Data and code for "Agricultural response intensity does not signal lower drought sensitivity"

This record contains the data, code, documentation, statistical outputs, and figure sources supporting the manuscript “Agricultural response intensity does not signal lower drought sensitivity.” The archive includes frozen processed analytical inputs, the complete source-data workbook, country and crop crosswalks, Python analysis and figure-generation scripts, pinned software environments, model outputs, bootstrap and permutation draws, all country- and crop-exclusion results, seven submitted figures, integrity checks, and SHA-256 manifests. The primary analysis evaluates a noncausal country-level association between an agricultural response-intensity score and an externally sourced, typology-excluded drought-sensitivity outcome across 69 matched dryland-core countries. The response-score coefficient is +0.800793 percentage points per one-standard-deviation increase, with an HC3 95% confidence interval of 0.077630 to 1.523956 and P = 0.029979. Higher outcome values indicate greater estimated drought sensitivity. The repository reproduces all statistical results and figures from frozen processed analytical inputs with one documented command. It does not claim to reconstruct every upstream public dataset from raw downloads. The analysis plan included in the archive is a reconstructed reproducibility record and is not a preregistration. Profiles A–D are overlapping descriptive score combinations and do not constitute natural classes, country rankings, or an ordered adaptation-quality hierarchy.