This paper formulates and estimates a model of producer behavior for U.S. manufacturing 1947-71 that simultaneously identifies substitution elasticities, scale economies, and the rate and bias of technical change. A nonhomothetic, nonneutral generalized Box-Cox cost function is employed which takes on the generalized Leontief, generalized square-root quadratic, and translog cost functions as special or limiting cases. Total factor productivity is estimated parametrically rather than being computed as the residual of growth in outputs minus growth in inputs. We find substantial economies of scale and relatively little technological change.

That fluctuations in primary commodity export markets have significant impacts on the economies of the producing countries is intuitively clear and widely accepted. But there is little consensus on the direction and significance of these impacts on economic development. Fluctuations in the value of primary commodity production and exports have the potential for affecting the growth path through a number of channels. This paper is concerned with delineating these channels of influence and with establishing quantitative dimensions for their effects. As an illustration we focus on coffee in relation to the economy of Brazil. The relationships between the coffee producing sector and the national economy are established econometrically and integrated into a macro model of the Brazilian economy. Alternative scenarios for coffee production and exports illustrate the nature of the impacts throughout the economy and the potential effects on Brazilian growth.