Structural Capacity, Food Security-Related Publications, and Crop Production: A Multilevel Global Analysis Across Income Settings

Agricultural performance is often interpreted through agronomic inputs and technological progress; however, the translation of knowledge into production depends on the structural environments in which food systems operate. This study examined the association between food-security-related publication activity and crop production across global income settings from 2000 to 2025, while testing whether governance, health-system, and financial indicators modify that association. A longitudinal ecological panel was constructed, integrating 61,158 Scopus-indexed peer-reviewed articles on food security and related dimensions of healthy food access and availability with 23 crop production indicators grouped into staple, horticultural, and commodity domains. Income-stratified regression models were followed by hierarchical mixed-effects models and moderator screening. In exploratory stratified models, 67 of 92 income-specific associations reached nominal significance; however, only 5 of those 67 associations (7.5%) remained statistically significant after multilevel modelling and false discovery rate correction. Robust associations were concentrated in selected staple and horticultural outcomes, whereas most commodity indicators lost significance after hierarchical adjustment. Structural moderators related to territorial control, corruption, healthy life expectancy, health researcher density, healthcare access and quality, and official development assistance shifted the conditional slopes linking publication activity to crop output. These findings do not support a uniform linear relationship between publication growth and production volume. Instead, they suggest that the alignment between research ecosystems and agricultural output is structurally conditioned and likely mediated by institutional capacity, health-system resilience, and implementation environments. The ecological design, the use of publication counts as an indirect proxy, and the reliance on production volume rather than yield or efficiency should be considered when interpreting these results.