Sources of Demographic Data in India: Census to Health Surveys

Ask someone how many people live in India, what the infant mortality rate is, or how the sex ratio has changed in the last decade, and you will get an answer almost instantly. But behind every one of those numbers sits a specific data collection system, built for a specific purpose, running on its own timeline. India does not rely on a single count. It runs six major, overlapping systems that together track everything from the size of the population to the number of women who received antenatal care last year.

Why one data source is never enough 🔗

Each source answers a different kind of question. The census counts every single person once every ten years. Civil registration tracks births and deaths as they happen, every single day. Sample surveys sit in between, giving quicker, more frequent estimates without counting the entire population. And specialised health surveys go deeper still, asking detailed questions about nutrition, family planning, and disease that a headcount exercise was never designed to capture. Understanding how these pieces fit together is the first step to reading Indian demographic data correctly.

The population census: counting every single person 🔗

The Census of India is the complete enumeration of the country’s population, carried out once every ten years, with every individual recorded separately across the entire national territory. The Census Division assists the Registrar General and Census Commissioner of India in conducting these operations under the legal authority of the Census Act, 1948. The last full census was conducted in February 2011, and the next one, originally due in 2021, was postponed because of the pandemic.

What the Census Act, 1948 actually does 🔗

The Census Act provides the legal backbone for the entire exercise. It empowers the central government to declare when a census will be taken, appoints census officers with legal authority to collect information, and guarantees confidentiality; individual responses cannot be used as evidence in a court of law. This legal protection is part of why the census achieves near-universal participation despite asking deeply personal questions about religion, caste, disability, and household conditions.

What makes the census different from every other source 🔗

Unlike every other data source on this list, the census is not a sample. It is a complete count, which makes it the foundation that other surveys build on. Sample surveys like the NSS, NFHS, and SRS all draw their sampling frames from the census, using the villages, towns, and enumeration blocks it maps out. Without the census, none of the sample-based systems would have a reliable population base to draw from.

Civil registration system: recording life events as they happen 🔗

While the census gives a ten-year snapshot, the Civil Registration System (CRS) works continuously. It is the unified, permanent, and compulsory recording of vital events, births, deaths, and still births, as they occur. This continuous recording is enabled by the Registration of Births and Deaths Act, 1969, which was enforced across most states from April 1970 and has since been strengthened by a 2023 amendment that makes it mandatory for medical institutions to issue a cause-of-death certificate and feeds data into a national digital database.

Why CRS is the backbone of public health planning 🔗

CRS data is not just administrative record-keeping. It generates the vital statistics that health departments, hospitals, and welfare schemes depend on, including the sex ratio at birth, infant mortality trends, and still birth rates. Because registration happens continuously, well-functioning CRS data can, in theory, replace the need for periodic sample surveys altogether. In practice, registration completeness still varies across states, which is one reason India continues to run a parallel sample-based check on births and deaths.

Sample registration system: bridging the gap between counts 🔗

That parallel check is the Sample Registration System (SRS). It was started on a pilot basis in 1964-65 and became fully operational in 1969-70, precisely because early civil registration and one-off recall surveys were both proving unreliable for estimating birth and death rates. The Ministry of Statistics and Programme Implementation’s own vital statistics manual notes that retrospective recall surveys from the 1950s and 60s consistently produced inaccurate birth and death rate estimates, which is what pushed the Registrar General’s office toward SRS as an alternative.

How the SRS actually collects its data 🔗

SRS runs on a dual-record method. Resident part-time enumerators, often teachers or Anganwadi workers, continuously record births and deaths in a fixed set of sample villages and urban blocks. Separately, SRS supervisors conduct an independent retrospective survey of the same areas every six months. The two records are then matched against each other, and any mismatch triggers a follow-up home visit to verify the correct figure. This cross-checking is what makes SRS estimates of the infant mortality rate, birth rate, and death rate more dependable than either method alone. The sample itself is refreshed roughly once every ten years, aligned to the latest census.

National sample survey: India’s ongoing socio-economic diary 🔗

Where SRS focuses narrowly on births and deaths, the National Sample Survey (NSS) covers almost everything else about how people live. Set up in 1950, it is a continuing, multi-subject survey programme that has, since its first round, gathered information across dozens of themes: household consumer expenditure, employment and unemployment, land holdings and livestock, migration, morbidity and disability, and family planning. The National Sample Survey Office, now merged into the National Statistical Office, organises this work into annual or six-month “rounds,” with the specific subject rotating according to government data priorities rather than staying fixed year to year.

What makes NSS particularly useful for demography is its flexibility. A single round might focus entirely on migration and fertility one year, and switch to household consumption the next. This rotating design lets the government track a wide range of social indicators without running a dozen separate survey machines.

National family health survey: the deep dive into health and nutrition 🔗

The National Family Health Survey (NFHS) takes a narrower but far deeper approach, focused specifically on population, health, and family welfare. The first round was conducted in 1992-93, and the International Institute for Population Sciences (IIPS), Mumbai, has served as the nodal agency for every round since, working with the Ministry of Health and Family Welfare. NFHS is modelled on the globally used Demographic and Health Surveys programme, which is part of why its indicators are comparable across countries.

How the survey has grown over three decades 🔗

Each round has expanded the scope of what NFHS measures: fertility, mortality, and family planning in the earliest rounds, followed later by reproductive health, women’s autonomy, domestic violence, HIV and AIDS awareness, and adolescent health. The most recent round, NFHS-5 (2019-21), covered 28 states, 8 union territories, and 707 districts, and, according to policy research organisation PRS India’s analysis of the data, recorded a national sex ratio of 1,020 women per 1,000 men, the highest of any NFHS round so far, alongside new questions on internet use and disability that had not featured in earlier surveys.

District level household surveys: zooming in to the district 🔗

Even NFHS could not always answer questions at a hyper-local level in its earlier rounds, which is the gap the District Level Household Survey (DLHS) was designed to fill. Conducted under the Ministry of Health and Family Welfare, with IIPS again playing a coordinating role, DLHS ran across four rounds between 1998 and 2014, building a reproductive and child health database at the district level to track progress under the country’s Reproductive and Child Health programme.

DLHS supplemented household interviews with facility surveys of sub-centres, primary health centres, community health centres, and district hospitals, giving planners a combined picture of both demand for health services and the ground-level capacity to deliver them. This made it especially valuable for targeting resources at the district rather than the state level, a scale at which the census and NSS rarely offer usable detail.

How these six sources fit together 🔗

None of these systems work in isolation. The census supplies the sampling frame that SRS, NSS, and NFHS all rely on to select representative households. CRS and SRS check each other’s numbers on births and deaths. NSS fills in the socio-economic picture that the census only touches on lightly. NFHS and DLHS then go further still, asking detailed health questions no other source covers. Layered together, they let planners move from a national headcount all the way down to a district-level health indicator, using the right tool for the right question.

What do you think? If India’s civil registration system eventually achieves complete, real-time coverage of every birth and death, would a sample-based system like SRS still be necessary? And which of these six sources do you think matters most for planning local health services in your own state?

References

1. https://censusindia.gov.in/census.website/en/node/378
2. https://dc.crsorgi.gov.in/crs/about
3. https://www.mospi.gov.in/sites/default/files/publication_reports/vital_statistics_2010.pdf
4. https://www.mospi.gov.in/national-sample-survey-office-nsso
5. https://iipsindia.ac.in/content/national-family-health-survey-nfhs-5-india-report
6. https://prsindia.org/policy/vital-stats/national-family-health-survey-5
7. https://iipsindia.ac.in/content/district-level-household-project