Structural Versus Reduced-Form Estimation of Optimal Stopping Problems

In this paper I examine several statistical, interpretive, and policy implications of reduced-form (probit or logit) estimation of optimal stopping problems. The discussion proceeds in the context of an examination of the timber harvest decision of nonindustrial private forest (NIPF) owners. For a large class of optimal stopping problems a reduced-form model which closely approximates the statistical performance of its structural counterpart is readily found. Still, failure to properly interpret the relationship between the reduced-form model and the underlying optimal stopping problem invites flawed econometric analysis and inappropriate interpretation of reduced-form coefficients. Copyright 1997, Oxford University Press.