Abstract

FAO is the agency responsible for 21 SDG indicators and, as such, is responsible for, leading methodological improvements, coordinating the development and dissemination of methodological support documents, and strengthening countries' statistical capacity for the relevant indicators. FAO's Office of the Chief Statistician, which has been mandated to coordinate FAO's efforts in this area, has identified the Economic and Statistical Observatory of Sub-Saharan Africa (AFRISTAT) as one of the strategic partners in implementing methodological development activities and strengthening the statistical capacities of African countries in monitoring the SDGs.
In this context, AFRISTAT selected two African countries (Burkina Faso and Mali) to test the implementation of the calculation of indicators for the calculation of indicators 2.3.1 "volume of production per work unit " and 2.3.2 " average income of small food producers".
The data used for the computation of SDGs 2.3.1 and 2.3.2 indicators for Burkina Faso and Mali are microdata from their respective annual agricultural surveys (EPA and EAC). This microdata is collected annually on a regular basis for several years without interruption in both countries. Using this data in the calculation of SDG indicators is interesting from several points of view.
When the data needed to calculate SDG indicators 2.3.1 and 2.3.2 are available, the procedure for computation of the indicators is not really a challenge. Country experiences have highlighted the lack of important information in the data of agricultural surveys regularly conducted in the countries. While data on agricultural production, land area and income are well collected, the same is not true for information on labor input and cost of production. Proxy variables were used in some cases and imputations under certain assumptions were made in others.
The paper will present in detail all these challenges encountered as well as the solutions that have been provided to compute indicators that are fairly in line with the recommendations. Some methodological changes are proposed as well to overcome the issues regarding data availability. The paper will finally provide some guidance for the improvement of agricultural annual survey questionnaires.