Paper short abstract:

Throughout the 1930s statistical planners advocated for model-based techniques that could measure the limits of uncertainty. This paper historicizes an international statistical estimation model, evincing the rise of a practical statistics enterprise and its achievement of model-based governance.

Paper long abstract:

This paper documents the rise of an international network of model-based planners during the Depression era. Statistical models were designed and rapidly employed in efforts to break away from a reliance on the Bayesian paradigm seen in late 19th and early 20th century statistical sampling practices. Here I focus on a particular model—a statistical estimation technique—developed to create a more efficient agrarian science program in Poland. This confidence-planning model emerged at the nexus of eastern and western science, converging mathematical modeling techniques from Moscow with British statistical programs. During his appointments with the Agricultural Institute of Bydgoszcz and the Nencki Institute in Warsaw, Poland (1921-1935) practical statistician Jerzy Spława-Neyman was concerned with the unexamined dimensions of agrarian population analysis—how to measure unknown potential yields of sugar beets—and how blind commitments to mathematical laws such as the law of random errors impacted empirical results. Neyman's own commitment to the critical examination of statistical sampling techniques resonated with Egon Pearson at the University of College London. Together they developed confidence planning in order to test the limits of certainty and uncertainty in statistical sampling. Significantly, these models quickly proliferated through various institutional and industrial contexts. Throughout Poland, England, and the United States economists, statisticians, bureaucrats, and corporate leaders partook in the celebration of these new statistical models for in them they saw unprecedented accuracy in planning the future.