Summary

There is clearly an existing problem with development data provision. For example, even population figures may be uncertain because of undercounting of some groups in society, and changes to the statistical basis of GDP estimates can make large differences. Ghana became a middle-income country overnight when its estimated GDP doubled in this way.

Food and Agriculture Organization (FAO) data on undernutrition – on which many poorer countries rely in the absence of official national statistics – has also been shown to be prone to significant errors.

This matters because the post-2015 agenda is likely to have a far greater monitoring burden, and Jerven has shown convincingly that paying for a full set of development data is not feasible. The second issue is that, even if it was available, throwing money at the problem will not solve it. The problem of lack of statistical capacity is hard to solve, and the chapter also points out the opportunity cost. The Partnership in Statistics for Development in the Twenty-First Century (PARIS21) group estimates that there has been an increase in data-gathering exercises in Africa because of the MDGs, but a shift away from surveys not closely geared to them. As Jerven correctly argues, macroeconomic, labor, and agricultural statistics have suffered in particular.

This chapter is a groundbreaking attempt to delineate the issue, map the extent of the problem, and make recommendations; essentially it is a wake-up call to the development community. It makes several important contributions, the most important of which is its careful enumeration of the costs of monitoring theMDG indicators. This quantification is extremely powerful in showing the need for prioritization of targets and indicators – and so is highly complementary to the Copenhagen Consensus exercise.

Because there is not yet a definitive list of targets and indicators, a precise costing is impossible, but the chapter usefully lays out the likely minimumdata requirements. However, the cost of data analysis and utilization is excluded from these estimates, and this is, of course, a necessary requirement for the information to be debated by politicians and the public.