For decades, farmers in the American corn belt have been battling a persistent and elusive foe. The enemy in question? A five millimeter corn rootworm, credited to be the most expensive pest in American agriculture, and responsible for annual losses of two billion dollars in the US alone. The weapons at the farmers’ disposal have been either traditional pesticides (which are prone to the worm developing resistance, and with dire environmental consequences), or crop rotation (an approach rendered ineffective after the bugs started preying on the crops used in the rotation). However, decades of private-sector research into RNAi technology[1] will finally come to fruition with the forthcoming release of a new corn trait, targeting the action of specific genes that are essential to the attack that the rootworm makes against the standing crop. The technology is projected to save American producers billions of dollars every year. Yet there is little prospect that these advances will directly benefit farmers in the rest of the world.

In a recent study, Jacob Moscona (Harvard) and Karthik Sastry (MIT) show that international transfer of such biotechnology breakthroughs are often hampered by a mismatch in local environmental conditions. While R&D was able to significantly support the fight against the corn rootworm in the Northern hemisphere, many pests primarily affecting other regions, such as the Maize Stalk Borer in Sub-Saharan Africa, have not been targeted by large biotech companies operating on the research frontier. Furthermore, the unsuitability of the RNAi corn to the problems of the global South means that agricultural fields in the richest areas of the world will become even more productive relative to the poorest ones, thus enlarging disparities that are already enormously consequential.

Moscona and Sastry draw from the inappropriate technology hypothesis, a 50-year-old idea that states that the process of development of new technologies by countries in the forefront of innovation inevitably focuses on what will work best in those contexts of prevailing climates, soils, pests, diseases, thereby limiting their applicability to other geographies. In a context where the US and Europe are overrepresented in R&D – with over 45% of all research expenditures, compared with Africa accounting for 1.1% and South Asia with 2.5% – the hypothesis posits that innovation will be biased toward the needs and demands of the developed world. Indeed, the authors document that biotechnology advances focus primarily on agricultural characteristics of the wealthy countries, with 67% of all new varietal innovations being firstly recorded in the US, Canada, or the EU. As with the new RNAi corn trait, we have good reasons to think that these innovations will better fit the fields of the already wealthy North.

To quantify the impact of this appropriate technology gap, the authors’ empirical strategy relies on the dissimilarity of the mix of crop pests and pathogens (CPP) between countries as a measure of appropriateness. The more dissimilar a country’s CPP mix is from a nation on the agricultural research frontier, the less likely it will be that new technologies will be diffused in the country. More technically, to measure the inappropriateness of the innovations, the researchers compute a measure of “CPP Distance” between countries, which captures the extent of their ecological differences. For instance, countries with similar ecological profiles share the same pests in the same crops, resulting in a CPP Distance of zero – a special case in which biotech innovations are perfectly exchangeable across geographies.

To examine the relationship between the CPP Distance (the inappropriateness) and the diffusion of innovations, the authors use an international catalog that records the introduction of new varieties and tracks their diffusion into other countries. The tested hypothesis is that the higher the inappropriateness, the lower the diffusion. And indeed the paper finds that for the same crop, on average, two countries fully dissimilar in their CPPs have 65% less biotech transfer between each other compared to countries that are fully similar. The effects of CPP distance on diffusion are substantially higher when the sending country is a country in the research frontier[2].

After proving the lower diffusion generated by inappropriateness, a step further is to study the production distortions generated by the CPP mismatch. Using country-level agricultural production data, the authors measure how the agricultural output relates to the country’s CPP Distance to the frontier of innovation[3]. In a CPP Distance scale from zero to one, with the upper value meaning complete ecological dissimilarity, a one standard-deviation increase in the dissimilarity from frontier countries reduces output by half a standard-deviation.

The estimates presented confirm that companies operating in countries on the agricultural R&D frontier neglect the pests and diseases of places where ecological conditions are different. Much of this situation is determined by what the large multinational companies carrying out the research expect to be privately profitable for their shareholders. And yet the broader economic returns to investments in agricultural R&D are regularly estimated to be just as high for the developing countries as for the global North. This wedge between privately profitable, and the social optimal, is a major rationale for the existence of CGIAR and its historical focus on delivering both global public goods and on carrying out research on the adaptation of innovations to local contexts.

Moscona and Sastry examine two case-studies of central importance to CGIAR – the Green Revolution period in Asian countries in 1960-1980; and the conundrum of continued low adoption of improved varieties in Africa. They argue that both cases are evidence of the inappropriate technology hypothesis. The Green Revolution effort started from a small number of high-yielding varieties in breeding centers, followed by their wide distribution to many countries, including some with dissimilar ecological conditions. This may have hampered adoption of varieties that proved to be less productive in places where ecological dissimilarity was higher (Figure 1). The authors also find a negative relationship between the household-level adoption of improved seeds in Africa[4] and the inappropriateness indicator (CPP Distance) to R&D leader countries: according to their estimates, the use of new varieties by African farmers would be 14% higher if R&D took into account the pests and diseases present in that continent.

Figure 1: During the Green Revolution, breeding centers developed high-yield varieties that were widely spread. However, the more dissimilar a country’s ecological profile is from the location of the center, the lower the crop output growth in the period that concentrated the effort of agricultural innovations.

If left to the market, and without an institution such as CGIAR that can act as a buffer, new agricultural technologies have a tendency to increase inequality in agricultural productivity by simply not being tailored to low-income countries’ environmental conditions. Moscona and Sastry estimate that the inappropriateness of the biotechnology developed by frontier countries reduces the world’s agriculture productivity levels by 41%.

This fascinating paper highlights and quantifies how ecological and biophysical differences are one of the key reasons that there are no silver bullets in agriculture. To counter the prevailing trend of inappropriate technology in plant varieties highlighted by Moscona and Sastry, CGIAR plant breeders continue to collaborate with their partner researchers in the national agricultural research systems. This is a key mechanism through which the best science can be brought to bear to address the constraints faced by farmers in the global South and as such reduce the productivity gap. By rigorously documenting which of the varieties resulting from such partnerships are being adopted by farmers in target countries, SPIA continues to provide evidence on the effectiveness of this approach. 


[1] RNAi technology was developed in 1990 and first incorporated into a crop in 2003.

[2] The authors prove that the results are stable for other links between countries that are not incorporated in the ecological distance, such as trade and distance, and for an additional measure of ecological distance.

[3] Findings are similar using a set of innovators or the US alone.

[4] Data from World Bank’s Living Standard Measurement Survey, Integrated Surveys of Agriculture (LSMS-ISA).