Household surveys omit the most vulnerable populations by design. Only considering households means that CRVS data on those living outside traditional household settings – including undocumented nationals, refugees or migrants and those living in institutions – is simply missing.

Timeliness is also an important issue. USAID’s Demographic and Health Survey (DHS) and UNICEF’s Multiple Indicators Cluster Surveys (MICS) are typically conducted once every three to five years and the microdata is usually released two years thereafter. With progress being made on digital registration as outlined in Chapter 3, household survey data is not timely enough to inform planning and target resources.

In addition to the missing populations, the survey data as it currently stands does not provide the level of detail required to inform the planning and implementation of CRVS systems; track the progress towards digitisation and interoperability with other foundational systems; or provide the detailed subnational information for health and local government administrators to plan and target their interventions more accurately and in a cost-effective manner.

Survey data also lacks detail on the timeliness of the registration itself, as this information is not captured by the standard questions used on surveys. For example, standard survey results show little difference between the birth registration rates of boys and girls as they are based on a question which asks whether children aged under five within the households have been registered. Recent research from Mexico highlighted a gender bias against girls, with the registration of girls more likely to be delayed until school age, while boys are typically registered nearer to birth. This is important as delayed registration is associated with incomplete, incorrect and unreliable information.^[1]

Another issue is that the two main global datasets on birth registration provide substantially different results and it is not clear which one is more authoritative. Recent analysis by Development Initiatives (DI) on Africa^[2] considered these two databases, one maintained by UNICEF^[3] and one by the UN Statistics Division.^[4] Over one third (35%) of countries had discrepancies between the two systems. Some of these were the result of different sources being used, but a number had different values recorded, despite coming from the same source.

There are also significant concerns about the quality of the data generated by questions from household surveys. For example, the household head may not understand the difference between birth notification and birth registration or whether the birth certificate is linked to a legal identity. In this instance, important information about whether the birth is recorded in a national database or whether the paper certificate is the only lifetime proof of identity that the child has would not be captured.

Given these issues, it seems appropriate to question our reliance on standard household survey data to monitor birth registration coverage.

“While data coverage has improved over time, this has largely been the result of investment in collecting data on birth registration in low- and middle-income countries through household surveys. In fact, in the absence of reliable administrative records, household surveys have become a key source of data to monitor levels and trends in birth registration. While it is clear that household surveys provide a rich source of valuable data, there are a number of problems if we continue to rely on surveys to monitor our progress towards complete civil registration coverage.” - UNICEF, 2019.^[5]

Most developing countries are now making steady progress in their development of national registries and administrative data. These systems, by their nature, contain a built-in monitoring framework. In this new monitoring framework, all performance statistics should be sourced wherever possible from cleaned, aggregated counts of records stored within the national CRVS and ID registries, or from considered, documented estimates produced by senior officials within the registration agencies. Monitoring the experience of detailed implementation would provide valuable global lessons on how to reach those being left behind. It would free up resources so that household surveys could be used to collect data that cannot be captured elsewhere and ensure that monitoring is owned, maintained and accounted for by those responsible for securing sustainable data infrastructures.

In October 2019, DI proposed a new sustainable and accountable monitoring framework that is governed by the implementing countries.^[7] Reporting directly from the CRVS system itself as implementation occurs, it would track actual registrations in a timely manner as well as building capability to produce vital statistics. Data from household surveys and national household and population censuses should continue to be used for quality assurance and sense checking of the new data sources.

As more and more countries adopt national databases to manage their registries the aggregation of performance indicators becomes an automated process, making timely, annual reporting a feasible option. Annual reporting will not only shine a light on progress within a timescale that is meaningful to policymakers, but over time it will ensure that exaggerated and inaccurate reporting is either eradicated or made transparent to those seeking to hold the registration agencies to account.

With statistics being drawn from the primary data source, national registration agencies – supported by national statistics offices – should be empowered to maintain both ownership of and accountability for the monitoring process. National registration agencies should be responsible for the timely and accurate validation and production of the required statistics. National statistics offices should be responsible for coordinating, collating and reviewing the data received from their partner agencies.

Table 2 shows a simple annual country monitoring framework proposed by DI. It contains six questions relating to the development of CRVS and ID systems, and their interoperability. The framework aggregates data from these systems in order to track a range of indicators.

There is a tendency within the development community to focus on the use of data solely for monitoring and evaluation. This approach often overlooks the fact that data is needed more urgently in order to inform action, so that the Sustainable Development Goals (SDGs) are met by 2030, not just monitored. Universal digital registration of births is not only about identifying dispossessed communities that are being left behind – it is about empowering those communities through better access to services.

The findings in Chapter 2 illustrate the complexities of registering those furthest left behind. For legal identity to have an impact on vulnerable communities, it needs to result in improved access to services. From a data point of view, improved service delivery requires the three foundational information systems – education, health and civil registration – to be embedded in every school, clinic and registry office and for the data collected and the point of service delivery to be shared as usable information in the same locality (see Figure 11).