Evaluating the impacts of an agricultural technology adoption program using a randomized control trial
Date
2019-08
Authors
Aramburu, Julián
relationships.isContributorOfPublication
Reynoso, Ana
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de San Andrés. Departamento de Economía
Abstract
El presente trabajo evalúa los impactos de la adopción de tecnología agrícola en un conjunto de resultados que incluyen el ingreso de los hogares, la seguridad alimentaria, y diferentes medidas de productividad. En particular, se evalúa el Programa Agrícola de Innovación Tecnológica (PATCA) en la República Dominicana, que tuvo como objetivo aumentar la productividad agrícola y los ingresos de los pequeños agricultores mediante el fomento de la adopción tecnológica. Este trabajo hace uso de la asignación aleatoria del programa para inducir una variación exógena en la adopción de tecnología. Se utiliza una encuesta integral de hogares para una muestra de 2.214 agricultores, incluidos beneficiarios directos, beneficiarios indirectos y controles. Las estimaciones de 2SLS indican que la adopción de una tecnología agrícola provista por el programa PATCA mejoró el estado de seguridad alimentaria de los hogares beneficiarios en alrededor de 19%, aunque no se encontraron efectos en los ingresos del hogar. Se encuentra diferentes patrones de adopción e impactos en las medidas de productividad para cada una de las dos tecnologías analizadas (riego y rehabilitación de pastizales). La inferencia se basa en un método de corrección de Bonferroni que controla por pruebas múltiples. Además, una evaluación preliminar de los efectos indirectos no validó las hipótesis de que podrían producirse efectos indirectos a nivel geográfico.
In this paper, I evaluate the impacts of agricultural technology adoption on a set of outcomes including household income, food security, and different measures of productivity. In particular, I evaluate the Agricultural Program for Technological Innovation (PATCA) in the Dominican Republic, which aimed to increase agricultural productivity and income of small-farmers by encouraging technological adoption. I exploit the random assignment of the program in order to induce exogenous variation in technology adoption. I use a comprehensive household survey for a sample of 2,214 farmers including direct beneficiaries, indirect beneficiaries and controls. 2SLS estimates that account for the presence of non-compliance indicate that the adoption of an agricultural technology provided by the PATCA program improved the food security status of beneficiary households by around 19%, although no effects are found on household income. Different patterns of adoption and impacts on productivity measures are found for each of the two technologies analyzed (irrigation and grassland rehabilitation). Inference is based on a Bonferroni correction method that accounts for multiple outcomes testing. Also, a preliminary assessment of spillover effects did not validate the hypotheses that indirect effects might take place at the geographical level.
In this paper, I evaluate the impacts of agricultural technology adoption on a set of outcomes including household income, food security, and different measures of productivity. In particular, I evaluate the Agricultural Program for Technological Innovation (PATCA) in the Dominican Republic, which aimed to increase agricultural productivity and income of small-farmers by encouraging technological adoption. I exploit the random assignment of the program in order to induce exogenous variation in technology adoption. I use a comprehensive household survey for a sample of 2,214 farmers including direct beneficiaries, indirect beneficiaries and controls. 2SLS estimates that account for the presence of non-compliance indicate that the adoption of an agricultural technology provided by the PATCA program improved the food security status of beneficiary households by around 19%, although no effects are found on household income. Different patterns of adoption and impacts on productivity measures are found for each of the two technologies analyzed (irrigation and grassland rehabilitation). Inference is based on a Bonferroni correction method that accounts for multiple outcomes testing. Also, a preliminary assessment of spillover effects did not validate the hypotheses that indirect effects might take place at the geographical level.
Description
Fil: Aramburu, Julián. Universidad de San Andrés. Departamento de Economía; Argentina.