Exploring the Complementarity Between Traditional Econometric Methods and Machine Learning - An Application to Adoption and Disadoption of Conservation Practices

Aug 2, 2026

Upload by Riley

Study examines how traditional econometric methods and machine learning complement each other when analyzing adoption and disadoption of cover crops, a key agricultural conservation practice. Using a longitudinal panel survey of Iowa farmers, the research compares logistic regression with a Random Forest model to identify drivers of both initial conservation choices and later discontinuation. Results show logistic regression supports economic-theory interpretability, while Random Forest improves prediction and captures complex, non-linear effects. SHAP highlights adoption scale, past behavior, and environmental factors; larger prior acreage and consistent histories reduce disadoption, while cost-share long-run effects appear limited.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages52

Reading Time2 h 11 min

AccessFree to download

Abstract

Introduction

Post-initial adoption and disadoption

Limits of traditional econometric approaches

Role of machine learning models

Exploring the Complementarity Between Traditional Econometric Methods and Machine Learning – An Application to Adoption and Disadoption of Conservation Practices Zhushan Du, Hongli Feng, and J. Arbuckle Working Paper 25-WP 668 February 2025 Center for Agricultural and Rural Development Iowa State University Ames, Iowa 50011-1070 [www.card.iastate.edu](www.card.iastate.edu) An updated version of this working paper is forthcoming in Applied Economics Zhushan Du is PhD Student, Department of Economics, Iowa State University, Ames, Iowa, 50011. [E-mail: zhushand@iastate.edu](E-mail: zhushand@iastate.edu). Hongli Feng is Assistant Professor, Department of Economics, Iowa State University, Ames, Iowa, [50011. E-mail: hfeng@iastate.edu](50011. E-mail: hfeng@iastate.edu). J. Arbuckle is Professor, Department of Sociology and Criminal Justice, Iowa State University, Ames, Iowa, [50011. E-mail: arbuckle@iastate.edu](50011. E-mail: arbuckle@iastate.edu). This publication is available online on [the CARD website: www.card.iastate.edu. Permission](the CARD website: www.card.iastate.edu. Permission) is granted to reproduce this information with appropriate attribution to the author and the Center for Agricultural and Rural Development, Iowa State University, Ames, Iowa 50011-1070. For questions or comments about the contents of this paper, please contact Rabail Chandio, [rchandio@iastate.edu](rchandio@iastate.edu). Iowa State University does not discriminate on the basis of race, color, age, ethnicity, religion, national origin, pregnancy, sexual orientation, gender identity, genetic information, sex, marital status, disability, or status asa U.S. veteran. Inquiries regarding non-discrimination policies may be directed to Office of Equal Opportunity, 3410 Beardshear Hall, 515 Morrill Road, Ames, Iowa 50011, Tel. (515) 294-7612, Hotline: (515) 294-1222, email [eooffice@iastate.edu](eooffice@iastate.edu). Exploring the Complementarity Between Traditional Econometric Methods and Machine Learning – An Application to Adoption and Disadoption of Conservation Practices Zhushan Du, Hongli Feng, J. Arbuckle An updated version of this working paper is forthcoming in Applied Economics Author name and affiliations Zhushan Du, [zhushand@iastate.edu](zhushand@iastate.edu), Department of Economics, Iowa State University, United States Hongli Feng, [hfeng@iastate.edu](hfeng@iastate.edu), Department of Economics and Center for Agricultural and Rural Development, Iowa State University, United States J. Gordon Arbuckle, [arbuckle@iastate.edu](arbuckle@iastate.edu), Department of Sociology and Criminal Justice, Iowa State University, United States Abstract: This study explores the potential complementarity between traditional econometric methods and machine learning in analyzing the adoption and disadoption of a key conservation practice in agriculture, cover crops. While the adoption of conservation practices has been widely examined, the literature on their disadoption is limited. Using a unique longitudinal panel survey of Iowa farmers, we compare logistic regression models with a Random Forest algorithm to examine factors driving conservation adoption and disadoption. Our findings show that while traditional logistic regression models offer interpretability grounded in economic theory, Random Forest provides superior predictive power and reveals complex, non-linear relationships among key factors such as past adoption behavior, cost-share participation, and farmer perceptions. SHAP (SHapley Additive exPlanations) analysis identifies adoption scale, past adoption behavior, and environmental factors as primary drivers of disadoption. Farmers with larger previous cover crop acreage and consistent adoption history are significantly less likely to disadopt, while the longterm impact of cost-share programs on continued use appears limited. By combining machine learning’s predictive power with the interpretability of

FAQs

What conservation practice and data does the study focus on?

The study focuses on cover crops and uses a unique longitudinal panel survey of Iowa farmers to track adoption and disadoption over time.

How do the authors compare econometric methods with machine learning?

They compare logistic regression models with a Random Forest algorithm to assess factors driving both adoption and disadoption decisions, evaluating interpretability and predictive performance.

What does the SHAP analysis identify as key drivers of disadoption?

SHAP analysis indicates adoption scale, past adoption behavior, and environmental factors as primary drivers of disadoption, with consistent prior adoption reducing the likelihood of discontinuation.

Aug 2, 2026

Upload by Riley