Population Data: UNFPA's Role in Census and CRVS Support

EXECUTIVE SUMMARY

UNFPA is the leading multilateral supporter of national population and housing censuses and civil registration and vital statistics (CRVS) systems in developing countries — a role that is less visible than its programme work but arguably among its highest-leverage investments. Census and CRVS systems are the foundation of all population-based planning: without them, governments cannot accurately know how many people they serve, where those people live, how many are being born and dying, or what is killing them. For SRHR programming specifically, functioning population data systems are preconditions for measuring maternal mortality, tracking fertility change, identifying unmet contraceptive need, and monitoring progress toward SDG health targets.

An estimated 1 billion people globally lack any legal identity — they have never been registered at birth, hold no documentation, and are invisible to government planning and legal systems. This "invisible people" problem disproportionately affects women and girls, indigenous communities, migrants, and stateless populations — precisely the groups whose SRH needs are greatest and whose deaths are most likely to go uncounted. Solving the invisible people problem is not primarily a data problem; it is a civil registration problem that UNFPA addresses through systematic CRVS support.

The 2020 census round (2015–2024) was severely disrupted by the COVID-19 pandemic: more than 100 countries that planned to conduct censuses in 2020 or 2021 were forced to delay. This has created a significant data gap that will affect planning through the 2030s, making post-COVID census support an urgent UNFPA priority. At the same time, new data approaches — satellite imagery, mobile data, administrative data integration — are offering partial alternatives to traditional census methods that UNFPA is beginning to integrate into its data support portfolio.

For frontline programme staff, this document explains what UNFPA's data work means for programme design and targeting. For funders, it makes the case for census and CRVS investment as high-leverage development financing. For researchers, it identifies the significant methodological debates around census alternatives and CRVS quality that shape the reliability of global health statistics.

KEY FACTS

UNFPA is the UN system's primary supporter of national population censuses in developing countries, providing technical and financial support to over 100 countries across successive census rounds.
An estimated 1 billion people globally lack any legal identity; approximately 166 million children under 5 are unregistered at birth (UNICEF 2023 data).
In sub-Saharan Africa, only approximately 46% of births and fewer than 30% of deaths are officially registered — meaning the majority of demographic vital events go unrecorded.
The 2020 census round was the most disrupted in history: more than 100 countries postponed censuses planned for 2020–2021 due to COVID-19, creating a data gap that affects planning accuracy for a decade.
Maternal mortality can only be accurately measured if deaths are recorded with cause of death; the wide uncertainty ranges on WHO maternal mortality estimates for low-income countries (sometimes ±100%) reflect poor CRVS coverage.
A girl without a birth certificate cannot legally prove her age if challenged about child marriage — meaning birth registration is a concrete legal protection mechanism, not an administrative formality.
The UN recommends a census every ten years; approximately 40 developing countries have not conducted a census in the past fifteen years due to cost, conflict, or political sensitivity.
UNFPA has committed to supporting 100% of programme countries with population data collection through its 2022–2025 Strategic Plan — a more explicit data commitment than in previous strategic plans.
The Demographic and Health Survey (DHS) Programme — the primary alternative data source to census/CRVS for reproductive health statistics — is funded primarily by USAID with UNFPA support, covering approximately 90 developing countries.
Civil registration reform — moving from paper-based to digital registration systems — is a major focus of UNFPA's current CRVS support, potentially enabling real-time population monitoring between census rounds.
Stillbirths are among the most underregistered vital events globally: an estimated 2 million stillbirths per year are not registered, making them invisible in national statistics and in SDG monitoring (SDG 3.2 tracks neonatal mortality but not stillbirths separately).
Remote and nomadic populations, stateless persons, and internally displaced people are systematically undercounted in standard census enumeration — requiring specific methodology adaptations that many national statistical offices lack.
Geographic Information System (GIS) technology is increasingly central to census support: UNFPA provides GIS support for census mapping, allowing enumeration areas to be defined, household listings to be georeferenced, and data to be disseminated in spatially disaggregated form.
The global cost of a typical developing-country census is USD 1–3 per capita — making a census for a country of 50 million approximately USD 50–150 million, an amount that many governments cannot finance without external support.
UNFPA co-chairs the UN Expert Group on Civil Registration and Vital Statistics with WHO — providing normative leadership on CRVS standards and practices.

BACKGROUND AND CONTEXT

Why Population Data Matters: The Policy Foundation

The relationship between population data quality and policy effectiveness is direct and consequential. Every major development planning decision — how many health facilities to build, where to place them, how many health workers to train, how to allocate education funding, how to target social protection — requires accurate data on population size, location, age distribution, and vital trends.

For SRHR specifically, the data dependencies are:

Measuring progress: The SDG maternal mortality target (SDG 3.1: reduce MMR to below 70 per 100,000 live births) can only be tracked if countries have functioning civil registration that captures maternal deaths with cause of death. In sub-Saharan Africa, where most maternal deaths occur, the majority of deaths happen outside health facilities, are not registered, and do not enter any statistics system. WHO's maternal mortality estimates for these countries are derived from statistical modelling — useful for broad comparison but with uncertainty ranges that span hundreds of deaths per 100,000. You cannot achieve what you cannot measure.

Targeting programme resources: A programme to reduce unmet need for family planning in Nigeria cannot be targeted efficiently without knowing where unserved women of reproductive age actually are. Census data provides the geographic population distribution that makes efficient programme targeting possible. Without it, resources are allocated based on guesswork.

Detecting change: Demographic and Health Surveys (DHS) measure contraceptive prevalence, fertility rates, and maternal health coverage — but they are conducted only every five years, have large sample sizes, and provide national- and regional-level estimates, not community-level data. Census data, combined with DHS, enables more granular monitoring.

Understanding inequality: The most marginalised populations — the poorest, the most rural, ethnic minorities, indigenous communities — are precisely the groups most likely to be undercounted in weak census and CRVS systems. Without disaggregated data, SDG commitment to "leaving no one behind" cannot be operationalised.

The CRVS System Architecture

A functioning CRVS system consists of several interconnected components:

Civil registration: The legal registration of vital events (births, deaths, marriages, divorces). In a functioning system, every birth is registered by a civil registrar within days of occurrence, producing a birth certificate. Every death is registered with cause of death coded to the International Classification of Diseases (ICD-10 or later). Registration is universal, compulsory, and continuous.

Vital statistics compilation: Registered vital events are compiled by the national statistical office into vital statistics — annual birth rates, death rates, cause-specific mortality rates, total fertility rate. These statistics are the inputs to national health planning, SDG monitoring, and epidemiological surveillance.

Population register: In the most advanced systems, civil registration feeds into a continuously updated population register — a real-time count of the national population, available between census rounds.

Most developing country CRVS systems are severely deficient in at least one of these components. The most common failure modes: incomplete birth registration (children born outside facilities are not registered); incomplete death registration (deaths at home, especially in rural areas, are not registered); inadequate cause of death coding (deaths are registered but cause of death is missing or miscoded); non-digital systems that cannot be compiled or analysed efficiently.

The Census as Demographic Baseline

A population and housing census is a complete enumeration of a national population at a specific point in time. Well-conducted censuses provide:

Total population count, disaggregated by age, sex, and geographic location.
Household composition, enabling analysis of dependency ratios, female-headed households, and household-level poverty proxies.
Education and literacy levels, providing human capital baseline data.
Housing conditions, providing proxies for poverty and access to basic services.
Geographic distribution, enabling facility planning and service mapping.

Censuses are the foundational demographic baseline from which all subsequent population estimates and projections are calculated. Population projections (such as those used for SDG monitoring) are extrapolations from the most recent census; the further in time from the census, the greater the uncertainty in projections.

WHAT UNFPA DOES: PROGRAMME DETAIL

Census Support

Pre-census technical assistance:

Census design and methodology: UNFPA provides technical assistance for the design of census questionnaires (ensuring inclusion of key topics relevant to development monitoring, including disability, ethnicity, migration status, and household characteristics), sampling methodology for any sample-based components, and enumeration area delineation.
GIS mapping: Census enumeration requires accurate maps of every enumeration area — defined geographic units of approximately 100–200 households that a single enumerator can cover in a given period. UNFPA provides GIS support for creating and digitising these maps, often using satellite imagery and existing geographic data. GIS-based enumeration has replaced manual cartography in most UNFPA-supported census operations.
Census pilot testing: Full pilot censuses in sample areas test questionnaire content, enumerator training, logistics, and data processing before the national enumeration.
Enumerator and supervisor training: UNFPA supports design of training programmes for the tens of thousands of census enumerators and their supervisors typically deployed in a national census.

Census implementation:

Operational logistics support: In countries with the most limited capacity, UNFPA may provide direct operational support — helping with procurement of census materials (questionnaires, tablets, GPS devices), coordination of enumeration logistics, and data processing support.
Financial support: In some country contexts, UNFPA provides partial financial support for census operations. This is less common than technical assistance but is used where the fiscal constraint on census conduct is most severe.
Post-enumeration surveys: After the census, post-enumeration surveys test coverage — whether all people were counted — and check for systematic undercounting of specific groups. UNFPA supports design and implementation of post-enumeration surveys.

Data processing and dissemination:

Data processing support: Supporting data entry, verification, and cleaning for paper-based census data; supporting data integration and quality checking for tablet-based digital enumeration.
Data analysis and dissemination: Supporting preparation of census analytical reports — not just tabulation of results, but analysis of demographic trends, spatial distribution of population, and implications for planning.
Micro-data release: Supporting governments to release census micro-data (anonymised individual-level data) for research use, in line with statistical confidentiality standards.

CRVS Support

National CRVS assessment and reform planning: UNFPA supports governments to assess the current state of their CRVS systems — completeness, quality, coverage, and capacity — and to develop national CRVS reform plans with specific targets and investment requirements.

Birth registration support:

Facility-based registration: Supporting integration of birth registration into maternity facility routines — ensuring that every baby born in a health facility is registered before discharge. This requires coordination between health and civil registration systems (typically managed by different ministries) and training of facility staff.
Community outreach registration: For births outside facilities (still the majority in many countries), community health workers and mobile registration units are the primary tools. UNFPA supports community registration campaigns and ongoing community-based registration systems.
Late registration: For older children and adults who have never been registered, UNFPA supports catch-up registration campaigns — mass registration events, mobile units in underserved areas, and simplified procedures for late registration.

Death registration and cause of death improvement:

Training of civil registrars: Supporting national systems to train civil registrars in accurate death registration, including the legal and procedural requirements.
Health worker training in cause of death coding: The quality of vital statistics depends on accurate cause-of-death coding. UNFPA supports training of health workers and verbal autopsy surveyors in ICD-10 cause-of-death coding.
Verbal autopsy: For deaths occurring outside health facilities (particularly in rural areas), verbal autopsy — a standardised interview with family members about circumstances of death — is used to assign probable cause of death. UNFPA supports verbal autopsy implementation as a complement to formal death registration in settings with low facility delivery rates.

CRVS system digitalisation: UNFPA supports the transition from paper-based to digital CRVS systems — civil registration software that enables real-time registration, automated statistics compilation, and eventually integration with identity management and social protection systems. Digital CRVS systems have significant long-term cost advantages over paper systems and enable data quality monitoring that paper systems cannot support.

New Data Approaches

UNFPA is actively integrating new data approaches into its data support portfolio:

Satellite imagery for population estimation: High-resolution satellite imagery, combined with machine learning algorithms that identify settlements and estimate population density, can produce population estimates in areas where census coverage is poor. UNFPA has piloted this approach in humanitarian settings and in countries with data gaps, working with partners including WorldPop at the University of Southampton.

Mobile data for population mobility: Anonymised mobile phone data can provide real-time information on population movement — useful for tracking displacement, understanding commuting and migration patterns, and monitoring population change between census rounds. UNFPA is exploring partnerships with mobile network operators for this data in programme countries.

Administrative data integration: In countries with strong administrative data systems (tax records, school enrollment, health facility data, social protection registries), integrating these sources can supplement census and CRVS data. UNFPA supports national statistical offices to develop administrative data integration frameworks.

Geospatial modelling of health outcomes: Combining DHS survey data with satellite-derived covariates (vegetation, nighttime lights, distance to facilities) through Bayesian geostatistical models can produce small-area estimates of health indicators (maternal mortality, contraceptive prevalence) below the level of DHS geographic stratum. UNFPA supports use of these methods by national statistical offices.

THE EVIDENCE BASE

Evidence on Why Census and CRVS Quality Matter

The uncertainty range problem in maternal mortality: WHO's maternal mortality estimates for sub-Saharan Africa countries demonstrate the consequence of poor CRVS coverage. Country-level MMR estimates for high-mortality countries often have uncertainty ranges of ±100–200 deaths per 100,000 live births — a range so wide as to make the estimate of limited value for programme planning or accountability. Countries with strong CRVS systems have narrow uncertainty ranges because the data comes from registration counts; countries with weak systems have wide ranges because the data comes from statistical modelling with uncertain inputs.

DHS versus CRVS data quality comparison: Multiple comparative studies (Hill K et al., 2007; Alkema L et al., 2016) have compared maternal mortality estimates derived from DHS data (sisterhood method) with estimates from vital registration systems. These studies consistently find that countries with strong vital registration systems produce more accurate, consistent maternal mortality estimates than countries relying on survey-based methods. The implication: investing in CRVS produces better data quality than investing in more survey rounds.

The GDP correlation with statistical capacity: Jerven M's research on African economic statistics (Poor Numbers, 2013, and subsequent work) documents the systematic underestimation of economic output in countries with weak statistical systems. The same methodological arguments apply to population and health statistics: the headline statistics used for global rankings and aid allocation are significantly affected by the quality of national statistical systems, independent of underlying realities.

Evidence on UNFPA's Census Support Effectiveness

What the evidence shows: Country programme evaluations that have assessed UNFPA's census support contributions generally find that UNFPA-supported censuses are conducted on schedule (when external factors don't intervene), achieve reasonable population coverage (post-enumeration surveys typically find undercounts of 3–7% in UNFPA-supported censuses, which is within acceptable ranges for developing country censuses), and produce data that governments and international agencies use for planning.

Limitations of the evidence: No rigorous controlled evaluation of UNFPA's census support has been conducted — this is not feasible given that census support is a country-specific, one-time activity. The evidence is based on process assessment (was the census conducted to methodological standards?) rather than outcome assessment (did the census improve government planning?). The link between census data availability and improved development outcomes has not been rigorously evaluated in any context.

Quality assessment: Evidence for the value of population data for development planning is strong in principle (multiple studies document the correlation between statistical capacity and development outcomes). Evidence that UNFPA's specific census support investments improve statistical capacity at country level is moderate, based on process evaluations and programme monitoring data.

Evidence on Birth Registration as Child Protection

The evidence that birth registration provides concrete legal protection is well-established in principle but harder to demonstrate empirically:

A birth certificate prevents verification of a child's age at marriage; without a birth certificate, a child cannot challenge a marriage by proving they are underage. Multiple UNICEF country case studies document this protection mechanism.
Birth registration is required for school enrollment in many countries — unregistered children are excluded from formal education.
Unregistered individuals cannot access formal health insurance, social protection, or legal rights that require proof of identity.

The counterfactual — would child marriage or school exclusion be lower in communities with higher birth registration rates? — has been assessed in a small number of studies with positive findings (Duflo E et al. on identity documentation and girls' outcomes in contexts where registration affects service access), but the evidence base is thin.

IMPLEMENTATION REALITIES

Political Sensitivity of Census Data

Census operations are politically sensitive in many country contexts. Population counts determine allocation of legislative seats, government transfers to subnational governments, and resource distribution. Groups that believe they are undercounted lobby for methodological adjustments; groups that believe they are overcounted argue the opposite. In several countries, census results have been disputed, contested in courts, or rejected by governments — most notably:

Nigeria (2006): The Nigerian census results were disputed by multiple states who questioned the enumeration methodology; the data was not published in final form for years.
Multiple Sahel countries: Ethnic tensions have made specific population counts politically sensitive, with some groups refusing to participate in enumeration or contesting results.
Post-conflict contexts: In countries emerging from conflict, census enumeration of displaced populations, returnees, and contested-territory populations creates significant technical and political challenges.

UNFPA's role in these contested contexts requires diplomatic sensitivity alongside technical capacity. UNFPA must maintain technical integrity while managing relationships with governments and population groups that have conflicting interests in census outcomes.

The COVID-19 Census Gap

The disruption to the 2020 census round is the most significant data challenge UNFPA faces in the current period. More than 100 countries that planned to conduct censuses in 2020 or 2021 were forced to delay, some for multiple years. Countries that had already invested in pre-census preparation (mapping, questionnaire design, pilot testing) lost that investment; enumerators trained before COVID-19 needed to be retrained; and data gaps widened.

The consequence for SDG monitoring: population denominators for SDG indicator calculations are becoming increasingly uncertain in countries that have delayed censuses. Maternal mortality ratio estimates, for example, require accurate estimates of live births (denominator); without current census data and registration data, both the numerator and denominator of MMR calculations are estimates with growing uncertainty.

UNFPA's response has been to: accelerate support for countries to conduct delayed censuses; explore temporary use of alternative data sources (DHS, modelled estimates) for planning purposes; and advocate for increased funding for the 2025–2034 census round that will follow the disrupted 2020 round.

CRVS Reform as a Long-Term Investment

Civil registration system reform is not a project — it is a generational investment in state capacity. Moving a country from 30% birth registration to 90% birth registration takes fifteen to twenty years of sustained investment in legal frameworks, administrative systems, community awareness, and staff capacity. The returns on this investment compound over time as data quality improves and as increasingly reliable population statistics enable better planning.

The challenge is that CRVS reform is difficult to fund through project-based development assistance, which operates on three to five year cycles. UNFPA's advocacy for long-term, sustained CRVS investment — ideally through domestic government budget commitments rather than external project funding — reflects this structural challenge.

FUNDING, SCALE AND RESOURCES

UNFPA's Data Investment

UNFPA does not publish a single budget line for its data work (census + CRVS + data systems support). Estimates suggest total investment of approximately USD 40–70 million per year, including contributions to the DHS Programme, direct country census support, CRVS programme activities, and data capacity development. This is one of UNFPA's more significant investments on a per-dollar-of-development-impact basis, though it receives less visibility than UNFPA's direct service delivery programmes.

Comparative Value of Census and CRVS Investment

The development economics literature suggests that statistical capacity investments — national statistical systems, surveys, censuses — have among the highest returns of any government institutional investment. The ability to accurately measure population size, distribution, and vital events enables resource allocation efficiencies that dwarf the cost of the statistical investment. Jerven's research suggests that misallocation of development resources due to poor population data costs hundreds of millions of dollars annually across developing countries.

For donors, census and CRVS support is an unusually high-leverage investment: a USD 5–10 million contribution to a national census that costs USD 50–150 million total can be the catalytic contribution that determines whether the census happens at all. The development impact of that census — in better resource allocation over the following decade — is almost certain to exceed the investment by orders of magnitude.

KEY DEBATES AND CONTESTED QUESTIONS

1. Traditional Census Versus Modelled Alternatives

The traditional population census — complete enumeration every ten years — is expensive, logistically demanding, and (as COVID demonstrated) fragile. A growing body of statistical and data science research argues that modelled alternatives — combining satellite imagery, mobile data, DHS survey extrapolation, and administrative data — can produce population estimates of sufficient accuracy for planning purposes at much lower cost and on a more continuous basis.

UNFPA and national statistical offices have been cautious about embracing these alternatives, arguing that modelled estimates cannot substitute for the legal and definitional functions of a census (which produces legally authoritative population counts for electoral and administrative purposes). The debate is partly technical (how accurate are modelled alternatives?) and partly institutional (who owns and validates modelled population estimates?).

The emerging consensus is that modelled approaches should complement rather than substitute for census enumeration — providing between-census estimates and coverage checks, while the decennial census remains the baseline.

2. How to Handle Political Manipulation of Census Data

When census results are politically contested or governments attempt to manipulate enumeration to serve political interests, UNFPA faces a dilemma: technical integrity versus diplomatic relationship maintenance. In several country contexts, UNFPA has had to decide whether to publicly challenge census methodology or data quality, or to work quietly through technical channels while accepting results that do not meet methodological standards.

UNFPA's formal position is that it will not endorse census results that do not meet international methodological standards. In practice, this commitment is tested in contexts where the government controlling the census operation is also a major UNFPA programme partner.

3. DHS Versus CRVS as the Primary Data Investment

A resource allocation debate: given limited funding for population data in developing countries, should investment priority go to strengthening CRVS systems (slow, expensive, long-term return) or to maintaining and expanding DHS surveys (faster data, lower per-indicator cost, immediate usability)?

The mainstream position, endorsed by UNFPA and WHO, is that CRVS systems are the long-term investment and DHS fills the gap while CRVS develops. The argument against over-reliance on DHS: DHS data is collected by an external USAID-funded programme, creating dependency on external funding; DHS sample sizes do not support small-area estimation needed for local planning; and DHS cannot substitute for vital registration in providing continuous demographic surveillance.

4. Privacy and Data Sovereignty in New Data Approaches

The use of satellite imagery, mobile data, and administrative data integration for population estimation raises significant privacy and data sovereignty concerns. Mobile data is personally sensitive; aggregated anonymisation may not fully protect individual privacy. And the use of private sector data (mobile operators, digital platforms) for government population estimation creates dependency relationships and potential conflicts of interest.

UNFPA has committed to privacy-by-design principles in its new data work, but the governance frameworks for satellite and mobile data use in population estimation are not yet settled.

IMPLICATIONS BY AUDIENCE

For Frontline Staff and Practitioners

Population data directly affects your programme targeting. If you are designing a programme to reduce maternal mortality in a specific district, you need to know: how many women of reproductive age are in that district, how many deliveries occur per year, what is the proportion occurring in facilities, and what is the current skilled birth attendance rate. These questions are answerable only if the national statistical system — census and DHS — has been maintained.

If you find yourself working with out-of-date or missing data — a census that is ten or more years old, or no recent DHS survey for your district — advocate for data investment in your programme planning. A programme without accurate population data cannot be targeted efficiently or evaluated meaningfully.

For CRVS specifically: health facility staff have a critical role in birth and death registration. Ensure that your facility has a functioning system for registering births at delivery — the form should be completed before mother and baby are discharged. Where deaths occur at your facility, ensure cause of death is recorded using ICD coding standards. This is a legal requirement in most countries, but it is also a public health responsibility.

For Programme Managers and Decision-Makers

Use subnational data wherever available. National DHS data provides useful context but cannot guide programme targeting below the provincial level. Push for additional data investment — mini-DHS surveys, facility assessments, and local population registers — to enable community-level programme targeting in priority areas.

When programme evaluations use national-level MMR or fertility data as the primary outcome measure, apply appropriate caution: UNFPA's country programme cannot be attributed responsibility for national-level indicator trends, and the data quality of these indicators in many programme countries is insufficient to detect programme-level change. Invest in programme-specific data collection that can provide more sensitive outcome measures.

Engage with national statistical offices as programme partners, not just data suppliers. In many programme countries, the national statistical office is an important ally for health data quality improvement — particularly for integrating health management information systems with civil registration data.

For Donors and Board Directors

Census and CRVS investment is among the highest-leverage development investments available. The data produced by well-functioning statistical systems multiplies the effectiveness of all other development investments — by enabling efficient resource allocation, accurate accountability monitoring, and evidence-based policy. Donors who invest in statistical capacity are investing in the foundation of all development progress measurement.

The catalytic funding model is particularly applicable: UNFPA's contribution to a national census may be 10–20% of total census cost, but without it, the census may not happen at all. This leverage ratio — where a relatively small external contribution enables a much larger national investment — is unusual in development finance.

CRVS reform requires long-term funding commitments. Census support is episodic (every ten years); CRVS reform is continuous. Donors who are serious about improving population data quality in programme countries should consider multi-year CRVS programme funding (five to ten years) that aligns with the time horizon needed for system reform, rather than short-term project grants.

For Researchers

The population data landscape is changing rapidly with new data sources and methods. Key research directions:

Validation of alternative population estimation methods: How accurate are satellite-based, mobile-data-based, and modelled population estimates compared to census counts? Under what conditions do they provide adequate accuracy for planning purposes? Systematic validation studies using countries with both census and alternative estimates are needed.
CRVS quality and maternal mortality measurement: The relationship between CRVS completeness and the accuracy of maternal mortality estimates has been theorised but not systematically evaluated. Studies comparing maternal mortality measured through complete CRVS versus survey-based methods in the same populations would clarify how much data quality improvement translates into estimate quality improvement.
Political economy of census data: Under what conditions do governments manipulate census enumeration or data processing for political purposes? What UNFPA programme design features reduce this risk? This is an under-researched governance question.
Birth registration and child outcomes: Does birth registration improve development outcomes for girls (protection from child marriage, school enrollment) in causal analysis? The cross-sectional association is established; causal identification is less robust. Natural experiments from birth registration campaigns would enable stronger inference.
Small-area estimation for SRHR planning: How can DHS survey data, combined with satellite covariates and geographic modelling, produce reliable small-area SRHR indicators (unmet need, MMR, contraceptive prevalence) at community or district level? Methods are developing rapidly but validation against ground-truth data is limited.

CURRENT STATUS AND FUTURE DIRECTIONS

UNFPA's data work is a stated priority in the 2022–2025 Strategic Plan, with more explicit commitment to population data support than in previous plans. UNFPA has developed a specific data strategy — emphasising census support, CRVS reform, DHS programme co-investment, and new data approach integration — that is being operationalised in country programmes.

The priority challenge is the COVID-19 census backlog. Approximately 40 countries are conducting delayed censuses in 2023–2025; UNFPA is mobilising technical and financial support to help these countries complete their delayed enumerations. The quality of these delayed censuses will be critical for SDG monitoring accuracy.

Investment in digital CRVS systems is growing. Several programme countries have implemented or are implementing national civil registration platforms that enable real-time birth and death registration from health facilities. Where these systems are functioning, they represent a major improvement in data timeliness and completeness. The challenge is maintaining these systems beyond the initial investment — CRVS digitalisation requires ongoing technical support and IT infrastructure maintenance that many governments struggle to fund.

SOURCES

UNFPA (2022): UNFPA Population Data Strategy. Internal document outlining UNFPA's approach to census, CRVS, and new data support. Referenced in country programme planning documentation.

UNICEF (2023): Birth Registration Data. Annual update of global birth registration coverage. Available at data.unicef.org.

UN DESA (2023): World Population Prospects. The primary global population data source. Available at population.un.org.

WHO (2023): Trends in Maternal Mortality 2000–2020. The definitive source for global and country-level maternal mortality estimates. Documents uncertainty ranges that reflect CRVS quality. Available at who.int.

DHS Programme: All country DHS surveys available at dhsprogram.com. The most comprehensive source of country-level reproductive health data; the primary complement to census and CRVS data for SRHR programming.

Jerven M (2013): Poor Numbers: How We Are Misled by African Development Statistics and What to Do About It. Cornell University Press. Critical analysis of statistical quality in Africa; important context for understanding why CRVS investment matters.

Mikkelsen L et al. (2015): "A Global Assessment of Civil Registration and Vital Statistics Systems." The Lancet. Systematic assessment of CRVS quality globally; the primary academic source on CRVS completeness and its implications for health statistics.

Hill K et al. (2007): "Estimating Maternal Mortality at the Country Level Using Survey Data." International Journal of Epidemiology. Methodological analysis of survey versus registration-based maternal mortality estimation; demonstrates the data quality implications of CRVS coverage.

CRVS Digitisation Institute: Technical resources for CRVS system reform and digitalisation. Available at crvs.org.

WorldPop Project: Technical documentation of satellite-based population estimation methodology. Available at worldpop.org. The primary technical reference for UNFPA's satellite imagery-based data work.