Productivity, Scale Economies and Technical Change: Reconsidered

Abstract

Recently several writers have presented theoretical arguments and empirical evidence that suggests that an electric utility's long-run productivity growth rate can be decomposed into a variety of factors.' For example, Gollop and Roberts [12] decompose a utility's total factor productivity growth rate into three distinct components: (1) technical change, (2) a scale effect and (3) an environmental regulatory factor. Results for the 1970s indicate that declining rates of technical change and output growth coupled with increased regulatory intensity lead to a noticeable decline in the firm's long-run productivity growth rate. Similarly, Nelson and Wohar [15] examine the impact of technical change, scale economies, and rate of return regulation on an electric utility's longrun productivity growth rate. Evidence suggests that the technical change and scale components behaved in a manner similar to the Gollop and Roberts [12] findings. The evidence concerning the impact of the regulatory factor was less convincing due to the difficulty in obtaining reliable estimates of the regulatory constraint parameter. A common characteristic of this literature is that the firms under investigation are assumed to employ all inputs at their long-run cost minimizing levels. As such, the productivity decomposition model is primarily concerned with distinguishing between shifts in the production function versus movements along the production structure. According to Caves, Christensen and Swanson [5] a violation of the long-run equilibrium assumption adds an equally important factor to the decomposition of productivity growth, specifically the movement toward or away from longrun equilibrium. As a consequence, researchers must further distinguish between long-run and short-run productivity growth estimates.2 In a recent article [9], Daly and Rao illustrate theoretically and empirically (using a long-run cost function and data pertaining to Ontario Hydro) that a firm's long-run productivity growth rate can be decomposed into: (1) technical change, (2) non-constant returns to scale, and (3) a capacity