WEAI/AERE 2012 - Individual Paper Abstract
Title: Substitution with Biased Technological Change: the Dairy Product Industry in the United States
Author(s): Wei ZHANG,Department of Agricultural & Resource Economics, University of California, Davis, 2158 Social Sciences & Humanities Building, Davis, CA 95616, USA, 530-219-8472, 530-752-5614, weizhang at primal dot ucdavis dot edu; Julian Alston, University of California, Davis [Picture credit: Wei Zhang]
Abstract:
This paper examines the rate and biases of technological change in the dairy processing and manufacturing industry. We incorporate the distinctive characteristics of the industry into our theoretical modelling of technological change. Using an econometric approach, we obtain estimates of the elasticities of substitution between energy and other inputs (in particular milk, as well as capital, labor, and other processing materials) and measures of biases of technological changes in the dairy product manufacturing industry.
This research is motivated by our concurrent project on quantifying the effects of California's greenhouse gas cap-and-trade program on the dairy processing and manufacturing industry. Policies aimed to mitigate greenhouse gas emissions or to reduce energy consumption often lead to higher energy prices. An increase in energy price will affect energy use and the consequent carbon emissions through four channels: (1) input substitution, (2) induced energy-saving innovation (Hicks, 1932), (3) induced changes in the mix of products produced, and (4) induced changes in the scale of production.
Our model of the U.S. dairy processing and manufacturing industry incorporates the distinctive features of the industry. First, the industry faces a complex menu of government policies. An important economic consequence is that minimum processor prices are set such that uid milk plants pay a higher price for farm milk than do other types of dairy processors. Second, marketing and processing cooperatives play a significant role in the dairy manufacturing industry. Different from profit-maximizing farms, dairy cooperatives have a priority to process the total milk supply of their members. Third, the dairy manufacturing industry is a multi-product industry, with different products having different energy intensities. Moreover, products are related differently in production. Most dairy products compete for milk in production, but butter and nonfat dry milk are complements in production since they use different components of milk.
Alternative approaches have been used previously to specify technological change in econometric analysis. One approach is to model technological change as an extra variable in the production or cost function, such as a time trend (Binswanger, 1974). However, the rate and biases of technological change are not directly observable. In this analysis, we adopt the econometric method developed by Jin and Jorgenson (2010), representing the rate and biases of technological change by latent variables. The values of the latent variables are recovered by applying the Kalman filter (Kalman, 1960, 1963). Different from the price function approach used by Jin and Jorgenson (2010), we model the production of the dairy processing and manufacturing industry and extend the analysis to multiple products.
For the empirical implementation of the model, we take data from the Manufacturing Industry Database maintained by the National Bureau of Economic Research. The database covers all 6-digit 1997 North American Industry Classiffcation System (NAICS) manufacturing industries from 1958 to 2005, including six dairy processing and manufacturing industries. We also take data from the California Manufacturing Cost Annual, a publication of the California Department of Food and Agriculture.