This project seeks to summarize, systematize, and make publicly available basic data on the agricultural production and consumption behavior of the global poor. Using existing household survey datasets from developing countries, the project aims to characterize food production and consumption patterns across rural and urban areas, income classes, and food groups. In particular, the project will focus on characterizing the net food consumption/production position of households (i.e. whether a household produces more than it consumes), across income classes, food groups, and individual crops, as well as describing vulnerability characteristics and the range of substitution options available to households in these different categories. Collecting and systematizing such data across a geographically varied range of two dozen poor countries, and making the summarized data publicly available in a searchable database, will fill a large void in the development field. The results of this project will inform efforts to prioritize and target agricultural-related interventions and policy reforms, and to understand and manage the distributional effects of various market developments on a regional-to-global scale.
The emphasis on net consumption/production is essential in order to fully comprehend the lives of the majority of the global poor. Three quarters of the world's poor, the 2.5 billion people who live on less than $2 a day, live in rural areas. Agriculture is especially important for these people as a means of income generation and food security, and as a sector, is credited for driving economic growth in the rural economy. As rural households are simultaneously consumers and producers of agricultural commodities, net consumption/production position is crucial for understanding the heterogeneous impacts of public policies, changes in global markets and prices, and environmental changes. In conjunction with data on indicators of households' abilities to respond, improved targeting of both research and interventions is possible.