Can smallholder extension transform African agriculture?

Agricultural productivity in Sub-Saharan Africa (SSA) lags behind all other regions of the world (Block, 2014; World Bank, 2008) despite substantial government expenditures in agriculture. The slow growth of agricultural output, modern input use, and commercialization serve as a kind of litmus test in this space: to some, the existing constraints to technology adoption and commercialization suggests that there is a lot of low-hanging fruit. To others, widening productivity gaps indicate that policy-makers should instead focus their resources on other parts of the economy.

Even among those who believe that agriculture is important for economic growth, it is hard to dispute that agricultural policy experiments in SSA have yielded more evidence on what fails than on what works.

In theory, extension services can be an important tool for addressing low productivity. Unfortunately, existing evidence shows mixed results, and many studies have faced measurement issues and low comparability across programs (Aker, 2011; Anderson and Feder, 2007). Even when extension does encourage farmers to adopt new practices, measurable increases in yields or profits remain rare (Udry et al., 2019; Beaman et al., 2013). This is troubling, but do these findings mean that extension services have no hope of improving agricultural yields?

In a new working paper, we provide evidence of a rare success story. The program we evaluate already operates at scale and targets smallholder staple crop producers. This project shows that high-quality extension services can meaningfully increase farmer production and profits. Figure 1 below compares the maize production for farmers who participated in the program (treatment) and those who did not (control).

NGO - researcher collaboration

We worked with One Acre Fund, an NGO that currently works with over 700,000 farm households in seven countries in eastern and southern Africa. Their core service targets small farmers of staple crops like maize. Farmers sign up for a bundle of services based on the amount of land they want to cultivate and enroll in the program. Once enrolled, they receive training on improved farming practices, high-quality inputs (seeds and fertilizer) on credit based on the size of their enrolled land, and insurance on the cost of those inputs. In 2017, we worked with One Acre Fund on a randomized trial of this core small-farm program in western Kenya.

Does it work?

In short, yes it does! In our paper, we show that participation in the One Acre Fund small-farm program causes meaningful increases in maize yields and in total maize output. Perhaps even more impressively, the study detects a statistically and economically meaningful increase in participating farms’ profits. The profit calculations take into account labor and input costs, and would be positive even if we subtract the small donor-funded portion of the program.

Figure 1

Our core takeaway from these results is that a well-designed, well-managed program that measures farmer behavior and output carefully can be an effective tool for improving the performance of smallholder agriculture.

Why does it work

As researchers, we tend to assume that the impacts of development interventions–and especially agriculture programs–will vary dramatically across the population (as in for example Carter et al. 2019). In this study, we find that nearly all farmers benefit from the program. Further, the benefits are remarkably similar across different types of farmers. We think the bundled nature of the program–offering training, high-quality inputs, credit, and insurance–may be key to this broad effectiveness. Some farmers may need one component more than another, but by offering a simple, scalable, uniform program, One Acre Fund seems to be able to provide meaningful assistance to many farmers at scale. Since the program is mostly farmer-funded, this last point is important for thinking about the sustainability and scalability of approaches like the One Acre Fund program.