Africa Infrastructure Country Diagnostic (AICD) Working Paper 19
Paul Dorosh, Hyoung-Gun Wang, Liang You, and Emily Schmidt
“This study adopts a cross-sectional spatial approach to examine the impact of transport infrastructure on agriculture in Sub-Saharan Africa using new data obtained from geographic information systems (GIS). Our approach involves descriptive statistical analysis and econometric regressions of crop production or choice of technology for each location (a 9×9 kilometer pixel) in Sub-Saharan Africa on (a) agroecological zones and crop production potentials by the Food and Agriculture Organization (FAO) and the International Institute for Applied Systems Analysis (IIASA), (b) GIS data on crop production from the International Food Policy Research Institute’s (IFPRI) spatial crop allocation model (SPAM), and (c) road infrastructure based largely on data from the United Nations Environment Programme (UNEP) and estimated travel times.
“We address three main issues. First, we analyze the impact of road connectivity on crop production and choice of technology when we control basic supply and demand factors. Second, we investigate the impact on agricultural output of investments that reduce travel time on roads of various types. Third, we provide an example of how this type of analysis could be used to construct benefit-cost ratios of alternative road investments in terms of enhanced agricultural output per dollar invested.
“We find that agricultural production and proximity (as measured by travel time) to urban markets are highly correlated, even after taking agroecology into account. Likewise, adoption of highproductive/ high-input technology is negatively correlated with travel time to urban centers.
“There is substantial scope for increasing agricultural production in Sub-Saharan Africa, particularly in more remote areas. Total crop production relative to potential production is 45 percent for areas within four hours’ travel time from a city of 100,000 people. In contrast, it is just 5 percent for areas more than eight hours away. These differences in actual versus potential production reflect the relatively small share of land cultivated out of total arable land in more remote areas.
“For remote regions, low population densities and long travel times to urban centers sharply constrain production. Reducing transport costs (travel time) to these areas would expand the feasible market size for these regions, easing the constraint on production. If the expansion in production from these areas were small in terms of the relevant regional, national, or subnational market, average market prices outside the formerly remote region would be unaffected, and significant aggregate production increases could result.
“We find some interesting differences between East Africa and West Africa. On average, East Africa has lower population density, smaller local markets, and lower road connectivity—the average travel time to the nearest city is more than twice that in West Africa. While average suitable area for crop production is similar in East and West Africa, average crop production per pixel in East Africa is just 30 percent of that in West Africa. Road connectivity has different impacts in the two regions. In East Africa the results are similar as for all Sub-Saharan Africa. Longer travel time decreases total crop production, and reducing travel time significantly increases adoption of high-input/high-yield technology in East Africa, but the impacts are insignificant for West Africa. This may be because the more densely roads are connected, the smaller the marginal benefits of more connections. West Africa already has a relatively well-connected road network.”