-level data to the public. That let researchers from India and overseas examine the issues involved. The end result was a rich trove of literature on various details, including the ‘recall period’ used for consumption and other elements of poverty measurement. All this gave us the ‘Great Indian Poverty Debate.’

Also Read: NSS: Not just nerdy stats but a data arsenal for business and policy

The second big dust-up was over the consumption survey of 2017-18. However, this time, the agency withdrew the entire survey, although its data had already been vetted by the National Statistical Commission. This led to a vacuum in India’s poverty statistics. 

Recently, the NSSO released its consumption surveys for 2022-23 and 2023-24. Hopefully, this will again generate the same kind of debate and eventually help enrich the country’s statistical system.

However, the same cannot be said of the country’s other major statistical agency, the Registrar General of India (RGI).

The RGI plays an important role, as it is responsible for India’s decennial population census. Again, a first for the RGI, the Census of 2021 has been indefinitely postponed, although covid is long gone. This headcount, which predates independence, was last disrupted only by World War II and has otherwise served as a valuable source of information on several aspects of our social and economic structure.

In fact, it is the only data with information on social and economic characteristics at the village, ward and district levels, which is vital for decentralized governance. 

Also Read: India's Census needs an urgent update: Vital data mustn't be delayed

But it also has another crucial role. 

Census data is the basis of sampling for all surveys, including those conducted by the NSSO. The RGI is also responsible for other important data, such as statistics on fertility, mortality and various demographic variables, which are collected as part of the Sample Registration System (SRS). This sample survey is crucial for monitoring our demographic changes. Unfortunately, even this publication has not been released after 2020.

In this backdrop, an important development is the Telangana government’s tabling of basic numbers in its assembly from a caste census conducted by the state. While it released only basic community-wise aggregates, it will be useful if the government also releases granular data on not just population estimates, but also socioeconomic characteristics collected as part of this exercise.

The history of such experiments has not been encouraging. The nationwide caste count conducted in 2011 along with the Socio Economic and Caste Census (SECC) has more or less been buried. Karnataka, which conducted its own caste census in 2015, has not yet released the data. The only state to have released some data is Bihar. It released the results of its 2023 caste census, although only partially.

Also Read: The Bihar caste survey marks a big step towards a reality check

The initiatives of Telangana and Bihar are welcome for two reasons. 

First, attempts by states to conduct their own surveys and headcounts in the absence of national data help fill the vacuum created by the absence of national data. 

Second, these are also important for governance, in which states play a major role. Under India’s constitutional structure, most of the legislative subjects that matter for citizens at large—such as land, agriculture, education, health and social protection—are largely in the domain of states.

The availability of up-to-date data at the granular level of villages, wards and blocks will allow states to design better policies, at least in fields of application where the state is the major decision-maker.

Attempts by state governments to conduct their own surveys are also important for the creation of state-level statistical capacity. Even though states have their own directorates of economics and statistics (DESs), the Centre does most of the data collection and generation of statistics in India. 

Also Read: TCA Anant: The budget represents a missed opportunity to strengthen Indian statistics

Earlier, for most NSSO surveys, a state’s DES would conduct its own survey with a matching sample. Unfortunately, this is no longer done. Yet, since states differ in their social, political and economic structure, a strong statistical system at the state level is necessary for them to generate, own and use data-sets that are best suited for local governance. State governments have therefore done well to conduct their own census surveys.

These efforts cannot take the place of a national census, though. That’s primarily because these are conducted at different points of time. Also, their methodologies and coverage may differ from one state to another. 

However, if Telangana or Bihar manage to utilize such data for a granular approach to planning and public-service delivery, they could set an example for other states. More importantly, such an approach could serve as a state-level safeguard against the Centre’s apparent tendency to delay or withdraw vital statistical information.

The author is associate professor at Jawaharlal Nehru University and visiting fellow at the Centre de Sciences Humaines, New Delhi.