Standard microeconomic methods consistently suggest that, in the short run, higher food prices increase poverty in developing countries. In contrast, macroeconomic models that allow for an agricultural supply response and consequent wage adjustments suggest that the poor ultimately benefit from higher food prices. In this paper we use international data to systematically test the relationship between changes in domestic food prices and changes in poverty. We find robust evidence that in the long run (one to five years) higher food prices reduce poverty and inequality. The magnitudes of these effects vary across specifications and are not precisely estimated, but they are large enough to suggest that the recent increase in global food prices has significantly accelerated the rate of global poverty reduction. The policy implications of these findings are therefore nuanced: short-run social protection is justified in the face of high food price volatility, but passing on higher prices to producers in the long run is an important means of reducing poverty in the poorest countries.
In this paper we aim to further inform this debate by exploring the relationship between changes in poverty and changes in domestic food prices across a large swathe of developing-country poverty episodes that vary between one and five years in length. Although a large and related literature examines statistically the drivers of poverty reduction using cross-country data (Christiaensen, Demery, and Kühl 2011; de Janvry and Sadoulet 2010; Loayza and Raddatz 2010; Ravallion, Chen, and Sangraula 2007), and within that literature a handful of papers explores the links between inflation and poverty reduction (Easterly and Fischer 2000; Ravallion and Datt 2002), the present paper appears to be the first cross- country examination of whether higher food prices help or hinder poverty reduction. While cross-country analyses typically have limitations in terms of causal identification (Durlauf, Johnson, and Temple 2005), we use the exogeneity of international food prices as a natural instrument for domestic food price movements. We also engage in a wide range of robustness tests, as well as tests for structural breaks and parameter heterogeneity.