I set forth a generalized stochastic time trend approach, based upon the Kalmanfilter, as an alternative to the general index approach to measure technologicalchange. Technology is treated as a latent variable in a state-space model of theproduction function. In data sparse settings, where panel data are unavailable,the method provides results which encompass insights from the general indexapproach, but provides more detailed estimates. I revisit an analysis oftechnological change in the Lofoten fishery. The estimated technology timeprofiles agree to some extent between the methods, but my more detailedresults demand a new historical interpretation.
展开▼