Research dataset supporting "Climate-Smart Agriculture and Household Food Security: Evidence of Heterogeneous Adoption Drivers from Eastern Kenya

Description

This dataset supports the manuscript entitled **"Climate-Smart Agriculture and Household Food Security: Evidence of Heterogeneous Adoption Drivers from Eastern Kenya."** It contains anonymized household survey data collected from **384 smallholder farming households** in Eastern Kenya to investigate the determinants of Climate-Smart Agriculture (CSA) adoption and its impact on household food security. The dataset includes socioeconomic, demographic, farm, institutional, and climate-related variables used to examine heterogeneous adoption drivers across household food-security groups. Household food security was measured using the Household Dietary Diversity Score (HDDS), and households were classified into three categories: food insecure, breakeven, and food secure. The empirical analysis employed a Multinomial Logit model to identify determinants of CSA adoption across food-security groups and an Endogenous Switching Regression (ESR) model to estimate the causal effect of CSA adoption on household food security while accounting for selection bias. The repository contains the cleaned Stata dataset (.dta) and the Stata do-file (.do) used to reproduce all statistical analyses. This repository is intended to promote research transparency, reproducibility,and the reuse of data for future studies on climate change adaptation, climate-smart agriculture and household food security.

Steps to reproduce

### Steps to Reproduce 1. Download all files contained in this repository. 2. Open the dataset (`Food Security Data.dta`) using **Stata 18** or a later compatible version. 3. Place the dataset and the analysis script in the same working directory. 4. Open the file `New FS dofile` in Stata and set the working directory to the location containing the repository files. 5. Run the do-file from beginning to end. The script performs data preparation, descriptive statistics, estimation of the Multinomial Logit and Endogenous Switching Regression (ESR) models, and generates the statistical outputs reported in the manuscript.

Institutions

• University of Embu

  Embu County, Embu

Categories

Licence