One main aim of this project is to maintain and improve the Land Matrix Global Observatory, a global open database on large-scale land acquisitions. As part of the Land Matrix Initiative (LMI), we provide basic analyses of the Land Matrix Global Observatory’s data.The Land Matrix database is a constantly updated data set that includes deals made for agricultural production (for food or agrofuel production), timber extraction, carbon-trading, mineral extraction, conservation and tourism. Deals included in the database must meet the following criteria:* They entail a transfer of rights to use, control or own land through sale, lease or concession.* They were signed sometime since 2000, when the annualised value of the FAO real food price index was at its lowest level.* They cover an area of 200 hectares or more.* They entail the conversion of land from local community use or from important ecosystem-service provision to commercial production.The Land Matrix Initiative is presently involved in a process of decentralisation and expansion of its network. As a result, regional focal points support the LMI on regional level data collection, research, policy advocacy and networking and communication. Records are derived from a variety of sources: media reports; reports by international and local organisations, NGOs, and field-based research projects; company websites; and government records. Moreover, the "crowdsourcing" function plays an increasingly important role.The project also intends to zoom into specific countries and investment cases. At the country-level, we aim to assess socio-economic impacts of large-scale land acquisitions by linking data from the Land Matrix to national household surveys (or similar micro data, for example on the farm-level). This approach will not only allow to relate the presence of large-scale investments to welfare outcomes of nearby households, but also to shed light on the transmission channels through which these investments may (or may not) affect rural livelihoods, in general, and smallholders, specifically. Methodologically, we will rely on standard impact evaluation techniques, in particular differences-in-differences estimates at a small geographical scale. Finally, the project complements the quantitative assessments by qualitative case studies of individual investment cases. Specifically, we intend to identify and document cases that are exemplary in certain respects, e.g. cases that maintain out-grower schemes.**Results**We maintain the Land Matrix Global Observatory; data and the most recent newsletter with data analysis can be accessed here: [http://landmatrix.org/en/](http://landmatrix.org/en/).Around the world, 26.7 million hectares of agricultural land have been transferred into the hands of foreign investors since the year 2000. This means that these investors possess approximately 2 per cent of the arable land worldwide, or roughly the equivalent to the total area covered by United Kingdom and Slovenia together. This finding comes from Land Matrix's report "Land Matrix Analytical Report II: International Land Deals for Agriculture”.The report provides detailed information on who is buying up farmland in which regions of the world and how this land is being used. It also highlights the economic, social, and political impacts of land investments. One of the main conclusions is that land deals are increasingly being implemented while the land targeted has often been used before, mainly for agricultural activities, pastoralism and forestry. As land acquisitions only rarely take place on idle land, they can potentially have serious implications for people living on the land or using it.Findings from our research in Zambia confirm this: land targeted by investor in Zambia is by no means "idle" but in close proximity to small-scale farmers. Hence, the question how local communities are affected by the presence of commercial farms is crucial. For instance, we find that smallholders in regions with large-scale farms increase their area cultivated and maize yields, but have lower fertiliser usage. This hints at positive spillovers at the extensive and intensive margins but not at improved access to agricultural inputs. It is likely that these results are also driven by the emergence of medium‐scale farms in these regions. Our case studies show the challenges of setting up a commercial farm in a poor rural setting: High expectations (on all sides) are hard to fulfil, especially in areas marked by a lack of infrastructure, low educational levels and poor land governance. Continuous communication with local communities is a prerequisite for projects to be successful under such circumstances. Moreover, we have analysed the employment implications from large-scale agricultural investments (LSAIs). Our analysis shows that LSAIs massively crowd out smallholder farmers, which is only partially mitigated through the cultivation of labor intensive crops and the application of contract farming schemes. This holds true for all regions targeted by LSAIs, although regional differences are found in terms of magnitude. These effects tend to be large on the local scale (i.e. in the immediate surroundings of the investment site) but small in relation to total national employment in agriculture.In our research published in the Journal of Economic Geography, we used Land Matrix data to show that the determinants of foreign land investments only partly overlap with the determinants of other forms of foreign direct investment (FDI). For example, rich investors target economies with abundant land and water resources. In contrast to other forms of FDI, high‑quality institutions do not attract land investment.
