Using Machine Learning to Estimate the Heterogeneous Effects of Livestock Transfers

Conner Mullally, Mayra Rivas and Travis McArthur

American Journal of Agricultural Economics, 2021, vol. 103, issue 3, 1058-1081

Abstract: We evaluate a program in Guatemala offering training and transfers of a local chicken variety using a randomized phase‐in design with imperfect compliance. We do not find strong evidence for or against positive average intent‐to‐treat effects on household‐level outcomes, including indicators of expenditure, calorie and protein intake, diet quality, egg consumption and production, as well as chicken ownership and management. Among girls between the ages of six and sixty months, we find that the program reduced stunting by 23.5 (± 19.4) percentage points while also improving other height and weight outcomes. Boys are more likely to suffer from intestinal illness, which could explain differences in program impacts by gender. Using machine learning methods, we show that the poorest households enjoyed the largest impacts on diet quality and animal protein consumption, whereas children in the poorest households experienced the largest impacts on the probability of consuming animal source foods. Larger effects on animal source food consumption among children in relatively poor households did not translate into greater impacts on height or weight.

Date: 2021
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
https://doi.org/10.1111/ajae.12194

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wly:ajagec:v:103:y:2021:i:3:p:1058-1081

Access Statistics for this article

More articles in American Journal of Agricultural Economics from John Wiley & Sons
Bibliographic data for series maintained by Wiley Content Delivery ().