A bed net is distributed in the forest and no one is there to track its delivery and use. Does it contribute to reducing the incidence of malaria?
This question gets to the heart of a key issue in development policy and one that is extremely important to Giving What We Can: data. And with the much-heralded ‘data revolution’ demanded by the UN High-Level Panel on the post-2015 development agenda, it seems to be rising up the list of priorities for governments and multilateral institutions. In this post I want to take a look at this ‘data revolution’ and why data are important for tackling global poverty.
But first, the ‘so what?’ question. Why is data important? The first reason comes from an efficiency argument central to Giving What We Can’s mission: you need to gather data to judge whether an intervention is having the effects it was supposed to have, and thus whether it is a success. What we really mean when we talk of investing in data collection is investing in effective school systems, in well-run hospitals or in successful nutrition programmes. That much is familiar. But this can also contribute to the democratic political process, by equipping people with the information they require to hold their leaders accountable.
Claire Melamed, of the Overseas Development Institute, makes a further, ‘equity’ argument for gathering data: we tend to have least information about the people who are most marginalised (e.g. slum dwellers, women, the disabled), which leads them to receive less attention in terms of policy responses**.If we had more information about these groups we would provide better services for them**. These are three compelling reasons for caring about data or evidence.
The problem is that the current state of data collection in many developing countries is extremely poor - so much so that Morten Jerven has written an entire book about the poverty of GDP statistics in Africa. And here we are not talking about minor mistakes: for example, when Nigeria’s GDP was re-based, the new estimate showed a 90% increase. And to give another example, the High-Level Panel’s briefing document on the data revolution states that every year around 50 million births go unregistered.
Given this situation, some sort of ‘revolution’ seems appropriate. But the term ‘data revolution,’ officially comprising ‘two main objectives’ and ‘four components’, is a slippery one. As Claire Melamed discusses on a Development Drums podcast on the subject, it holds different meanings for different people.
For some it means bringing ‘big data’ into the mainstream of measurement and policy. This could mean using mobile phone data to track refugee movements , or satellite imagery of trees to measure rates of deforestation .
For others, it means strengthening capacity in traditional government statistics agencies: training more statisticians, getting computer systems to work, etc. Perhaps, then, the ‘data revolution’ is simply an aspirational term, pushing policymakers to prioritise “more data, better data, disaggregated data, frequent data,” as Amanda Glassman of the Center for Global Development expresses it in the same podcast. It has certainly succeeded in drawing political attention.
For Glassman, one of the key issues is addressing the fundamental problem of incentives underlying the poor data. Administrative, government-collected data remain low-quality in many countries partly because of donor actions: they have demanded accurate household surveys to be able to evaluate their programmes, and to track indicators of their interest (e.g. the Millennium Development Goals). This has produced a system of little coherence and with certain types of information prioritised over others, for example at the national level at the expense of more data disaggregated into population groups.
Faced with this situation, one concrete proposal for donors is to engage in funding compacts with recipient-country governments to achieve certain standards on data, following the payment by results strategy discussed in an earlier blog, and with civil society oversight . This could perhaps be nested within what the High-Level Panel calls a ‘Global Partnership for Development Data’, which would also help set common global standards for data reporting, reducing fragmentation and double-collecting.
What else can be done here and now to improve the data we collect? Melamed identifies some such steps, for example ensuring that data are made available, particularly by publicly-funded bodies such as the UN, and also linking existing data together to create a baseline for 2015 by which we can judge progress on the Sustainable Development Goals, the successor to the Millennium Development Goals. These are both linked to the further point that the final packaging of data must be ‘actionable data’ i.e. visible, user-friendly and practical . The UK Department for International Development leads the way on this matter with its Development Tracker.
Whichever particular proposal you favour - making better use of ‘big data’ for development, forming ‘data compacts’ to improve statistics agencies in developing countries, standardising the way data are reported, or any other initiative - the key point is to take advantage of the current political moment to create a list of specific actions which can be taken forward even after attention has moved on from the data revolution (which it will), balancing the need to strengthen existing systems with the potential advantage to be gained from tapping into new and exciting data projects.
Giving What We Can puts gathering good data on an intervention at the heart of its philosophy. The ‘data revolution,’ if successfully translated into a series of concrete measures and integrated into the post-2015 development settlement, can reap the benefits of this approach on a global scale.
Image credit: CGDev