I just returned from the IFPRI 2020 Vision Conference entitled Leveraging Agriculture for Improved Nutrition and Health. The Conference in Delhi had 10 co-sponsors, attracted about 1000 participants, featured 150 presentations, and was opened by Indian Prime Minister Manmohan Singh. As ever it was all done very professionally—I wouldn’t expect anything less from IFPRI.
1. The goal of the conference was not terribly clear to me. Was it a rallying call to maintain the momentum for agriculture, using its impacts on nutrition as added justification? Or was it a rallying call to maintain the momentum on nutrition by looking at how an additional potential resource, agriculture, could do more for nutrition? In other words, who was leveraging whom? I felt the conference featured both of these strands. Both are important, but the ambiguity did not help us to focus on the key issues.
2. Whatever the motivation, there was a clear appetite for agriculture and nutrition to work together more closely, but less of a sense of how to do it? Should we analyse multisectorally and act sectorally? Or analyse and act multisectorally? Or play to sectoral strengths and co-locate interventions? Marie Ruel of IFPRI rightly said this is a researchable question, and it reminded me of the complex systems work that Danny Burns does at IDS.
3. In terms of the evidence base, we had a nice paper by Derek Heady of IFPRI updating the cross-country regression paper that Lisa Smith and I did in 2000, significantly upgrading and disaggregating the data and the methods. The paper finds that agricultural growth has a positive impact on underweight reduction and non-ag growth has a zero impact, while the reverse is true for stunting: ag growth has no impact, but non-ag growth has a positive impact on stunting reduction. India is the key outlier—when its data are excluded, the relationship between agricultural growth and stunting becomes much stronger.
4. So we know that agricultural growth at the aggregate level is important for child underweight and stunting (when Indian states are excluded). But are there key features of agricultural programmes that make them more or less potent in reducing undernutrition? Here I cited some of the preliminary results from a systematic review that I blogged about last week (by Edoardo Masset). That paper concludes that (a) the set of high quality impact studies looking at the impacts of agricultural interventions (which intend to reduce undernutrition) on undernutrition is very small at around 20, with about a third of them showing positive results and none showing negative results. Clearly we have to grow this set of studies if we are to work out how to increase the positive impact of agriculture on growth.
5. But we cannot stop there…we then need to understand the design features of these interventions that made then successful in reducing undernutrition—were they multisectoral in design or delivery? Were they staffed by people with multisectoral training? Were they community led? Did they invest in strong relational networks? Were the supported by pooled funding from different sectors? Which kinds of indicators did they track and who influenced the selection of indicators? How strong was the leadership behind the interventions? In other words we need to get behind the successful interventions, identify the enablers and then figure out ways of supporting the emergence of those enablers.
6. More evidence is not a prerequisite for action, we can learn as we go along, but the resources for doing that must be there (a minimum of 5% of the overall intervention budget I would say). And when we find things that work, we must communicate and share these innovations in creative and imaginative ways that facilitate the uptake and adaptation of these ideas using wikis, turning social networks into social capital networks, using SMS technology, and multimedia.
Was there anything terribly new about the conclusions?
For some perspective I went to look at the report from a 1984 workshop on “International Agricultural Research and Human Nutrition”, edited by Per Pinstrup Andersen, Alan Berg and Martin Forman. The first recommendation for follow up activities says that the international agricultural research centres (i.e. the CGIAR) should “begin research to estimate the effects of past agricultural research on human nutrition, either globally or for selected countries”. Post 1984, and with a few notable exceptions, I would say this sensible recommendation was ignored.
The conclusion of the 2011 conference is not that different. Post 2011, with rising levels and increasing volatility of food prices and stubborn rates of undernutrition, we simply cannot afford to let history repeat itself.
I am cautiously optimistic this time around, but no more than that.