An interesting idea is that farmers themselves have a better idea of how they will benefit from the new technology. And this should be reflected in their willingness-to-pay to experiment with the new technology . Experimentation has shown that incentivizing contact farmers with the payment of bonuses proportional to their success with diffusion of the innovation in their community may be necessary and can be effective . It remains however unclear whether this recommendation is scalable. Experiments include sending SMS reminders to farmers on agronomic practices and using SMS to make recommendations customized to farmers’ own idiosyncratic soil conditions . In some specific cases, there may be simple techniques that can substantially improve farmers’ ability to obtain information through self-experimentation. An interesting experiment provides farmers with leaf color charts that guide them in their fertilizer doses decisions . The extension models described above rely on what can be called a “push approach”, whereby contact farmers are expected to pass information to others in their social networks. There is also potential for a “pull approach” to the diffusion of information through social learning. In this case, information is broadly broad casted in the community that something is to be learned from informed farmers. This “buzzing” can motivate farmers to seek information from informed farmers that may be inside or outside their normal social networks. Dar et al. thus show that “buzzing” through visible demonstration plots that use a counterfactual can be just as effective as seeding information through central farmers .
Under this approach, farmers know that there is something to be learned and they know with whom to engage in conversation to be informed. Banerjee et al. found a similar result in helping people understand the demonetization process in India: inform selected individuals in the community and widely tell the community who they are to induce information-seeking conversations. The advantage is lower cost and ability to reach less privileged farmers who do not belong to well informed social networks. In general, drying rack weed results show that information remains a serious constraint on SHF modernization. Extension services are even more under-funded than Research-and-Development and in need of new approaches that may work. Social learning can be made more efficient by a better choice of contact farmers and giving them high-powered incentives to diffuse information. Social learning can be reversed from push to pull for broader impact. Using IT services such as Digital Green and digital platforms offer interesting new options . Innovative approaches in addressing the information problem are in need of conceptualization and experimentation, with significant opportunities to make a large difference on SHF modernization.Incentives to adopt require good access to well performing markets. In contrast, SHF typically face poor infrastructure and high transaction costs, limited access to information on prices, lack of competition on local markets, and problems with quality recognition for inputs and outputs. Distance to market and poor infrastructure are major contributors to higher input prices and lower product prices for net sellers, which in turn act as a tax that discourages the adoption of innovations. Aggarwal et al. thus show that distance to market is equivalent to a 6% advalorem tax per kilometer for villages in Tanzania. Reducing travel cost by 50%, which is said to be equivalent to paving rural roads, would in this case increase local maize prices and double fertilizer adoption.
Improved infrastructure may however increase or decrease the prices of local crops depending on the competitiveness of local goods with those from further away. For Sierra Leone, Casaburi, Glennester, and Suri show that improving rural roads lowers prices on rural markets, benefiting consumers and hurting producers. They find that only when cell phone services help traders reach markets further away does improved infrastructure raise local prices, benefiting producers. For teff in Ethiopia, Vendercasteelen et al. show that proximity to cities increases the price received by farmers as well as the use of fertilizer and improved seed, resulting in higher yields, especially proximity to primary as opposed to secondary cities. Improved price information can also have mixed effects on farmer welfare. The role of IT services in reducing search costs, lowering local price volatility, raising average producer prices, and lowering consumer prices was observed by Jensen for fish in India and by Aker for grains in Niger. Better information on prices was also observed by Svensson and Yanagizawa in Uganda to have a positive effect on the level of prices received by farmers. However, better price information may have no effect on prices received by farmers if they have no option to sell on these markets and no bargaining power with local merchants . Local markets may also not be competitive, with the possibility of extensive collusion among merchants as shown by Bergquist in Kenya using an experiment exogenously varying the number of merchants on local markets with no consequent impact on price. This finding is however not supported in an extensive literature review by Dillon and Dambro who found that local markets in Sub-Saharan Africa overall tend to be competitive.Finally, quality recognition is a major issue on local markets, even though urban consumers may be willing to pay a price premium for higher quality, particularly for higher phytosanitary standards.
Quality recognition via third party certification, as for onions in Senegal, resulted in higher prices for good quality produce and created incentives for farmers to adopt quality enhancing technology . On input markets, Bold et al. find that there is extensive cheating on the quality of fertilizers, contributing to low adoption in Uganda. An experiment by Hasanain, Khan, and Rezaee shows that quality recognition in services markets via IT ratings can lead to improved veterinary services for artificial insemination in Pakistan. This was due to increased veterinarian effort once success rates were known to cattle owners. Lack of quality recognition for domestic production is a major issue for the competitiveness of SHFs with imported food on urban markets. It creates an increasing disconnection between what farmers produce and what urban households consume. The large number of SHFs and aggregation of production by traders high in the value chain prevents creating incentives for farmers to increase the quality of what they produce. This contributes to rising dependency of urban consumers on imported foods and low prices for domestic producers, discouraging technological upgrading. The frequent poor performance of markets due to high transaction costs, partial transmission of information, frequent lack of competitiveness, and lack of quality recognition remains a major obstacle to profitability and hence to technological upgrading. Addressing these constraints requires not only institutional innovations, but also costly public investments in infrastructure and marketing facilities.The supply-side approach to constraint removal and value chain development has helped identify a large number of technological and institutional innovations with potential to enhance adoption. In spite of this, technology adoption and modernization has been modest. As shown by continuously rising cereal yield gaps, and in spite of many local success stories, a global Green Revolution for Africa is still in the waiting. A major difficulty for technology adoption under rainfed conditions is heterogeneity of conditions. At the household level, this applies to three dimensions: farmer circumstances, farmer objectives, and farmer capacity. If these dimensions are immutable or too costly to change, technological innovations must be customized to fit these dimensions. Farmers’ circumstances such as agro-ecological conditions vary widely over short distances and across years in particular regarding rainfall patterns and soil fertility . For Zambia, Burke et al. show that only 8% of farmers can profit from basal chemical fertilizer applications due to lack of a complementary factor, in this case lime to achieve the desirable level of soil acidity.
In Western Kenya, Marenya and Barrett find that only 55% of plots can profitably use chemical fertilizers due to lack of a complementary factor, racking system in this case soil organic matter as measured by carbon content. Barghava et al. similarly find that there is complementarity between soil organic carbon and modern inputs. For adoption to go beyond farmers with complementary factors in place, technological innovations must either be customized to fit heterogeneous contextual conditions, or complementary factors must be delivered jointly with the technological innovation. Farmers’ objectives are different from breeders’ who typically focus on maximum yields in experimental plots with highly favorable controlled conditions . Farmers maximize profit or utility weighting return and risk. They may also have labor calendar objectives such as labor-saving at peak periods and labor-smoothing in the rest of the year. Labor constraints on farming may come from involvement in rural non-farm economy activities and seasonal migration, requiring to fit farming systems to accommodate complementarities between on- and off-farm labor engagements, including a gender division of tasks. The household will have nutritional objectives if part of the harvest is home consumed, and diversity of diets matter for the choice of farming systems. These specific objectives must feed into the design of new customized technological innovations. Farmers’ capacity may be improved through the acquisition of information and skills, but other dimensions of capacity are fixed factors to which technological innovations must adapt. T.W. Schultz and Foster and Rosenzweig famously showed that farmers’ education matters for technology adoption. Low skills may reduce the capacity and the speed of learning . Again, limits on capacity must be taken into account on the supply side of technology if it cannot be addressed as a demand-side constraint that can be relaxed. Technology must be kept relatively simple to use. An example is SwarnaSub1 that requires the same agronomic practices as the widely used Swarna rice variety. Another is the leaf color chart to adjust the quantity and timing of fertilizer applications. It is thus possible that available technology is not adapted to the circumstances and demands of a majority of farmers. Either it has to be adapted to the lack of key complementary factors, or the complementary factors have to be jointly delivered as a technological package. Unless this is done, lack of technological upgrading for a majority of farmers may not be an adoption issue but a supply-side issue concerning the availability of technologies that are profitable and adoptable by a majority of farmers. Lack of investment in Research-and-Development to address the specificity and heterogeneity of SubSaharan conditions noted above adds credibility to this interpretation. This is documented by Pardey et al. who shows that there is both under-investment in agricultural research in Sub-Saharan Africa as revealed by an estimated average internal rate of return of 42% for 25 countries over the 1975-2014 period, well in excess the expected return on public investment, and a continuing deterioration of the situation. Goyal and Nash document a net decapitalization of agriculture Research-and-Development capacity in Sub-Saharan Africa over the last decade. Conclusion is that an approach to using Agriculture for Development that seeks to remove constraints on adoption of existing technology from the supply side is essential, but likely to hit a low ceiling due to heterogeneity of conditions, lack of complementary factors, and diverse farmers’ objectives and capacity. Lack of quality recognition also creates increasing disconnectedness between what farmers produce and what urban consumers demand. A complementary approach to address these issues is development of inclusive value chains and constraint removal starting instead from the demand side.A demand-side approach consists in creating incentives for SHF to modernize through their participation in vertically coordinated value chains that provide links to markets for products with a profitable effective demand, while at the same time potentially offering solutions to market and institutional failures. The advantage of a demand-side approach is that it does not predetermine the solution to adoption but seeks instead broad complementarities in the ways of achieving modernization that are specific to the agent in question. Referring to the Byerlee and Haggblade classification mentioned above, we include under the category of vertically coordinated value chains, resource providing contracts and the more complex multi-stakeholder structure. Which elements are included in contracts and the specific structure and institutional form that vertical coordination will take is endogenous, depending on the particular needs of producers, the end buyers, and the context of market failures and institutional deficits.A key element of modern value chains that can result in SHF modernization is implementation of resource-providing contracts .