Sub-Saharan Africa is typically regarded as land abundant, and previous efforts to estimate the trueextent of potentially available cropland (PAC) have largely affirmed this perception. Such efforts, how-ever, have largely focused on production potential and have underemphasized economic profitabilityand other constraints to expansion. This paper re-estimates PAC for Africa in a more explicit economicframework that emphasizes the returns to agricultural production under a variety of assumptions, usingrecent geospatial data. Existing PAC estimates for Africa are shown to be highly sensitive to assumptionsabout land productivity and market access, and are moderately influenced by the use of alternative datasources.

The region’s underutilized land resources are concentrated in relatively few countries, many ofwhich are fragile states. Between one-half and two-thirds of the region’s surplus land is currently underforest cover; conversion of forests to cropland would entail major environmental costs. Most of the con-tinent’s unexploited land resources are located far from input and output markets, limiting their eco-nomic attractiveness. In the long run, improvements in infrastructure and agricultural productivity andthe growth of hinterland towns will enhance the economic returns to cropland expansion. In the shortto medium term, however, the potential for profitable smallholder-based cropland expansion in mostAfrican countries is likely to be much more limited than it is typically perceived to be.

Sub-Saharan Africa has about 900 million inhabitants at present – a number expected to grow to 1.4 billion by 2030 (UN, 2013). Substantial increases in agricultural output will be required to meet the rising demand for food within the region. Such output growth will have to come from higher yields on existing cropland and/or expansion of land under cultivation. Over the past four decades, crop production growth has occurred predominantly through area expansion (Brink and Eva, 2009, Evenson and Gollin, 2003). While there would be distinct advantages if additional food supplies could be generated mainly through intensification of existing farmland, this may not be realistic. It is almost certain that African governments will face intensified political and economic pressures to allocate much of the region’s underutilized land for exploitation, including at the expense of the region’s forest land (Gibbs et al., 2010), to meet the growing food and energy needs of African cities and global markets.

The recent “land grab” phenomenon – driven largely by rising international food and energy prices – has sparked renewed interest in determining the true extent of under-utilized land in Africa. It is well understood that rural Africa is highly heterogeneous and that much of its land is either unutilized or underutilized even while a considerable fraction of its rural population resides in densely settled smallholder farming areas facing land shortages (Table 1; e.g., Tittonell and Giller, 2013; Jayne et al., 2014). Less well understood is how land transfer decisions made today will affect the viability of future agricultural development patterns in the coming decades, particularly the potential for cropland expansion under a smallholder-led development strategy. Unlike other developing regions, Africa’s rural population will continue to grow by almost 50% between 2015 and 2050 (UN, 2013).

Young, 1999, Young, 2000, Young, 2005 has been a vocal critic of such estimates, noting that they simply do not conform with many of the empirical regularities of smallholder agriculture in the region.

“If there is so much spare land, why has cultivation been so widely extended onto steep slopes, onto extremely infertile soils, and into semi-arid zones liable to frequent crop failure’? Why is there so much illegal incursion into forest reserves and national parks? Why has average farm size in some countries fallen below one hectare, and why are infertile soils which need rest periods cropped continuously? Why is there so much land degradation, widely attributed to the interaction of land shortage with population increase and poverty? Above all, if it were possible to bring further land under cultivation, why do 800 million people suffer from endemic undernutrition? Inequitable land distribution is certainly part of the answer, but these indicators are so widespread as to suggest that in many regions, the supply of available land is approaching zero”
Young (2000: 51).

On the basis of these observations, Young suggests that there is a systematic combination of overestimation of cultivable land, underestimation of land already cultivated, and/or underestimation of competing non-agricultural land uses.