But the question remains whether the conceptual problems and implicit assumptions made are so problematic that it is impossible to come up with a reliable aggregate measure of ‘hunger’. Given the discussion above, I would submit that it seems indeed impossible to develop a reliable measure of ‘hunger’ using such an aggregate input-oriented approach for several reasons. First, it is not very clear whether average caloric deficiency over a three year period is really a good measure of ‘hunger’. It ignores inter-seasonal variations, shortterm shocks, and the role of other nutrients. Second, we know simply too little about critical assumptions of this approach, including how many calories get wasted, how calories are distributed among the population in different countries (which is likely to be different across space and changes over time), and what the caloric requirements are for different population groups, depending on their circumstances. I find it also implausible that we will reliably fill these knowledge gaps anytime soon, certainly not before 2015. So frankly my (probably contentious) conclusion is that further research in this area is unlikely to generate a reliable and comparable aggregate indicator of ‘hunger’ so that it is probably better to think about alternative approaches.