Improving Food Aid Targeting for Program and Global Impact: Improving Food Aid Targeting for Program and Global Impact David Tschirley Michigan State University Presented at “Reconsidering Food Aid” A Workshop Hosted by the Partnership to Cut Hunger and Poverty in Africa March 15-16, 2006


Introduction: Introduction Targeting should really be about who gets what type of assistance, when Typically, though, the focus is on who Macro (which countries and regions) Micro (which households and individuals) I’ll do the same, but will return to the what and when at the end Focus on SSA


Why does targeting matter?: Why does targeting matter? An increasingly limited resource, which we want to use efficiently Exclusion error matters for two reasons Despite widespread poverty, research shows very large income and asset differences between the most poor and the least poor in rural SSA Most shocks are idiosyncratic to households As a result, exclusion can lead to poverty traps


Why does targeting matter? (2): Why does targeting matter? (2) Inclusion error also matters for at least two reasons: It increases the well known price disincentive effects of food aid and thus the risk that food aid will have overall negative developmental effects It increases the cost of response Thus reducing resources – and potentially costing lives – in other needy areas


What do we know about targeting?: What do we know about targeting? Targeting is often poor At macro level, because it is not only a humanitarian endeavor At the micro level, due to limited administrative capacity combined with mixed motivations at national, sub-regional, and local levels So at the micro level, targeting may be poorest precisely where most of it is going – in SSA


What do we know about targeting? (2): What do we know about targeting? (2) The method of targeting may matter less than how it is designed and implemented Hoddinott reports that only 20% of the variation in targeting performance was due to the method of targeting 80% was due to variation within any given method Resources are needed to allow practitioners more systematically to learn how to improve the details of design and implementation


What do we know about targeting? (3): What do we know about targeting? (3) Some mis-targeting is rational even from a humanitarian perspective Due to rapidly escalating cost as targeting criteria become more narrow Especially in a place like SSA And to inevitable trade-offs between exclusion and inclusion errors Due to imperfect information and administrative capacity The costs and benefits of different approaches need to be better understood


Food aid disincentive effects: Food aid disincentive effects Research shows strong reason to be concerned about: price disincentives frequently poor targeting, especially in SSA Despite this, little evidence that food aid creates dependency (at least at hh level) or negative production effects And some recent evidence of positive effects on labor supply and ag production


Food aid disincentive effects (2): Food aid disincentive effects (2) Along with clearly positive (but not always large) impacts on nutrition and survival during emergencies The positive effects from food aid’s resource transfer element seem, on average, to outweigh the known negative effects


Food aid disincentive effects (3): Food aid disincentive effects (3) Yet this resource transfer is very inefficient And better targeting (the who) is only one of the ways to improve it Being more flexible in what we provide … … cash and non-food assistance And finding ways to provide it in a more timely fashion (the when) … e.g., through local procurement Would dramatically increase the impact of our food assistance programs And hasten the day when Africa is a growing commercial market for our output


Slide11: Thank You!