Growing recognition of biodiversity’s role in supporting agricultural productivity has prompted economists and other scholars to quantify the “productive value” of crop biodiversity using observational data from real‐world conditions. We present the first meta‐analysis of that literature, drawing on 158 estimates from 52 observational studies published between 1998 and 2025. Overall, crop biodiversity exhibits a positive and statistically significant effect on agricultural productivity, with an average elasticity of 0.75—meaning a 1% increase in crop biodiversity (proxied by diversity indexes) raises agricultural outcomes by 0.75%. Yet, this average masks considerable heterogeneity across and within studies (standard deviation of 1.23). Complementary meta‐regression analyses attribute much of this heterogeneity to differences in research settings: biodiversity indexes based on species richness yield larger effects than evenness‐based ones, profit‐related outcomes show stronger associations with biodiversity than production‐related ones, and instrumental variable estimates exceed those from other methods. We further detect signs of selective reporting in our sample, with published estimates clustering just above conventional significance thresholds. Adjusting for this bias reduces average effects substantially, though the corrected evidence still points toward a positive—though more modest—productive value of crop biodiversity.
