Lost in aggregation? On the importance of local food price data for food poverty estimates

Abstract This paper explores within‐country variations in food price dynamics and food poverty estimates by employing local market price data and national consumer price index (CPI) data. Our results show that national CPI data may be useful for approximating national trends but they fail to detect and identify spatial variations in local trends, which have relevant welfare implications. Indeed, sourcing cereal prices from local markets, instead of national CPI statistics, leads to classification divergences in 9% of food poverty predictions, with a substantial variation in estimation errors across countries. Our findings thus underscore the need for the systematic collection of local price data to ensure effective policymaking, such as adjusting social transfers to better reflect local differences and optimizing the allocation of social funds.