Energy Substitution in U. S. Manufacturing: A Regional Approach

Abstract

One response of economists to energy-market developments of the 1970s was to launch a series of studies in which the objective was to estimate energy substitution parameters characterizing important sectors of the U.S. economy. Relatively high substitution elasticities between energy and other inputs would permit firms to mitigate the effects of energy price increases by shifting to less energy-intensive factor proportions. The seminal work in this effort was carried out by Berndt and Wood [5], Griffin and Gregory [14], Pindyck [28], Halvorsen [16], and Fuss [12].'The research was substantially abetted by developments in production theory, where the work of McFadden [23], Diewert [8], and Christensen, Jorgenson and Lau [6] yielded new and flexible production and cost function forms. Empirical work on energy substitution has continued.2 For the most part the data analyzed have been highly aggregative, applying to, say, total U.S. manufacturing, or perhaps to the various two-digit sectors comprising that total. To date little effort has been given to investigating energy-substitution relationships at the regional level in the U.S. This is unfortunate since we know that the U.S. economy is actually composed of regions which are differently endowed with resources, have developed at different rates, and are comprised of different sectoral mixes. The distinct possibility exists that regions differ in terms of the way in which energy inputs relate to other inputs in production. Furthermore, even though the energy crisis may seem to be a national phenomenon, national policy responses are worked out within a context of shifting and colliding regional conflicts, stemming in large part from real or imagined perceptions of how energy market developments impact different regions. It would seem especially important, therefore, to identify