Draft
From Harvest Plus to Harvest Driven: How to Realise the Elusive Potential of Agriculture for Nutrition?
Lawrence Haddad1
Institute of Development StudiesUK
October 28, 2010
The potential for agriculture to accelerate improvements in nutrition is large. The standard pathways are well known, but are they being accessed and are new pathways being created? This short discussion piece touches on three questions: First, what are the pathways between agriculture and nutrition? Second, is the potential being realised? Third, what can be done to increase the realisation of the potential connections? The paper concludes by arguing that we need to move from the era of thinking of improved nutrition as an optional extra for agriculture to one where improved nutrition status of the population is driven by agriculture as its main reason for being. Agriculture has never and will never be the only or even the main driver of nutrition. First, agriculture is not the only instrument or sector delivering food and income—other sectors provide key wealth creation opportunities and social protection programmes are vital where markets are weak. Second, food and income are not the only drivers of nutrition status: care, water and sanitation quality, health services and the status of women are equally vital drivers. But nutrition should be the main driver of agriculture. What else is agriculture for?
1. What are the pathways between agriculture and nutrition?
The standard pathways are well known (World Bank 2007, Haddad 2000):
(a) Greater farm productivity leads to greater farm income which can generate economy-wide income growth. We know that income growth does improve dietary diversity but that in terms of anthropometry of infants (a key nutrition outcome) it is a rather underpowered and hit and miss driver.
(b) Lower food prices as supply and efficiency of production increase. Lower food prices generate de facto income increases and lead to improvements in nutrition as in link (a). If the price declines are in fruits and vegetables and fish/livestock/dairy, then there will be additional nutrition impacts as the prices of key micronutrients decline.
(c) More nutritious production for own consumption. We also know that there is not a complete separation of what is eaten from what is grown. If on farm income generation is more geared towards high nutrition value crops then we can assume more of these will be consumed from own production.
1 My thanks to Howdy Bouis who 25 years after being my PhD adviser is still giving me excellent comments on papers. All errors (as in 1985) are mine.
1
Draft
(d) As (c), but with more general consumption effects. Biofortification comes in strongly here as a way of potentially increasing the supply of key micronutrients without compromising (and even possibly increasing) the supply of macronutrients.
(e) Empowering women to enhance nutrition impacts of (a)–(d). Greater control by women at all stages in the agriculture-nutrition chain will tend to reflect their preferences and priorities more and this tends to enhance nutrition outcomes.
What are the key policy levers?
In terms of generating poverty reduction we know the work of Fan and others (e.g. Fan and Zhang 2008) that agricultural research, development and investment is important, but we don’t know enough about how the portfolio in terms of crops and attributes affects diet and nutrition. We also know that investment in agricultural infrastructure is vital for poverty reduction, but which types are most potent for nutrition and when: irrigation, processing facilities, cold chains, or communications? We know that there is a gap between microfinance and the formal banking system when it comes to small enterprises such as farms, but how important is this finance gap for smoothing consumption across shocks? There is a lot we don’t know about the choices we make in agricultural research, development and investment and the impacts they have on nutrition.
In terms of influencing demand for certain types of foods and nutrients and how well they are utilized, we have nutrition knowledge campaigns which are shown to be effective when in combination with other non-nutrition interventions (Leroy et. al. 2009). We know that empowering women via political quotas, via enhancing asset and income control and in terms of legislation that enhances their agency, if it enhances their own nutrition status, will on average be good for family nutrition (Birner et al. 2010).
In terms of influencing the supply of certain types of foods and nutrients biofortification seems promising (HarvestPlus Orange Flesh Sweet Potato in Uganda and Mozambique 2010) although we should not globalise about its cost effectiveness in all contexts and for all crops. A national Homestead Food Production (Ianotti et. al. 2009) programme in Bangladesh has “convincing evidence of impact on household production, improved diet quality, and intake of micronutrient-rich foods, although its contribution to reducing the prevalence of deficiencies in vitamin A, iron or zinc has yet to be determined.
But how do we make sure these multiple pathways are actually travelled?
2. Is this potential being realised?
Clearly the potential is there. Is it being realised? For several reasons, this is a difficult question to answer.
First, the impact evaluations of agriculture that are outcome focused at the human wellbeing level, let alone nutrition focused, are hard to find. The CGIAR’s own Standing Panel on Impact Assessment (SPIA) lists impact evaluations done throughout the CGIAR.
2
Draft
Table 1 shows that out of the 761 listed by the CGIAR as having been published from 1995-2008, only 83 listed impact focusing on welfare indicators such as income or nutrition/health status.
Table 1: CGIAR Impact Assessment Studies
Impact evaluations focusing on income as an outcome
variable
Impact evaluations focusing on (income or
nutrition/health) as an outcome variable
All Impact Evaluations
2008 0 02007 1 22006 4 42005 0 02004 4 52003 5 6Total 1995-2008 67 83 761
As of August 2009 http://impact.cgiar.org/
Neither the Poverty Action Lab (J-PAL) nor the International Initiative on Impact Evaluation (3ie) have undertaken or commissioned many agricultural project impact studies. As of mid 2009 the project database search at the Poverty Action Lab website shows 25 health evaluations, 38 in education, only 5 in agriculture (and these are all in Kenya) i. And only 2 of 18 funded applications in round 1 of 3ie (the International Institute for Impact Evaluation) funding were awarded to agriculture projects (irrigation, low cost farm equipment) compared to 6 in health. Presumably this reflected some combination of low submissions (perhaps due to the size of funding chunks available) and lack of quality of submissionsii.
Second, the aggregate data on the impacts between agricultural growth and income or nutrition are inconclusive. Cross-country econometric work (Ligon and Sadoulet, 2008) reported in the 2008 World Development Report shows that a 1% gain in GDP originating in agriculture generates a 6 % increase in overall income for the poorest 10% of the population. This compares with a 4% increase in overall income for the next poorest, and 3% for the subsequent decile. In stark contrast, GDP growth originating in non-agriculture sectors generates zero growth for the poorest 10% of the population, a 1% increase in income for the next 10% and a 2% increase thereafter. A more recent empirical study by Christiansen et al. (2010) comes to similar conclusions. Using cross country econometric evidence they report “Irrespective of the setting, a one percent increase in agricultural per capita GDP was found to reduce the total $1-day poverty gap squared by at least 5 times more than a one percent increase in GDP per capita outside agriculture”p 30. For a large set of countries within a cross-country regression framework, Loayza and Raddatz (2009) found that growth in labour intensive sectors was the most poverty reducing. Cross-country regressions simply represent average associations between variables. It is useful to contrast their results with careful large country time series studies. IFor Brazil Ferreira et. al. (2006) found that growth in the service industries was the most poverty reducing for the 1985-2004 period. For India, Datt and Ravallion (2010) found that pre-1991, rural growth was more poverty reducing than urban growth, but for the post 1991 period the reverse held true.
3
Draft
In one of the few recent careful cross-country studies on agricultural growth and nutrition (as opposed to income) Heady (2010) found:
“sectoral growth effects do not seem to explain much of the variation in aggregate growth of nutrition outcomes, at least in the short run. We did find long run (levels) evidence of a much larger elasticity between malnutrition and agricultural growth relative to nonagricultural growth, but this pattern disappeared in shorter run episodes, except for adult BMI” p.31.
So the evidence seems to point to positive impacts of agricultural growth on the income of the poor, but is a little less clear when it comes to nutrition outcomes.
Third, the literature reviews that have been conducted are good quality in general, but not systematic in terms of protocols for inclusion and exclusion, grouping around outcomes and interventions. The Del Carpio et al. (2009) meta-evaluation of the general impacts of agricultural interventions provides a good example of the kind of study that is needed in agriculture-nutrition. Figure 1 below describes the selection of studies for inclusion in their meta-study.
Figure 1: del Carpio et al’s (2009) protocol for their systematic review of the impact of agricultural innovations
Note: IE=impact evaluation
The following table summarises Del Carpio et. al.’s results which show input technologies to have the lowest percentage of non-positive impacts, something that bodes well for biofortification (although biofortification relies on these other interventions too).
4
Draft
Table 2: Relationship between interventions and agricultural outcomes
Fourth, we don’t yet know enough about the impacts of biofortification on nutrition. Hopefully this conference will give us more results on the science and the efficacy, and possibly even effectiveness, but most likely we will have to be patient. The Meenakshi et. al. (2009) ex-ante study reviews the evidence along the theory of change of biofortification and uses these assessments to construct optimistic and pessimistic assessments of costs per DALY averted and then compares these to supplementation and fortification interventions. Biofortification comes out relatively well under the optimistic scenarios -- but not under the pessimistic ones. I have already mentioned the HarvestPlus Orange Flesh Sweet Potato study (2010) and the positive impacts on farmer uptake and dietary intake and it will be interesting to see the impacts on serum retinol and health related outcomes.
Overall then, weak and poorly organised evidence makes it hard to assess whether the potential for agriculture to increase its impact on nutrition is being realised. My hunch would be that we are only beginning to scratch the surface of the potential.
3. What can be done to increase the realisation of this potential?
So what needs to be done to increase this potential? For example, how do we make the optimistic assumptions around biofortification’s theory of change a reality? While technical ideas around how to dovetail nutrition and agriculture are necessary, they are not sufficient. What is needed to make the agriculture and nutrition innovations work together is institutional innovation to facilitate and generate political pressure.
Fundamentally, getting agriculture and nutrition together is a political problem. But how can the political pressure for agriculture and nutrition to work together be generated and sustained?
a. Map nutrition outcomes in real time
New methods for monitoring nutrition outcomes are needed. Real time monitoring of nutrition outcomes makes nutrition harder to ignore and can guide action to reduce malnutrition. Mindful of the past successes and failures of nutrition monitoring and what it
5
Draft
takes to sustain them in terms of organizational incentives to collect and use nutrition-relevant data we need to work with the web 2.0 community (e.g. Frontline SMSiii) to identify, develop and test new monitoring possibilities afforded by mobile technologies and cloud computing (see Bhawsar 2009 for a review of several areas where pilots are taking place). If effective these methodologies will be particularly valuable in fragile contexts where events change rapidly and unpredictably and where conventional data systems are extremely weak. Fresh streams of nutrition data will keep the issue in the public mind and put pressure on agriculture to act.
b. Capitalise on the increasing need to demonstrate impact in MDG terms
More and more donors are emphasising impacts of intervention on outcomes rather than only inputs and outputs. The impacts have to be framed within the MDGs and therefore have to be able to show impact at the human level. I was just involved in a donor review of multilateral organisations, and this was a key criterion. I imagine that donors will make it become so for the CGIAR and NARs. This creates an opportunity for advocates of closer links between agriculture and nutrition within the donor community: insist on agricultural projects and programmes being evaluated in terms of nutrition outcomes. There will be push-back along the lines of: the causality chain is too long, attribution is too difficult, and we don’t have the skills. All of these are challenges of course but they are not insurmountable. The IFPRI commercialisation of agriculture studies from the 1980s showed that it can be done and shared some methods on how to do it (Von Braun and Kennedy 1994). We certainly need donors to invest more in measuring nutrition status methods reliably and quickly (especially for micronutrient status), but difficulty of measurement is not, in my opinion, a valid excuse for stopping measurement halfway along the agriculture-nutrition chain.
c. Develop diagnostic tools to help identify the points of greatest leverage of agriculture on nutrition
We have heard many policymakers complain that because nutrition is such a multi-sector issue, they lack guidance on how to prioritise and sequence action so that it addresses binding constraints in the context within which they work. This is precisely the dilemma faced by Ministries of Finance in stimulating economic growth. Practical work undertaken by the economic growth diagnostics community (e.g. Hausmann et. al. 2008) shows the way forwards for nutrition. We need processes and tools to develop typologies for action and then ways of deciding how to sequence and prioritise them in ways that are sensitive to capacity and political opportunities.
One simple typology for action is highlighted in Table 2 which has two critical axes over which to map the landscape: (1) whether food output and input markets are functioning well or not and (2) whether women are disempowered and excluded from decision-making.
If markets are functioning well, then what is grown does not have to be closely matched to what is eaten (production and consumption are separable). Here the task is to maximise farm income in a sustainable way and influence diet choices. If markets are really thin, then what is eaten is much more dependent on what is grown and the task is to directly influence the upstream and downstream agricultural investment choices. If male-female power
6
Draft
relations are really skewed against women, then their preferences are discounted and their entrepreneurship is denied. Where women are relatively empowered, the task is to influence them as decision makers. Where they are not empowered, the task is to get them into decision making positions.
Within each cell in Table 2 are suggestions for key elements of a strategy. The actual strategy developed will be determined by the context—nutritional needs, agricultural possibilities, political space, capacities at the organisation and institutional levels, fragility of context (conflict and environmental). My guess is that much of sub-Saharan Africa is in the top left hand corner.
7
Draft
Table 2: Organising and prioritising action to enhance agriculture’s impact on nutrition
Weak food markets Strong food markets
Gender exclusion
Reduce gender asymmetries with respect to decision making around agriculture—traits, crops, technology, information, time use, storage and consumption properties -- food production will map more closely into food consumption needs. Focus on institutions that can help women articulate voice, promote accountability to those voices, and be responsive to those voices
Biofortification potentially vital to get more micronutrients into local food supply without significant challenging of gender asymmetries in power
Reduce general gender asymmetries in power to:o increase farm income and
overall rural incomeoincrease nutrition impact of
farm income
Use rights, legislation, representation at the basic level and at the underlying and immediate levels of nutrition status determination to rebalance male-female power relations
Biofortification potentially still useful
No Gender Exclusion
Ag extension and research more closely linked to linear growth promoting diets
Behaviour change on farm production to link it more closely to linear growth promoting diets (e.g. dairy, F&V, livestock, aquaculture)
Biofortification potentially very important
Ensure women involved in market strengthening interventions
Maximise farm income through production of what is most profitable—make sure risk mitigation and management mechanisms are in place
Undertake behaviour change on diet and health where it has the greatest leverage
Biofortification potentially still useful
d. Develop indicators to measure commitment to nutrition
Strategies guide policy, legislation, resource allocation and civil society action. But these commitments can only be realized through implementation. How can fidelity to these commitments be assessed? If they could, they would provide nutrition stakeholders with an effective transparency and accountability tool. Work by IDS and Action Aid work to develop food security commitment indices (Masset 2010) might have potential here. Obviously working out the key components of such an index, establishing the most effective ways of collecting credible data on commitments, and the most effective ways of communicating the results are challenges. But if they can be met, the indices promise the ability to enhance media attention on nutrition and agriculture.
e. Build the next generation of nutrition-agriculture champions
One common element of nutrition success stories is effective nutrition leadership. Leadership seems to be necessary although not sufficient for sustained improvements in
8
Draft
nutrition. Leadership is needed to build teams to secure the financial and political resources to undertake and respond to nutrition monitoring, to develop and communicate nutrition strategies and to be accountable for commitments to those strategies. How can nutrition champions and leaders be developed? Mindful of related efforts in health (e.g. Lister 2007), we need to begin analysing how leadership in nutrition works, how to generate more of it, and how to embed it in agriculture. It is not a coincidence that so many involved in the IPFRI commercialisation se studies came out of graduate programmes that emphasised cross-sectoralism and cross-disciplinarity. With perhaps the exception of Cornell, training people to think and analyse and act to connect nutrition and agriculture is a key missing ingredient in today’s graduate programmes. Where is the next generation of Howdy Bouis’ going to come from?
Conclusion: From Harvest Plus to Harvest Driven?
I’m not sure how Harvest Plus got its name. It may have been an attempt to reassure the agriculture community that in striving for additional nutritional impact there would be no tradeoff with yield. Even if that was not the reason for Harvest Plus’ name, in my experience this kind of thinking is prevalent in much of the agricultural establishment. Agriculture, they say, is about food production, less about income generation, and certainly nothing much to do with nutrition. It is, in fact, about all three. The first (food production) is especially important in subsistence economies, the second (income generation) especially important in semi-subsistence systems and beyond where markets work fairly well, and the third (improved nutrition status) is the ultimate impact of agriculture. There will be tradeoffs between these goals, but ultimately there has to be a convergence of understanding and commitment that agriculture is essentially about reducing hunger and malnutrition. We need to move from a situation where each outcome has its non-overlapping constituency (I exaggerate, but not much) to a situation where each is seen as a tactical route towards the strategic goal of improved nutrition and where context rather than ideology and habit dictates tactics.
We need to move from the era of thinking as improved nutrition as an optional extra for agriculture to one where nutrition is driven by agriculture as its key raison d’etre. We need to move from Harvest Plus to Harvest Driven. This paper has offered some ideas for how to do this.
References
Berti, P, J. Krasevec and S. FitzGerald. 2004. A review of the effectiveness of agriculture interventions in improving nutrition outcomes. Public Health Nutrition: 7(5), 599–609
Bhawsar, R.D. 2009. Review of the Use of PDAs/Mobile Phone Approaches in Improving Health and Nutrition Systems. MPTAST (Madhya Pradesh Technical Assistance Support Team. For DFID India).
Regina Birner, Agnes R. Quisumbing, Nazneen Ahmed. 2010. Cross-Cutting Issues: Governance and Gender, IFPRI and Bangladesh Institute of Development Studies Prepared for the Bangladesh Food Security Investment Forum May 2010.
9
Draft
Bonnard, P. 2000. Increasing the Nutritional Impacts of Agricultural Interventions. FANTA.
Christiaensen, L, L. Demery, and J. Kuhl. 2010. The (Evolving) Role of Agriculture in Poverty Reduction: An Empirical Perspective. WIDER Working Paper No. 2010/36.
Datt, G, and M. Ravallion. 2010. Shining for the Poor Too? February 13, vol xlv no 7 Economic and Political Weekly.
del Carpio, X. 2009: Measuring the Impacts of Agricultural Projects: A Meta-analysis and Three Impact Evaluations May 25, 2009. Independent Evaluation Group, World Bank.
Fan, Shenggen and Xiaobo Zhang. 2008. Public Expenditure, Growth and Poverty Reduction in Rural Uganda. African Development Review. Volume 20, Issue 3, pages 466–496, December/Décembre 2008
Ferreira, F. H. G. & Leite, P. G. and J. Litchfield. 2006. "The rise and fall of Brazilian inequality, 1981-2004," Policy Research Working Paper Series 3867, The World Bank
Haddad, L. 2000. A conceptual framework for assessing agriculture–nutrition linkages. Food and Nutrition Bulletin; 21(4): 367-373.
HarvestPlus. 2010. How Orange Flesh Sweet Potato can Reduce Vitamin A Deficiency. Findings from a HarvestPlus Project in Uganda and Mozambique. HarvestPlus c/o IFPRI. www.harvestplus.org
Hausmann, R., B. Klinger, R. Wagner. 2008. Doing Growth Diagnostics in Practice: A ‘Mindbook’. CID Working Paper No. 177 September. Harvard University.
Hawkes, C. and M. Ruel. 2006Understanding the Links Between Agriculture and Health for Food, Agriculture and the Environment. Agriculture and Nutrition Linkages: Old Lessons and New Paradigms. 2020 Briefs. Focus 13.
Heady, D. 2010. Hit And Miss: Why Is The Relationship Between Economic Growth And Malnutrition So Varied? International Food Policy Research Institute (IFPRI), Addis Ababa Office.
Iannotti, Lora, Kenda Cunningham, and Marie Ruel. 2009. Improving Diet Quality and Micronutrient Nutrition. Homestead Food Production in Bangladesh IFPRI Discussion Paper 00928.
Leroy, Jef L. , Ruel, Marie and Verhofstadt, Ellen. 2009. The impact of conditional cash transfer programmes on child nutrition: a review of evidence using a programme theory framework. Journal of Development Effectiveness, 1: 2, 103 — 129
10
Draft
Ligon, E and E. Sadoulet. 2007. Estimating the Effects of Aggregate Agricultural Growth on the Distribution of Expenditures. Background Paper for the World Development Report 2008.
Lister, G. 2007. Building Leadership for Health. Bull World Health Organization vol.85, no.12. Geneva.
Loayza, N, and C. Raddatz. 2009. The composition of growth matters for poverty alleviation. The World Bank.
Masset, E. 2010. A review of hunger indices and methods to monitor country commitment in fighting hunger. Food Policy. Forthcoming.
Meenakshi, J.V., N. Johnson, V. Manyong, H. De Groote, J. Javelosa, D. Yanggen, F. Naher, J. Garcia, C. Gonzales and E. Meng. 2009. How cost effective is biofortification in combating ‐micronutrient malnutrition? An ex ante assessment. World Development. 38. (1): 64-75.
von Braun, Joachim, ed. Kennedy, Eileen T., ed. 1994. Agricultural Commercialization, Economic Development, and Nutrition. Published for the International Food Policy Research Institute (IFPRI) by Johns Hopkins University Press
World Bank. 2007. From Agriculture to Nutrition: Pathways, Synergies and Outcomes.
11
i www.povertyactionlab.orgii http://www.3ieimpact.org/page.php?pg=round1iii http://www.frontlinesms.com/
