Why everyone’s counting on the Census — and why a delay hurts

This cascading effect of a delayed census prompts a few questions: Is a census needed at all? Can India do without a census? What and how does the country stand to lose if it does not have timely census information?

Why have a census?

Since the start of economic reforms in 1991, there have been two big changes when it comes to publicly available data. One, India’s economy is becoming increasingly digital and generates a lot of data — for example, telecom and internet data — that can guide policymakers. Two, with the rollback of the centralised plan economy and the broad-based acceptance of the ‘minimum government’ paradigm, it can be argued that the government’s dependence on one central database to arrive at policy decisions has come down. Not to mention a whole bunch of surveys done by private firms. As such, an argument can be made that the country’s everyday functioning will not be terribly impacted if the census is delayed, or indeed, even done away with altogether.

The 2021 Census was put off due to the Covid-19 pandemic — the first time since 1881 that India’s decennial census could not be held. (File)

To begin with, the idea of dropping the census is not completely unheard of.

“Even in a country such as the US, their Census Bureau has considered stopping their census,” says Sonalde Desai, one India’s leading social demographers who is Distinguished University Professor, Department of Sociology, University of Maryland, and Professor and Centre Director, NCAER-National Data Innovation Centre, New Delhi.

Desai says the debate in the US was whether to instead simply rely on something called the American Community Survey, which is a “survey” that carries on continuously (every month, every year) and provides vital information on a yearly basis. Of course, being a survey and not a census, it covers only a sliver of population — 30-odd million in a decade.

But eventually, the US stuck to conducting the decennial census partly because of the constitutional requirement to have a census but more importantly, says Desai, because the ACS may not accurately capture from a policy perspective some of the most important segments of the population such as migrants and the homeless.

Many of these concerns are even more heightened in India’s case.

Even so, can India do without a census and instead look at some other source to supply the data? For instance, administrative statistics — that is, the data collected by government departments.

One could argue that over time, the Indian government has improved its capacity to know about the public. Since governments run a whole host of schemes, providing everything from general sanitation to individual housing, can’t all that data – instead of relying on an exercise as monumental as a census — be used to govern effectively?

Desai says the problem with basing policymaking on administrative statistics is that many Indians fall outside of their scope. For instance, the migrant workers. Since many schemes may not be effective in targeting them, using data from such schemes will lead to obvious exclusions.

The Census schedule of 1951, Independent India’s first Census. (Credit: Census of India)

Yamini Aiyar, President and CEO of Centre for Policy Research, points out that there are huge standardisation issues with MIS data across states. “Every government department has an MIS system, each one follows its own set of data standards, definitions etc. So comparability becomes a huge problem. Whereas in the census, the data is comparable across the country and over time,” she says.

Pronab Sen, former Chief Statistician of India, also warns against thinking that administrative statistics or any such set of data can replace the census. Administrative statistics can be faulty for a variety of reasons — ranging from neglect to incompetence to corruption.

Sen gives the example of the data on open defecation. “Recall that the Prime Minister had announced that India was open defecation free and that every household had a toilet based on the administrative data. But then came the National Family Health Survey and it showed that this wasn’t the case and that 30% of the households don’t have toilets,” points out Sen.
So the Census is needed for a reality check on administrative data instead of the latter being used as a replacement for the census.

Census enumerators at a village in Delhi in 1981. (Credit: Census of India)

If, for instance, data from a housing scheme claims that all the intended beneficiaries have received houses, it is only by conducting a census that one can be sure. A confirmation means the housing scheme has to end and if the census shows gaps, then it has to be continued.

How the delay hurts

Given that there are no real reliable alternatives to a census, how does a delay affect India?

• Data and the statistical system

Census is the foundation for the entire statistical system of the country. Sen explains that all the household surveys that are undertaken in the country draw their sample based on the census data. “When you are drawing a sample, you try to draw one which is as representative of the actual population — the picture of which is provided by the census — as possible,” says Sen.

However, if the census is delayed, the original picture is distorted and, as such, samples based on it will also fail to be representative of the reality. The same logic applies to the results of a survey. Ordinarily one could project the survey results to the whole population but with delays one runs into problems. “So in the absence of a relatively recent census, the quality of survey data also becomes suspect,” says Sen.

Desai says that in the short-term, projections based on the last census will be reasonably accurate but as the years roll on, a few metrics start getting out of whack.

The first metric where the data starts to falter is rural-urban population distribution. “We must use census town definition as against the statutory town definition because as things stand, there is a lot of political incentive for people to stay ‘rural’ that villages really don’t want to convert themselves into towns,” says Desai.

Vallabhbhai Patel, then Deputy PM, addressing a Census meeting in 1950. (Credit: Census of India)

Second, internal migration. Typically, after each Census, the Census Commissioner also produces annual population projections for the next two decades. “(However), population projections are based on estimates of only fertility and mortality data. On migration, we never have reliable projections because that is one that changes quite rapidly,” says Desai.

Contrary to popular imagination, most of India’s international migration happens intra-state and intra-district. This has massive implications for policymaking whether it is in reference to how much housing is needed in different parts of the country or how the tax revenues are divided between states.

The third metric that suffers in the absence of a timely census is mortality data. Desai says that while the National Family Health Surveys give “very good quality” fertility data, we do not have good data on mortality. Unless we have a good understanding of mortality data, our age-distribution projections could suffer significantly.

“India is clearly going to face an ageing challenge. Lower age groups are affected by fertility while the upper ages are affected by mortality. So unless we have a good understanding of the mortality data and consequently the age distribution, we will not be able to frame an accurate policy response such as assessing the extent of old-age pension requirements etc.,” says Desai.

• Policymaking

The lag in census data has already started claiming its victims.
Himanshu, professor of economics at JNU, says that almost all the major policies are decided on the basis of census data — be it the number of Primary Health Centres and their location or the allocation of subsidised foodgrains under the National Food Security Act (NFSA).

“The worst example of what a delay in census can do is to look at how genuine beneficiaries are being excluded under the NFSA,” he says. Since the Act specifies a ratio of population that needs to get the benefits (up to 75% of the rural population and 50 per cent of the urban population), the use of a dated population base has meant that as many as 100 million Indians are getting excluded from NFSA.

Rajendran Narayanan, who teaches at Azim Premji University and has traveled extensively across rural India to study the efficacy of government programmes, provides a very real example of how the delay in Census results hurts some of the weakest citizens. His example ties in with the mortality data problem that Desai pointed to as well as the policy implementation problem that Himanshu mentioned.

“Consider a household of a man and woman that was not eligible for a ration card as per the 2011 Census. There is evidence to show that in India, life expectancy of men is lower than that of women. If the man dies, that leaves the woman as the lone bread winner of the house. If she is successful in getting all the nods, then, under the National Social Assistance Programme, she will get a widow pension of Rs 200 per month. Getting her included in the list of ration card holders will be difficult because targets would be set based on the 2011 Census,” says Narayanan.

While the census may have started out just as a population count, and even now is often seen as a central government exercise, its biggest contribution is not in providing a macro estimate, but in providing the exact and granular detail of India’s reality. For instance, a census can tell us how many people in a particular village of a particular district in a particular state were disabled? How many had a roof over their heads? What kind of roof was it — mud, cement or something else? How many were looking for a job? How many had migrated and why? How many still use wood as fuel to cook their meals? How many get clean drinking water?

All such metrics not only matter for framing policies but also in assessing whether policies have been effective or not.

• Private sector

As significant as it is for public policy, it would be a mistake to think that the private sector does not benefit from the census.
For instance, one sector which is most affected, according to Sen, is the insurance industry. “All life insurance is based on Census data. Moreover, almost all market studies are based on census because it provides you the context,” says Sen.

Sen, in fact, calls the census a public good. Census 2021 was supposed to cost over Rs 8,750 crore of taxpayer money. In the absence of a census, the market surveys done by private firms will not only have to be bigger in size but also turn out to be significantly more costly.

Population projections as per Census

Rama Bijapurkar, long-time observer and researcher of Consumer India and author of We Are Like That Only and A Never Before World, echoes the sentiment on costs. “I think there is a severe lack of rigorously collected, large-scale private data because it is so expensive to collect. Just look at the price tag on the census. In heterogeneous India, you can’t get away with small samples and few locations for proper estimation,” she says.

She also points to the role the census plays as a benchmark for such private surveys. Often, she says, a private firm would claim their data is good data because it looks similar to what the census has thrown up.

An enumerator collecting information from then PM Jawahar Lal Nehru, 1961. (Credit: Census of India)

On the practical application of what the census provides, Bijapurkar says that India is becoming increasingly heterogeneous and for companies, micro-marketing is crucial. Micro-marketing is the private sector’s version of targeted subsidies. It essentially involves targeted marketing to specific groups of people. The granular data of the census provides the basis for this. “If you read the chairman’s speeches of Nestle and Hindustan Lever in recent AGMs, they talk about winning in many Indias. India is increasingly being viewed as a collection of distinct micro-markets,” says Bijapurkar, who has taught at IIM Ahmedabad as well as served on the boards of several of India’s blue chip companies.

Nikhil Gupta, chief economist of Motilal Oswal Financial Services, says a lot of decisions are based on census numbers. Data on migration to urban areas, for instance, gives a good understanding of the region-wise infrastructure needs in the country. Secondly, census provides a good understanding of penetration of consumer durables (such as two-wheelers, cars and TVs etc.) in different regions. “For instance, according to census 2011, the penetration of four-wheelers was pretty low, barely 10%, and it told us that there was huge potential in this sector,” says Gupta.

Concern over delay

There is considerable concern among policy experts and observers when it comes to the delay in India’s census. For one, not having a timely census undermines the comparability of data that was collected every 10 years.

Moreover, many believe that a timely census would not only have addressed the repeatedly questioning India’s GDP data faces but also provided the government’s claims — that it has brought about widespread prosperity in the country — much needed credibility.

Even now, for some, linking the Census to processes such as delimitation and legislation like women reservation is deeply problematic.

“I think that is a very dangerous thing that we will be doing. Ideally no census should not be linked with anything. If you announce beforehand that we want to have the census because we want to do the delimitation exercise, then it will completely spoil and vitiate the quality of the census data. That’s because there will be a tendency to over-report population. That will be dangerous. Ideally, the Census should be absolutely neutral,” said one of India’s top demographers who spoke on condition of anonymity.