Reply: The Concept Of Capital

Edward Denison and I agree that the correct theoretical concept of capital is to consider two capital goods equivalent if they generate the same real net revenue, defined as gross revenue minus variable operating costs measured at a fixed set of output and input prices. Although I showed in my book that the correct concept could be fully implemented for commercial aircraft and electric generating equipment, for other products I was able only partially to take operating costs into account. As a result, both Denison and I agree that my radical revision to the official capital goods deflators does not go far enough and is biased toward understating improvements in quality. Our disagreement comes down to research strategy: I believe that I have progressed partway toward the ultimate goal of implementing the correct concept, while he views such a full implementation as infeasible. As a result, he advocates a return to the traditional criterion of base‐period production cost, even though this yields price deflators that ignore improvements in performance (as for computers) and improvements in operating efficiency (as for successive generations of jet aircraft) made possible by technological advances that reduct the cost of production.