摘要：Two increasingly shared perspectives within the international development community are that (a) geography matters, and (b) many government interventions would be more successful if they were better targeted. This paper unites these two notions by exploring the opportunities for, and benefits of, bringing an explicitly spatial dimension to the tasks of formulating and evaluating agricultural development strategies. The paper was originally conceived to address the more specific goal of proposing a spatial characterization to underpin deliberations on appropriate development strategies for the “fragile” or “less-favored” lands of Sub-Saharan Africa. In practice, however, we considered that goal to be not only impractical but, perhaps, ill-conceived. The multiple senses in which land may be considered fragile, coupled with the myriad of potential development pathways would result in either an overly complex characterization or, more likely, a need to aggregate and generalize that would render the characterization of little use when confronted with any specific, real-world problem. We first review the lingua franca of land fragility and find it lacking in its capacity to describe the dynamic interface between the biophysical and socioeconomic factors that help shape rural development options. Subsequently, we propose a two-phased approach. First, development strategy options are characterized to identify the desirable ranges of conditions that would most favor successful strategy implementation. Second, those conditions exhibiting important spatial dependency – such as agricultural potential, population density, and access to infrastructure and markets – are matched against a similarly characterized, spatially-referenced (GIS) database. This process generates both spatial (map) and tabular representations of strategy-specific development domains. While there are many advantages to this tailored approach, it does depend on having access to a modest GIS capacity to re-characterize and re-interpret spatial datasets as the nature and focus of development problems change, and as new and improved data become available. This would be a significant step for many policy analysis units, typically run by economists. However, while acknowledging that not all aspects of strategic analysis necessarily benefit from a spatial perspective, we feel an important additional benefit of a spatial (GIS) framework is that it provides a powerful means of organizing and integrating a very diverse range of disciplinary and data inputs. At a more conceptual level we propose that it is the characterization of location, not the narrowly-focused characterization of land, that is more properly the focus of attention from a development perspective. IFPRI is expanding on these concepts in its work on policy-relevant applications of GIS linked more closely to economic perceptions of space. The paper includes appropriate examples of spatial analysis using data from East Africa and Burkina Faso, and concludes with an appendix describing and interpreting regional climate and soil data for Sub-Saharan Africa that was directly relevant to our original goal.
