Administrative data

This module is intended to serve as a reminder for the NSS to include administrative data among the priority considerations in the preparation of the NSDS. Suggested concrete actions are provided with reference to the relevant specific steps in the NSDS Lifecycyle.

Some examples of administrative data sources:

Administrative data   has been a key subject in statistical development forums for decades even though some administrative data systems have been in existence and are being used for much longer. Civil registration   and vital statistics systems  , population registers  , and statistical business registers as well as the educational management information systems   and health management information systems   are major administrative data systems that have been around and used in many countries for many years and are continuously being improved and expanded. 

There is also a variety of administrative data produced by the private sector and civil society organizations some of which NSOs use to supplement existing data or even substitute for data that is not complete or is not available. Many of private sector and CSO data constitute ‘big data’ such as credit data from banks, call records and internet data from telecom companies, and e-commerce data from private businesses, as well as sales and consumer trends data from industry associations, monitoring data from CSOs, among others. 

While in the past, administrative data is generally discussed in the context of being a support or an alternative to traditional censuses and surveys, it is increasingly seen as a primary, cost-effective and sustainable source of statistics in the NSS and a huge one. 

Increased need for administrative data:

Lessons learned from the MDGs include the need to strengthen administrative data systems, a lesson well amplified in the SDGs as many of the indicators to monitor goals and targets depend largely on availability of good quality administrative data in the countries. Similarly, planning and monitoring other national development goals typically rely on a considerable number of data that are mainly or solely generated from administrative and regulatory sources.

With rising costs of large-scale censuses and surveys and diminishing resources for statistics in the international community and NSSs, there is increasing impetus for NSSs to improve administrative data systems. Increasing rates of nonresponse is a general problem for NSSs and contributes to high costs and quality issues for many of the traditional surveys.  Strengthening administrative data systems to produce more and better quality administrative data for statistical purposes is a necessary strategy for NSSs.

Assessment of administrative data systems: 

It is important to assess the benefits and challenges   of improving and investing in the production and use of administrative data. While human, financial and technological resources and capacity are essential considerations, prioritization of sectors and subject matters as focal development areas is key to ensuring concrete results and strategic benefits for the NSS. 

Concrete actions

  • Assess existing or established systems in selected priority sectors and/or subject-matter areas (e.g., education, health, agriculture, CRVS, SBR, PopReg, etc.). Step 3.1 | Step 3.2 | Step 3.3 | Step 3.4
    • Explore and assess yet untapped data sources (e.g., sensor data, citizen-generated data, routine reporting and monitoring data).
  • Identify strategic goals and key outputs toward the improvement of administrative data. Step 4.2 | Step 4.3
  • Identify specific actions and corresponding costs, as well as key risk factors and mitigating measures. Step 5.1 | Step 5.2 | Step 5.3
  • Establish institutional partnership mechanisms between NSO and the ministries/agencies mandated in the selected priority sectors and/or subject-matter areas. Step 6.3
  • Consider partnerships with big data sources such as data science firms, telecommunication companies, and other service providers to share data and/or help develop applications to extract data. 
  • Study and consider existing systems of civil societies (many are location-specific) including citizen-generated data to produce statistics on specific issues.