Research Brief
Aid for Statistics in Africa

In the WIDER Working Paper ‘Aid and Investment in Statistics for Africa’ Jeffery I. Round discusses how the effectiveness of aid aimed at improving statistical capacity in Africa can be assessed. He begins by describing the reasons behind the increasing demand for data, and the institutions involved in helping to provide it. He then moves on to look at current approaches to measuring the effectiveness of aid invested in statistics, and suggests avenues for future research.

Aid and statistics

© Flickr / F H MiraThe increasing focus on results-based research and aid has led to an ever-increasing need for accurate statistical information on key indicators from the developing countries. Round highlights three major initiatives which have led to a higher demand for good statistics on sub-Saharan Africa, which is the region in focus in his paper. First the Millennium Development Goals needed improved statistics to be able to monitor progress towards their achievement. Second the introduction of poverty-reduction strategies as a prerequisite for World Bank and IMF assistance caused an increase in demand as previously unavailable data was required to be produced by them. Third the launch of the International Comparison Programme for Africa required the collection of large amounts of data in order to generate purchasing power parity price comparisons across African countries.

These increases in demands for data have led to a number of international organizations getting involved in programmes aimed at improving the statistical capacity of Africa. Perhaps key amongst these is the Partnership in Statistics for Development in the 21st Century, PARIS21 for short. It was set up by the UN, OECD, IMF, and the World Bank to foster a dialogue between those who demand and use statistics and those who are responsible for their production. PARIS21 has been involved in a large number of aid-funded programmes aimed at improving statistical capacity in all global regions. Other key institutions include the UN Economic Commission for Africa, AFRISTAT, the African Development Bank, the United Nations Statistics Division, and the World Bank Development Data Group. Round’s main question is: how can we measure whether the involvement of these development institutions has actually improved the statistical capacity of countries in Africa?

Approaches to measuring the effectiveness of aid to statistics

It is essential to the aid-effectiveness agenda to also have some way of assessing improvements in statistical capacity in a given country. However, Round points out, there is no universally accepted definition. Statistical capacity can be looked at either an input or output perspective.

  1. Logical framework approach: Measuring outcomes

Evaluations of aid programmes often draw on the logical framework approach, which either implicitly or explicitly use a logframe matrix. The logframe matrix has a dual logic. One part of the matrix looks at the process from input to output, output to outcomes, and outcomes to ultimate goals. The other looks at how each stage of the process can be accessed.

Round suggests that the logframe matrix can be usefully applied to the case of evaluating statistical aid, what is needed to identify the causal mechanisms through which aid can help improve the outcome of statistical activities. Inputs to the process will include things such as human resources, finances, and equipment. Outputs can be measured by looking at the production of data and information. These outputs should then have positive outcomes such as improved formulation and moderation of policy. These outcomes will then have impacts on things like growth, wellbeing, and the environment.

© Flickr / F H MiraThe World Bank Development Data Group (DECDG) is one institution that has adopted an output-based approach to measuring statistical capacity. Round outlines some of the results of the DECDG’s output approach which looks at three composite indicators; statistical methodology, source date, and periodicity and timeliness. The research shows that the overall statistical capacity indicator in sub-Saharan Africa has increased from 55 to 59 between 2004 and 2011; this is compared to a global increase from 64 to 68.

  1.  Measuring statistical capacity through inputs

The input perspective measures the increase in statistical capacity due to aid by assessing the resources a country has that can be used to generate statistics. The input indicators that have been suggested for use by PARIS21 are:

  • human resources (technical, administrative, support, plus data-producing agency staff) infrastructure
  • human resources management practices
  • finance
  • computing facilities
  • transport, communication and office equipment
  • statistical practices and the regulatory framework

Clearly having information about all the different types of resources used to generate statistics will give us a good idea of the statistical capacity of a given country. However Round points to the limitations, as well as the benefits, of both input and output approaches to measuring statistical capacity.

  1. Benefits and limitations of the two approaches

The benefit of the output approach is that it uses cheap and readily available data. However, it does have its limitations. In particular the output-based approach has been criticized because it does not tell as why statistical capacity has increased, and because it does not differentiate between an increase in statistical capacity and increased use of previously unused capacity.

The benefit of the input approach is that it avoids the problems associated with the output method and tells us exactly what statistical resources are available in a given country. However, collecting the data necessary to assess statistical capacity in this way is both expensive and time-consuming, and reporting on statistical capacity draws on the time of staff members who are needed elsewhere.

Round finishes by suggesting that further work needs to be done into determining the best approach to measuring statistical capacity. He suggests that the difficulties of measuring the effectiveness of aid aimed at improving statistical capacity should not put us off further enquiry. There is increasing demand for more and more data, understanding how effective aid is in achieving this if we are to appropriately balance aid spending in this versus other areas in Africa.