The geography of large-scale land acquisitions: Analysing socio-ecological patterns of target contexts in the global South

TLDR

The study examines local geographic contexts of large‑scale land acquisitions to link socio‑ecological patterns with investment processes, reflects on current methodologies, and calls for new empirical approaches. They analyzed 139 high‑accuracy georeferenced deals, assessing land cover, population density, accessibility, and agricultural potential, identified three distinct patterns consistent with stakeholder competition, and compared local‑scale idle‑land analysis to prior country‑level studies. The analysis revealed three patterns—densely populated croplands, remote forestlands, and moderately populated shrub/grasslands—while challenging earlier idle‑land claims and suggesting that better linkage methods could enhance policy relevance. The authors emphasize the need to improve large georeferenced land‑deal datasets and carefully match spatial scale to data resolution.

Abstract

This paper analyses local geographical contexts targeted by transnational large-scale land acquisitions (>200 ha per deal) in order to understand how emerging patterns of socio-ecological characteristics can be related to processes of large-scale foreign investment in land. Using a sample of 139 land deals georeferenced with high spatial accuracy, we first analyse their target contexts in terms of land cover, population density, accessibility, and indicators for agricultural potential. Three distinct patterns emerge from the analysis: densely populated and easily accessible croplands (35% of land deals); remote forestlands with lower population densities (34% of land deals); and moderately populated and moderately accessible shrub- or grasslands (26% of land deals). These patterns are consistent with processes described in the relevant case study literature, and they each involve distinct types of stakeholders and associated competition over land. We then repeat the often-cited analysis that postulates a link between land investments and target countries with abundant so-called “idle” or “marginal” lands as measured by yield gap and available suitable but uncultivated land; our methods differ from the earlier approach, however, in that we examine local context (10-km radius) rather than countries as a whole. The results show that earlier findings are disputable in terms of concepts, methods, and contents. Further, we reflect on methodologies for exploring linkages between socio-ecological patterns and land investment processes. Improving and enhancing large datasets of georeferenced land deals is an important next step; at the same time, careful choice of the spatial scale of analysis is crucial for ensuring compatibility between the spatial accuracy of land deal locations and the resolution of available geospatial data layers. Finally, we argue that new approaches and methods must be developed to empirically link socio-ecological patterns in target contexts to key determinants of land investment processes. This would help to improve the validity and the reach of our findings as an input for evidence-informed policy debates.