ABSTRACT
As part of the PRSP process, the World Bank is responsible for offering guidance on Poverty and Social Impact Analysis (PSIA), designed to assess the impact of economic policies on poverty. The Bank proposes a detailed review of the ‘toolkits’ used by economists and social scientists, which it puts in correspondance with the reform policies recommended by the PRSPs (World Bank 2002e). This sort of exercise is perfectly understandable and legitimate in a national or multilateral bureaucratic context, as all modern economic administrations have their ‘methods’ departments. Nonetheless, it has the disadvantage of producing a ‘shopping list’ that fails to take into account either the concrete conditions for applying the proposed methods, or the hypothetical and experimental nature of many of the evaluation methods. Indeed, these two aspects are connected, as good practice in evaluation methods requires that a certain number of conditions should be met among the users. These conditions include not only technical skills but also, and perhaps above all, an institutional infrastructure such that the ‘evaluation messages’ can, first, be demanded, second, understood, and third, integrated in a wider decision-making process. However, in a large number of countries concerned by the PRSPs, the world’s poorest countries, the conditions for implementing participatory evaluation and debate on the policies are often far from being met. In particular, it is clear that politicians and administrative staff take little note of work carried out by their Statistics, Forecasting and Planning departments. It is also clear that the democratic requirement to justify political decisions is weak. Hence, the problem with PSIAs is not so much the result of technical difficulties with the evaluation methods or data availability, but rather of the conditions under which the results of the evaluations are received. An initial evaluation of the PRSP process by the World Bank and the IMF observed that: ‘Poverty and Social Impact Analysis (PSIA) of major policies and programs has not been undertaken as part of PRSPs’ (IDA and IMF 2002). It explains the situation by underlining the lack of qualified staff within the administrations, technical
difficulties and the lack of data, as it goes on to say: ‘for reasons to do with national capacity constraints and its inherent technical difficulties. The data and capacity needs are formidable and serious methodological issues remain unresolved, despite some analytical advances in this area.’ We believe that the focus on technical difficulties, or what we could call the ‘technocratic’ explanation, is exaggerated. On the other hand, it is fair to point out that the methods have not been perfected yet, but surely the same can be said for all scientific activities. In the end, the problem in adopting the methods comes back to the more general problem of democratic ownership of the reforms, a subject that is referred to in detail in other chapters of this book.
