In developed countries poverty and its effects are counted by national and local government departments using data collected in the course of the planning, executing and monitoring of their day-to-day work. This is known as administrative data and is collected on a daily basis in the course of government’s work and is routinely turned into statistics for reporting purposes. National censuses, usually conducted every ten years, are used as key points of reference. Periodic surveys are used to sense check the data and to fill in gaps – on attitudes and behaviours, for example.
Few developing countries currently have the financial, technical and human resources to maintain all the administrative data they need. Civil registration is a case in point: many countries have rudimentary systems for the issuance of birth and death certificates (if people ask for them) but few are capable of turning these paper certificates into comprehensive and meaningful statistics that can be used for the planning and allocation of resources. National statistics offices, supported by the World Bank, USAID and other donors, currently employ a range of surveys as a substitute, the most common being the ‘household survey’. These are conducted using a representative sample of households spread across large statistical regions. In Uganda, for example (which has one of Africa’s best national statistical systems) about 1 in 1,000 households are surveyed in 10 regions each comprising 10 or more districts – the cornerstone of local government. District officials, responsible for service delivery at the cutting edge, can thus make little use of survey results as they need data that is disaggregated to a level below their area of responsibility.
The aim of the much heralded ‘Data Revolution for Sustainable Development’ is to ensure that not only is data available to meet (not just monitor) the Sustainable Development Goals, but also to ensure that sustainable data sources are established to ensure that in future – by 2030? – the digital divide between developing and developed countries is removed and the world shares common statistical methodologies. For this to happen two key requirements need to be met: firstly substantial domestic and international resources will be required to build, improve and maintain management information systems across government; secondly a strategy that recognises both short-term statistical needs (using existing survey-dominated methodologies) as well as a long-term, sustainable solution.
The World Bank Group is actively involved in this revolution and has earmarked $100m for investment in Civil Registration and is seeking to raise a similar amount for Statistical Capacity Development. Their commitment to the Global Partnership for Sustainable Development Data stated that “the WBG will support improvements to country systems by building on what works, ensuring country policy priorities determine where additional support is targeted by helping to maximize the value of what data is collected by making it accessible and usable.”
It is against this backdrop that we need to assess the World Bank’s announcement to raise and spend $1,5 billion ($300m every three years until 2030) on household surveys. This single commitment dwarves all other investments that have been made towards building a base for sustainable data collection and usage. It sends a signal that recurrent expenditure producing short-term results to meet the needs of global goals is preferable to capital investments producing resilient country information systems that will have a life way beyond 2030. A raid of this scale into the pot of limited resources available to the data revolution is short-sighted and unsustainable.