Every single person matters: traffic in Hanoi.
The World Bank’s operations are guided by two goals:
- eliminate extreme poverty by 2030 and
- foster income growth for the bottom 40 % of every country.
The Bank’s governors adopted these goals in the autumn of 2013. In a recent policy research report, the Bank has now assessed how to measure whether progress towards these goals is being made. The basic message is that statistics must improve.
The two goals are interrelated, but pose different statistical challenges. As the authors point out, it is impossible to totally eradicate extreme poverty, since there will always be some poor people who suffer in unusual circumstances. Moreover, poverty is said to be too “deep and widespread” in some countries to be eradicated by 2030. The Bank’s goal is thus to reduce the share of the extreme poor to three percent of the world population. This goal is deemed to be ambitious as well as achievable.
To judge success, however, the experts need reliable data on the size of the world population and poverty in every country. Obviously, similar problems haunt the UN Millennium Development Goals (MDGs) and will haunt the Sustainable Development Goals that are being drafted as the follow-up to the MDGs. As the World Bank authors argue, however, statistical services tend to be weak in countries that are strongly affected by poverty, and this is particularly true of countries that are rocked by violence. Unless national data are good, however, global data cannot be trustworthy.
To measure whether the lot of the bottom 40 % of a country is improving, one does not need global data. It is essential, however, to have data that is compiled consistently over time. If survey methodologies and questionnaires are modified too often, the World Bank warns, the results cannot be compared with one another. If, for example, farmers are asked about their situation immediately after they have harvested, their responses differ considerably from what they say in the lean season. If poor people are asked what they have eaten in the past seven days they will normally indicate more food than when asked about the past two weeks, simply because their memory is more precise.
The World Bank authors therefore suggest that surveys should only be modified very carefully. However, they state clearly that statistics should serve national purposes first of all, so any income and consumption surveys should fit the given country. The authors insist that governments need solid data to draft and implement policies prudently. The World Bank does not want surveys to be harmonised simply for the sake of better international comparability.
The World Bank publication argues convincingly that surveys, censuses and economics statistics on things such as growth or inflation are essential for monitoring development progress. It also discusses methods statisticians use to cope with data gaps. However, such methods always add up to some kind of informed guessing and do not result in entirely convincing substitutes for empirical data.
The report emphasises that recent trends will not continue for ever. Healthy development, for instance, may in the future be disrupted by economic crisis, political instability, civil war, environmental change or an epidemic. Nonetheless, it makes sense to improve statistics today. Doing so helps to understand what is happening, and the deeper that knowledge is, the better future change can be assessed too.
The World Bank publication tackles an important technocratic issue. Unfortunately, it is not well edited. There is too much jargon, too much convoluted syntax and too much showy rhetoric. An author may feel smart to state, for instance, that clarity of data is “a key quality to crystallising political traction around goals” (p. 114), but phrases like this are really very vague.