Estimating Distributed Lags in Short Panels with an Application to the Specification of Depreciation Patterns and Capital Stock Constructs

Abstract

This paper considers the problem of estimating distributed lags in short panels. Though the N time series contained in a panel may allow for relatively precise estimates of identified lag coefficients, identification requires restrictions on the contribution of the unobserved pre-sample x's to the current values of y, and the shortness of panels focuses attention on this matter. We investigate two such restrictions. The first constrains the relationship between the presample and insample x's, while the second constrains the lag distribution itself. An example, which investigates empirically how to construct "capital stocks" for the analysis of rates of return, closes the paper.