Agricultural yields in Sub-Saharan Africa are low and have remained stagnant for decades, even while yields in other regions like South Asia increased substantially following the Green Revolution. A primary explanation for this fact is that the adoption of technologies like chemical fertiliser is lower in Africa than in other parts of the developing world (World Bank 2008). For example, in our country of study of Tanzania, fertiliser usage was only 15% in 2012 (Tanzania National Bureau of Statistics, 2012). What factors explain such low usage?
One factor that is surely important is that road networks in Africa are sparse and transportation costs are high. Poor roads will discourage technology adoption in two ways: (1) by increasing the costs of inputs themselves; and (2) by decreasing the price farmers receive for their output. In addition to roads directly affecting prices by increasing costs to reach a buyer/seller, poor roads will have the additional effect of discouraging entry and reducing competition among intermediaries. While these problems have been widely documented in the prior literature, there is a fairly sparse literature which rigorously quantifies the welfare effects of poor roads through the lens of a trade model, especially in Africa (mostly due to a lack of data).
In this project, we seek to fill this data gap in a large number of rural markets (preliminarily we are planning to collect data in approximately 200 markets, serving a population of roughly 2.4 million Tanzanians). Our main proposed data collection includes:
- Farmer surveys: We plan to conduct surveys about input purchases, maize sales, prices, and other questions with several thousand farmers.
- Input and output price data collection: We plan to collect prices on a basket of goods (including fertiliser and maize) approximately every 2 weeks for about a year.
- Surveys with market actors: On the input side, we plan to interview fertiliser distributors and retailers. On the output side, we plan to collect surveys with maize intermediaries (called “agents”) and larger storage facilities in major markets which serve as hubs connecting areas (called “stores”).
- Driving distances and travel costs: We will collect information on distances between all relevant markets, and on travel costs faced by suppliers, intermediaries, farmers, and retailers.
At its broadest, our project seeks to understand how supply networks evolve and function in developing countries, especially in rural areas. We are specifically interested in understanding the impact of the density and quality of roads on transport costs and market access, and how those in turn affect the supply chain of agricultural inputs and outputs. The eventual aim is to use this data to estimate the potential gains from a hypothetical road building programme, and to benchmark them against the effects of other alternative interventions, such as input subsidies.