Africa’s Green Revolution? Raising Awareness of New Rice Crops for Africa (NERICAs)

January 13, 2014 • In-depth, Climate Change & Society, Middle East & Africa, Unprotected Post, World Development

By Aliou Diagne, Steven Glover, Ben Groom and Jonathan Phillips

A Green Revolution in West African agriculture will rely on new technologies such as NERICA rice varieties. Below, Aliou Diagne, Steven Glover, Ben Groom and Jonathan Phillips argue that raising awareness of NERICAs and improving access to its seed by farmers should be a priority. Strikingly, those that know about NERICAs are not the most likely to adopt them. By targeting ‘Detached Adopters’ it may be possible to rapidly raise the rate of adoption of NERICAs, starting with Gambia’s women.

West African agriculture is characterised by very low land and labour productivity. While there have been some success stories, the majority of production is subsistence based, leaving very little scope for growth and development. The Green Revolutions that pioneered and helped sustain Asian development in the 1960s have, thus far, proved elusive. To realise the potential productive capacity of West African agriculture, improving the technology used by farmers is crucial. In particular, widespread dissemination of improved seed varieties is known to have contributed to the success of agricultural production during the Asian miracle, alongside adoption of complementary inputs.

Technology constraints are typically thought of as bottlenecks in research and development, with government and industry failing to invest in technologies with long-term growth potential. Yet, around the world a host of agricultural technologies sit unused while farmers toil with degraded soil, worn-out tools and low-productivity seeds. In West Africa, significant frontier agronomical research was conducted in the 1990s. By 2000, the Africa Rice Center had succeeded in developing seven varieties of New Rice Crops for Africa (NERICAs) that were adapted to the West African environment. NERICAs typically offer much greater productivity, resilience to drought and pathogens such as blast, and reduced variability in yield, hence reduced risk. These new varieties can also offer up to 25% higher protein content than varieties of Asian rice, which would contribute to increased regional food security and improved nutrition.

The benefits to farmers of adoptingNERICAs are substantial; in Benin, incomes rose by US$277 per hectare for men and US$337 per hectare for women.1 There was a 6% increase in school attendance rates, a 14% increase in the gender parity index and a 36kcal per equivalent adult increase in calorie intake.2 Local economies would also benefit. West Africa depends on imports for 40% of its rice supply, making it vulnerable to international price fluctuations.3 Nigeria alone spends over $2bn per year on imports.4

The benefits to farmers of adopting NERICAs are substantial; in Benin, incomes rose by US$277 per hectare for men and US$337 per hectare for women.
Despite these benefits, the majority of West African farmers continue to use traditional varieties. The barriers to adoption are not the availability of an effective technology. While the Africa Rice Center, supported by member countries and international development agencies, has made the new varieties of NERICAs widely available, there are a number of reasons why farmers have not adopted them immediately. To understand why, household survey data was collected from farmers in Guinea, The Gambia and Côte d’Ivoire. The research focused on the proportion of farmers who are likely to adopt NERICAs in the coming years as their awareness grows, and – the counterpart to this question – identifying the barriers that inhibit adoption.

Understanding Adoption Choices

Learning which farmers will adopt NERICAs, and under what conditions, poses significant methodological challenges. A simple comparison of the characteristics of those who adopt and those who do not risks making two errors.

First, while NERICAs sit on shelves in stores, farmers may not have knowledge of the varieties, of their potential benefits, or of any risks their adoption might pose. Crucially, if a farmer does not have knowledge of NERICAs, they cannot adopt the new technology. That only 5% of the population has adopted NERICAs might not be so disheartening if only 10% of the population have knowledge about their benefits.

Second, those who do have knowledge about NERICAs might not be representative of the overall population of farmers. For example, those with the most to gain might have actively sought out information, or those in social networks that disseminate information rapidly may be more likely to adopt. In addition, research and extension agencies might actively target certain types of farmers. If the 10% of the population who have knowledge about NERICAs are also the only farmers likely to adopt the new varieties, then the current adoption rate of 5% might actually be a reasonable and disheartening guide to the maximum potential adoption rate that NERICAs could achieve in the population.

The impact of raising awareness of NERICAs is large. In The Gambia, over 90% of farmers are expected to adopt the new varieties once they know about them.
The methodology for research therefore involves separating the adoption question into two parts; first, the probability of a farmer being knowledgeable about NERICAs, and second, the probability of a farmer adopting NERICAs conditional on being aware of them. To calculate these probabilities in an unbiased way, it is crucial to establish the correct counterfactual: If a farmer who was unaware of NERICAs became aware, would they be likely to adopt them? While we do not (and cannot) have a direct answer to this question, programme evaluation techniques and selection models provide the apparatus to answer this question and correct for the above biases. However, these methods rely on having enough data on the characteristics of the respondents to model the processes of exposure to knowledge and the decision to adopt.

While the surveys of West African farmers were comprehensive, there are always limits to the amount of data that can be collected. Any analysis of adoption decisions must therefore make two critical assumptions. First, that any factors we don’t have data on do not affect the adoption choices of farmers; that there is `conditional independence’ between the adoption decisions of respondents who have knowledge of NERICAs and those who do not once we have accounted for the key characteristics we have data on. Second, that when one farmer adopts NERICAs, this does not affect the potential benefits of adoption for neighbouring farmers.5

With these two assumptions, it is possible to model the factors that affect which farmers are knowledgeable about NERICAs, and which of those farmers adopt NERICAs. In the first step, a `propensity score’ is calculated that estimates a farmer’s likelihood of being knowledgeable about NERICAs given their personal characteristics. In the second step, the binary decision to adopt or not is modelled conditional on this propensity score for adoption and the farmer’s characteristics.

The resulting models enable us to extrapolate from the current data to understand what would happen if all farmers were knowledgeable about NERICAs. This enables us to estimate the potential adoption rate in the population, where all farmers are knowledgeable, and therefore the prospects of NERICAs forming the backbone of a green revolution.  It also enables us to understand how important information is to promoting adoption, and what other barriers to adoption are likely to persist even when knowledge is universal.

Raising Awareness

The impact of raising awareness of NERICAs is large. In The Gambia, over 90% of farmers are expected to adopt the new varieties once they know about them. Even in Côte d’Ivoire, which has the lowest potential adoption rate, it is expected that four in ten farmers would adopt NERICAs if they knew about them; a large increase from the current rate of four in every hundred. Given the magnitude of benefits documented in other studies, these levels of adoption could translate into a radical transformation in the West African rice market. If supported by improvements in other key inputs, including fertilizer, irrigation and equipment, there are strong prospects for NERICAs to form the backbone of West Africa’s Green Revolution.



Who Knows about NERICAs?

The data reveal that knowledge about NERICAs is as subject to economic and social forces as the adoption decision itself. Moreover, these factors are culturally and locally specific. In Guinea and The Gambia, it is men who are most likely to know about NERICAs. Yet, in Côte d’Ivoire, it is women that are more aware. Education matters in the Gambia, but less so in the other two countries. Contact with extension agencies, practicing upland rice cultivation that NERICAs are best suited to, and knowing a lot of varieties of traditional rice are universally helpful. Extension and promotion activities by the Africa Rice Center are effective, with knowledge rates that are substantially higher in targeted villages.

The data also suggest the importance of social networks and diffusion of information between villages. In Guinea, for example, living in a prefecture where NERICAs were introduced had a positive effect on adoption even after controlling for NERICA introduction in the village and the activities of extension agencies.

Barriers to Adoption

Even if all farmers knew about NERICAs, adoption would be less than 100%. This reflects real and systematic constraints that farmers face. These include limited access to seed and the complementary inputs that make the benefits of NERICAs a worthwhile investment; lack of access to credit to support these complementary investments; and a limited supply of additional labour. In all countries, these constraints in access to additional inputs seem to be important in limiting NERICA uptake. Farmers that receive support from government agencies or NGOs are more likely to adopt NERICAs, reflecting the importance of technical knowledge and access to inputs that these organisations provide.

In some settings, education has also been recognised as a barrier to the ‘ability to choose the optimal bundle of input and output mix’.6 This `Schultz hypothesis’ suggests there may be large economic returns to improving rural education. For NERICA adoption, the effects of education vary between countries; in Côte d’Ivoire and Guinea education is no major barrier to adoption, but in The Gambia it is. Perhaps reflecting a more hierarchical or structured society, being an older farmer also increases the likelihood of adoption in The Gambia.

Gender related barriers to technology adoption have been identified, stemming from unequal access to complementary inputs or reluctance to adopt due to potential changes in the ownership of output.
Strikingly, while being female has a positive effect in Côte d’Ivoire and The Gambia, women are much less likely to adopt NERICAs in Guinea even where they are aware of them. Elsewhere gender related barriers to technology adoption have been identified, stemming from unequal access to complementary inputs or reluctance to adopt due to potential changes in the ownership of output. This is despite evidence that adoption has a higher impact on output for female than for male headed households in Guinea.7

Awareness and Adoption: Targeting Interventions to Detached Adopters

An innovative aspect of the data allows us to compare the farmers who are most likely to be aware of NERICAs with those who are most likely to adopt. We can ask whether these are the same people, and if not, how they differ.

Comparing the two groups, there is only a small selection bias in the propensity of farmers who are aware of NERICAs to subsequently adopt them, and it is a negative bias. This suggests that farmers that have knowledge about NERICAs are less likely to adopt than the average farmer. One reason for this finding might be that research and extension agencies may not have been targeting those most likely to adopt, but rather the poor or marginalised farmers to whom adoption might make the biggest contribution. Nevertheless, since these farmers are least likely to adopt NERICAs, this would be an expensive targeting campaign. To the extent there are significant systemic and national benefits to adopting NERICAs, arguably there is a strong case for raising awareness of NERICAs among precisely those farmers that are least likely to know and most likely to adopt. We call this group `Detached Adopters’, reflecting their detachment from information networks but their underlying propensity to adopt. A focus on this group could be a cheap and effective way of increasing NERICA adoption.

Using programme evaluation techniques to overcome the biases created by incomplete and selective awareness, the future for NERICAs appears promising.

Importantly, the Detached Adopters vary between countries based on local socioeconomic characteristics. In the Gambia, being female is an astonishingly powerful characteristic of Detached Adopters. Women represent a large untapped group of willing adopters who have failed to realise the benefits of NERICAs purely through lack of knowledge about their potential. Targeting older people, those who have little knowledge of traditional varieties, and those yet to be contacted by extension agencies are also likely to be effective at targeting Detached Adopters.

In Guinea, Detached Adopters tend to have adopted upland varieties in the upper and forest agro-ecological zones previously, know few traditional varieties, and receive support from NGOs but not from government. In Côte d’Ivoire, Detached Adopters are harder to come by. They tend to not be from the forest agro-ecological zone and to know fewer traditional varieties of rice.


A Green Revolution in West African agriculture will rely on new technologies such as NERICA rice varieties. Yet, developing these technologies will not be sufficient. Analysis of data from West African farmers suggests that lack of awareness of new technology can be a major constraint to adoption. Using programme evaluation techniques to overcome the biases created by incomplete and selective awareness, the future for NERICAs appears promising. With potential adoption rates ranging from 40% to 90%, they are well-positioned to form the backbone of a Green Revolution.

Raising awareness of NERICAs and similar technologies should, then, be a priority for governments, donors and researchers. Strikingly, those that know about NERICAs are not the most likely to adopt them. By targeting `Detached Adopters’ it may be possible to rapidly raise the rate of adoption of NERICAs. One group that should immediately be targeted is Gambia’s women farmers, for whom the impact may well be larger than for men.

This article summarises Diagne, A., Glover, S., Groom, B., and Phillips J. (2012), “Africa’s Green Revolution? The determinants of the adoption of NERICAs in West Africa” SOAS Department of Economics Working Paper Series, No. 174, SOAS, University of London

About the Authors

Aliou Diagne is Senegalese and is currently with The Africa Rice Center (AfricaRice) as Leader of its Policy, Impact Assessment and Innovation Systems Program. His current research concentrates on the economics of rice in sub-Saharan Africa. Aliou graduated in 1994 from Michigan State University, USA, with a dual PhD in Agricultural Economics and in Economics.

Steven Glover is a member of the Development Economics Research Group (DERG) at the University of Copenhagen, providing technical support on agricultural development to the Ministry of Planning and Development in Mozambique. Steven is a former ODI Fellow in Mozambique and holds an MSc in Development Economics from the School of Oriental and African Studies (SOAS), University of London.

Ben Groom is an Associate Professor in Environment and Development Economics at the London School of Economics and Political Science. He obtained a PhD in Economics from University College London in 2005. His research includes social discounting for long-term public projects, the economics of water resources and the economics of agriculture and food security. He has served as a Consultant for numerous international organisations, including the World Bank, the Asian Development Bank and the World Wide Fund for Nature.

Jonathan Phillips is a PhD Candidate in Government at Harvard University. His research focuses on the political economy of development, and particularly on explaining subnational variation in politics, accountability and service delivery within Nigeria. He previously worked as a Fellow of the Overseas Development Institute in Nigeria’s Presidency.


1. Agboh-Noameshie A, Kinkingninhoun, F.M.M. and Diagne, A. (2007), ‘Gendered Impact of NERICA adoption on farmers’ production and income in Central Benin’. Paper presented at the 2nd African Association of Agricultural Economists Conference, Accra, Ghana, 20-22 August.

2. Adekambi S.A., Diagne, A. and Biaou, G. (2006), ‘Impact de l’adoption des variétés NERICAs sur la scolarisation des enfants au Bénin’. Mimeo, Centre du riz pour l’Afrique (ADRAO), Cotonou, Bénin.

3. 3

4. 4

5. This assumption is known as the Stable Unit Treatment Value Assumption (SUTVA). Since there are likely to be positive spillovers from adopting NERICAs, as shown in Holloway et al (2002) for Bangladesh, this is a simplifying assumption. The estimates of adoption rates presented are therefore likely to be conservative.

6. Asfaw, A. and Admassie, A. (2003), ‘The role of education on the adoption of chemical fertiliser under different socioeconomic environments in Ethiopia’, Agricultural Economics, 30, pp.215-228.

7. AfricaRice Center (WARDA). 2008. NERICA Adoption and Impact: Summary of findings from four countries. Research and Development Brief, August 2008.  Cotonou, Benin: AfricaRice, 210 pp.

8. Holloway, G., Shankar, B. and Rahman, S. (2002), ‘Bayesian Spatial Probit Estimation: A primer and an application to HYV rice adoption’, Agricultural Economics, 27, pp.383-402.

You might also like:

Leave a Reply

Your email address will not be published. Required fields are marked *

« »