Content area
Abstract
The variance in functional measures of technological artifacts different from each other in kind is analyzed with an autoregressive stochastic model. An important feature of this model is that in the process of estimating its parameters it takes into account the moving-average structure of the error term its form implies. This capability is significant with respect to the analyses conducted in this investigation. Its exploitation generates results dramatically different from those generated by similar but less rigorous analyses.
The autoregressive form of this model derives from stated premises concerning the nature of technological change. Formalized, these premises permit two hypotheses to be tested, learning by doing/using and scale of utilization. Aircraft of the domestic trunk airlines and rail locomotives of the Class I Railroads of the United States railroad system embody their respective technologies and are the artifacts of this investigation. The variance in the functional measure of the aircraft is analyzed with the above two hypotheses from 1947 or 1952 to 1976. The results indicate that both hypotheses explain the variation over the estimation period of 1947 or 1952 to 1965 well. Furthermore, the estimates exhibit generality inasmuch as they prove stable outside the sample set in the prediction period, 1966-1976. The variance in the functional measure of rail locomotives is also analyzed for the period 1972 to 1984. Only the learning hypothesis is tested and no attempt is made to determine whether or not the parameters estimated are stable outside their sample set. The learning hypothesis explains the variance in the functional measure of rail locomotive technology well also. This is sufficient to permit a comparison of estimates to be engineered. The results of such comparisons determine that different technologies vary with respect to how responsive they are in terms of their artifacts' functional measures to variation in their respective arguments. This is suggestive for policy on productivity in which technology is generally acknowledged to prominently figure. This research implication, the link betwixt economic and functional measures, an agenda for refining and extending the application of the research program partially introduced with the analysis conducted etc. are also subjects of interest to which this investigation attends.
