Crop yields in Africa south of the Sahara are generally low, in large part because of low fertilizer use. A recent study of six countries in the region showed that only 35% of farmers applied fertilizer. There are many possible reasons why farmers do not use fertilizer. They may be unaware of its effectiveness; or have degraded soils that do not respond to fertilizer; they may not have the cash to purchase it; or unpredictable rainfall may make such investments risky. Local fertilizer prices may also cut into potential profits for many farmers.
Information about local fertilizer market prices (as well as for other inputs and for crops), is key to gauging the potential fertilizer demand in a particular location and addressing these problems. Yet such data is often lacking. A new study published in PLOS One compiles local market urea prices in 18 countries in Africa south of the Sahara for 2010-2018 and uses spatial interpolation models to predict local prices at locations for which no empirical data was available. The study, by Camila Bonilla Cedrez and Robert J. Hijmans (University of California, Davis), Jordan Chamberlin (International Maize and Wheat Improvement Center, or CIMMYT), and Zhe Guo (IFPRI), is the first major attempt to systematically describe the spatial variability of fertilizer prices within the target countries and test the ability to estimate the price at unsampled locations.
While national level fertilizer prices may be available, prices can vary considerably within countries, reflecting transportation costs and other factors. In the absence of such data, analysis of household-level behaviors requires making assumptions about prices—but such assumptions may not be valid. For example, evaluations of the returns to production technologies settings have often assumed that all farmers in a country face the same set of prices. This is at odds with what we know about economic remoteness and the highly variable market access conditions under which African smallholders operate.
“Our study uncovers considerable spatial variation in fertilizer prices within African countries and gives a much more accurate representation of the economic realities faced by African smallholders than the picture suggested by using national average prices. We show that in many countries this variation can be predicted for unsampled locations by fitting models of prices as a function of longitude, latitude, and additional predictor variables that capture aspects of market access, demand, and environmental conditions,” says Bonilla Cedrez, a PhD candidate at UC Davis.
Urea prices were generally found to be higher in remote places or away from large urban centers, ports of entry, or blending facilities. There were some exceptions, though. In Nigeria, Ghana, and Benin, for instance, prices went down when moving away from the coast, possibly because market prices are lower in areas with higher demand. In other locations, imports of fertilizer from neighboring countries with lower prices may be affecting prices. Political influence can also play a role, with politically well-connected villages receiving more input subsidies compared to less connected ones.
“The performance of our price estimation methods, and the simplicity of our approach, suggests that large scale price mapping for rural areas is a cost-effective way to provide more useful price information for guiding policy, targeting interventions, and for enabling more realistic applied microeconomic research. For example, local price estimates could be incorporated into household-survey-based analysis of fertilizer adoption,” says Chamberlin, a CIMMYT spatial economist. “In addition, such predictive ‘price maps’ can be incorporated into targeting and planning frameworks for agricultural investments—for example, to target technology promotion efforts to the areas where those technologies are most likely to be profitable.”
“The evidence we have compiled in this paper suggests that, while investments in more comprehensive and spatially representative price data collection would be very useful, we may utilize spatial price prediction models to extend the value of existing data to better reflect local price variation through interpolation. Even if imperfect, such estimates almost certainly better reflect farmers’ economic realities than assumptions of spatially constant prices within a given country. We propose that spatial price estimation methods such as the ones we employ here may serve for better approximating heterogeneous economic market landscapes,” says Hijmans, a UC Davis professor.
This study illustrates new ways for incorporating spatial variation in fertilizer prices into efforts to understand the profitability of agricultural technologies across rural areas in Africa south of the Sahara. One important avenue for future empirical work, the authors suggest, would be to evaluate the extent to which now-documented subnational price variation can be useful in explaining observed variation in smallholder fertilizer use in Africa south of the Sahara, after controlling for local agronomic responses and output prices. One way to do that may be to integrate input (and output) price predictions into spatial crop models, and then evaluate the degree to which modeled fertilizer use profitability predicts observed fertilizer use rates across different locations.
Joshua Masinde is a Communications Specialist with the International Maize and Wheat Improvement Center. He is based in Nairobi. This post also appears on the CIMMYT site.
Funding for this project was provided by the Feed the Future Sustainable Intensification Innovation Lab (SIIL) through USAID; the Bill and Melinda Gates Foundation, and the International Maize and Wheat Improvement Center (CIMMYT).
