Estimating Technological Change Using a
Stochastic Frontier Production Function Framework:
Evidence from U. S. Firm-Level Data
by
Abstract
This paper presents a methodology for estimating an index of technological
change using firm-level data in a stochastic frontier production function
model that takes into account time-varying technical inefficiency. In contrast
to the Solow divisia index approach, econometric estimation of the index
with panel data allows the researcher to separate technical progress from
the stochastic measurement error. Applying the econometric methodology
to a panel of 908 publicly-traded U. S. firms from the COMPUSTAT
database, we find evidence of a significant downturn in general technological
change for the period, 1970-1989, whereas the divisia index methodology
applied to the same data shows stagnation. When the sample is divided into
Manufacturing,
Services, and Miscellaneous categories we find that estimates
of technological change for the three groups display markedly different
stochastic behavior and that the Services group is the source of
the downturn.
Downloadable Copies
The paper is published in the Journal of Productivity Analysis,
Vol. 8, no. 4, Nov. 1997, pp. 431-446. Copies of the paper as submitted
to the Journal can be downloaded by clicking on the preferred file format
below. Of course, the usual copyrights apply.
Geoff Gerdes,
Last
Update: November 11, 1998