Heterogeneity of rural poverty: an application of the “ordered data” model

This study examines the heterogeneity of rural poverty in Peru, proposing an approach that classifies poverty into four types based on the size of the agricultural sector holding and income level. Based on data from the 1997 Living Standards Survey, a probabilistic discrete-choice model (“ordered data”) is used to assess the probabilities that a household belongs to a specific poverty category. The findings reveal that, although there was a decline in poverty in absolute terms between 1994 and 1997, extreme poverty continues to prevail in rural areas. In addition, educational variables, access to social programs, and housing conditions emerge as critical factors in determining the level of poverty. The policy implications suggest that strategies should differentiate between types of poverty to improve the targeting of social programs, emphasizing interventions that address the specific needs of the most vulnerable households, rather than a one-size-fits-all approach. (Abstract and audio: Department of Economic Publications)