A survey and analysis of the data requirements for stakeholders in African agriculture

As part of DFID's contribution to the G8 initiative on Open Data for Agriculture, a broad survey of key stakeholders in sustainable African agriculture was conducted to assess current and emerging trends related to data collection, processing, and dissemination. Stakeholders that promote and support sustainable intensification of agriculture in Africa require access to useful data upon which to base their decisions and evaluate current and future interventions amid limited resources. Information has value for supporting a decision only if it reduces the chance of being wrong and a cost of being wrong. Research across many fields has shown that quantitative decision analysis methods overwhelmingly outperform expert judgment in identifying the economic value of information and improving decisions. Therefore a key focus of the study was to assess the alignment of stakeholders’ perceived data needs with areas of decision uncertainty. A total of 281 stakeholders were contacted by email among 11 organizational categories. Of this list, there were 110 respondents to the online survey with 58 individuals further contributing to in-depth conversations. Results from the online survey were compiled along with live interview responses into a centralized database for analysis. In addition to searching for overall keyword trends and tendencies within groups, we examined motivations for data and use, and whether these data were informing specific decisions. Less than half of respondents (46%) could specify a decision of any kind in relation to their perceived data needs. Only 36% of respondents stated data needs that were consistent with their stated uncertainties and only 15% showed that perceived needs, uncertainties, and data gathering efforts are aligned. There was broad alignment among effort, perceived needs, and uncertainties for soil data. In other words, soil was the most frequently cited uncertainty, the most frequently stated perceived need, and the most frequently stated focus of current effort. Market data showed similar high priority and alignment. Climate data is frequently cited as both needed and satisfying an uncertainty, but is less frequently cited as a focus of current effort. Biodiversity and poverty data are frequently cited as a focus of effort but infrequently cited as a perceived need or uncertainty. Consistent with other case studies, overall there is evidence of a "measurement inversion" where decisions are either poorly defined or data priorities are disconnected from the decisions they potentially inform. Based on the survey results and analysis, we provide recommendations for improving the collection and use of data in African agriculture. Comprehensive, centralized web-enabled GIS databases could have large impact on improving decisions if efforts are prioritized by information values. Databases and information gathering requirements aimed at development impact should be based on information values identified by quantitative modeling of key decisions rather than by routine, intuition, or purely subjective means. Initiatives to develop awareness of the key decisions and what data is needed to support them should be widespread and routine. These will inform decision makers and researchers on how to spend their limited resources measuring only the most important information or variables. Researchers should also make greater use of decision analysis techniques that explicitly handle uncertain or incomplete data. This may help optimize the use of existing data for improving specific development decisions.