In spite of the central relevance of infrastructure in the development policy debate, there has been little rigorous research on the impact of large scale infrastructure investments in very poor economies. It is usually difficult to disentangle the causal impact of infrastructure on economic outcomes as infrastructure is often built in the fastest growing areas of a country. For example, governments often select for improvement roads that connect areas with perceived high economic potential. Estimating the impact of roads by comparing changes in agricultural development in areas with roads and those without, therefore, is likely to be biased upward.
In this project, we analyze how rural feeder roads rehabilitation may affect markets and trading behavior. The design of a particular road construction program funded by the European Union (EU) in Sierra Leone allowed us to cleanly identify some of these impacts. The EU selected roads to be reconstructed by first establishing a list of roads eligible for rehabilitation throughout four districts, and then creating a priority ranking for each road based on an index of quantitative economic data (such as economic output and population density in the local area). Funding was allocated toward the road projects according to this index, starting with the road with the highest value, until approximately 150km of roads were selected in each district. This method allows a rigorous impact evaluation of the roads rehabilitation program through a regression discontinuity (RD) framework.
Distance and access to markets have been shown to be a severe constraint for investment by farmers in Sub-Saharan Africa. The cost of agricultural inputs is extremely high in these economies, partly because of high transport costs due to poor infrastructure. Sierra Leone, in particular, is a country with low density rural population where infrastructure investment may be crucial to development. Agriculture is the country's largest employer, with 64% of households farming in 2008. However, that same year only 11% of farming households reported selling any sizeable fraction of their output, and imports of rice are on the rise. Sierra Leone has particularly good land for cultivation of cash crops such as coffee, cacao, and palm oil, yet high transport costs continue to deter investment from the private sector.
The rehabilitation program implemented by the EU involved creating motorable roads, filling pot holes, installing culverts and often constructing bridges on the selected feeder roads. By surveying farmers and traders and collecting high frequency data on prices along roads that were just above and just below each of the 4 district cutoffs for rehabilitation by the EU we will be able to assess the extent to which improved roads: made markets more integrated, increased access to and knowledge of improved agricultural technologies, increased investment in these technologies by farmers, lowered transport costs of traders, increased competition amongst traders and thus increased prices for crops received by farmers, improved access to agricultural storage, increased investment in cash crops, increased prices and profitability of farmers. Of which have the potential to increase the profitability of both farmers’ and traders’ economic activity. Investments by farmers in agricultural technologies (such as improved seed and fertilizer) could also have larger productivity gains. By assessing the magnitude of these effects this project will contribute to our understanding of whether the substantial cost of investment in road infrastructure is justified by the benefits accruing to farmers.