Empirically Testing Predictions of an Attrition Warfare Model for the War in Ukraine
The goal of this study is to empirically test hypotheses about wars of attrition by evaluating their predictions for the conflict in Ukraine. Evaluation will occur after the war is over and authoritative data sources become available for analysis. This pre-registration document presents two quantitative hypotheses that make opposite predictions about the course of the War in Ukraine: (1) the Economic Power hypothesis, which predicts a win for Ukraine and (2) the Casualties Rates hypothesis, which predicts a win for Russia. Additionally, I consider an alternative hypothesis, according to which the outcome will be determined by random unforeseen events. The document includes four main parts: 1. An introduction providing the conceptual background and the rationale for this study. 2. The mathematical framework and a computational model that incorporates both Economic Power and Casualties Rates hypotheses as special cases. 3. An analysis plan that defines model outputs (what is predicted) and model inputs (parameter values and initial conditions), which need to be estimated from data. 4. An interim assessment (as of Summer 2023) using non-authoritative sources illustrating how, after the end of the war, input parameters will be estimated and the accuracy of predictions assessed. At the time of pre-registration (November 2023) the conflict is still unresolved. Neither side has made significant territorial gains for over a year (since the late Fall of 2022). Furthermore, no authoritative source for data, needed to accurately estimate inputs, is currently available. Estimates published in the press differ wildly depending on the source. As a consequence, the alternative predictions discussed in the interim assessment should not be taken as predicting the future course of the conflict. They instead are meant to demonstrate how these specific scientific hypotheses about war dynamics will be assessed after the war concludes.