1992 John Rust "Optimal Replacement of GMC Bus Engines: An Empirical Model of Harold Zurcher" скачать в хорошем качестве

1992 John Rust "Optimal Replacement of GMC Bus Engines: An Empirical Model of Harold Zurcher"

John Rust “Optimal Replacement of GMC Bus Engines: An Empirical Model of Harold Zurcher” How do you estimate decisions that depend on the future? Firms and households constantly make forward-looking choices. Whether to invest, replace capital, retire, or continue operating depends not only on current costs, but on expectations about future payoffs. Estimating such dynamic decisions presents a formidable challenge. The value of waiting matters. Future uncertainty matters. And the decision rule emerges from solving a dynamic programming problem. John Rust showed that these problems could be estimated structurally using real-world data. Using detailed information from a bus company’s engine replacement decisions, Rust modeled the manager’s choice as an optimal stopping problem. At each point, the operator decided whether to continue using an engine or replace it, balancing rising maintenance costs against replacement expenses. Rust’s key innovation was computational: he embedded the solution to the dynamic programming problem inside the estimation procedure. This nested fixed-point algorithm made it possible to estimate fully dynamic, forward-looking models. The 1992 Frisch Medal recognized a breakthrough that reshaped applied econometrics. Dynamic discrete choice models — once seen as too complex to estimate — became operational. From this point forward, empirical economics could structurally model intertemporal optimization, opening new frontiers in labor, industrial organization, macroeconomics, and public policy.