Robustness Reproduction with Machine Learning

We conduct a robustness reproduction of Ammann et al. (2023), published in Food Quality and Preference, who examine how consumers evaluate the importance of animal welfare as an agricultural policy goal relative to potentially competing objectives. Using the publicly available data, we first provide a reproduction of the original results and then perform a robustness reproduction that relaxes the linearity assumption through a Machine Learning (ML) work-flow. The original association patterns are reproducible, while the ML extension uncovers modest nonlinearities with limited gains in explanatory power. Overall, this study illustrates a simple and transparent workflow that complements standard OLS-based replication with data-driven specifications, thereby supporting transparency and reproducibility in applied agricultural economics.