The socio-economic impacts of natural resource concessions in Liberia

The Ellen Johnson-Sirleaf administration has made natural resource concessions a key part of Liberia’s economic growth strategy. It has granted somewhere between 21% and 38% of the country’s land to investors in the hopes that they will stimulate economic activity in the agriculture and mining sectors, increase employment in rural areas, and build and maintain infrastructure. However, it remains unclear if this FDI-led growth strategy produces broad-based economic development benefits. There is also debate about which types of FDI (e.g. agriculture versus mining) yield the greatest development dividends. Another key question is whether the socio-economic effects of FDI depend on investor characteristics. Chinese investors, for example, have been criticised for importing foreign workers rather than drawing upon local labor, paying low wages, and making little effort to build domestic supply chains by prioritising local suppliers in public procurement.

In this project, we will use a novel sub-national dataset to answer two questions. First, do locations in which the Liberian government has granted concessions to foreign investors experience faster rates of economic development than areas where concessions are not present? To answer this question, we will utilize satellite data on nighttime lights and household survey data to measure development outcomes in locations with and without extractive sector FDI. Second, do concession and concessionaire attributes differentially affect economic development? To test our hypotheses, we have geocoded all available concession contracts the Liberian government has awarded since the end of the civil war and will merge these data with survey- and satellite-based outcome and covariate data at the 1km x 1km grid cell level. We will then use a novel matching propensity score matching approach that accounts for spatial autocorrelation, as spillover effects might bias the analysis if the treatment and control units are located next to each other. Using this quasi-experimental methodology, we will compare development gains in locations with and without concessions, in locations with different types of concessions (e.g. agriculture versus mining), and in locations with different types of concessionnaires (e.g. Western vs. non-Western).

The proposed study will build upon a collaborative data collection effort by AidData and the Concessions Working Group, a TrustAfrica-facilitated network of civil society organizations and researchers in Liberia that are engaged in efforts to monitor the activities of natural resource concessionaires. AidData and the Concessions Working Group have worked together over the last eighteen months to build a georeferenced database of all known natural resource concessions in Liberia. With support from the IGC, AidData and the Concessions Working Group will disseminate their study findings to in-country stakeholders, including the Office of the President, the Ministry of Finance and Development Planning, the National Bureau of Concessions, the Inter-Ministerial Concessions Committee, and the Ministry of Lands, Mines, and Energy