Econometric modeling of industries: Semiconductors and internal combustion engines

In this dissertation a small-scale econometric model of an industry is developed as a tool to analyze the impact of government purchases and other exogenous variables on the output, capital expenditures, and employment of the semiconductor industry in the United States. The model is also applied to the internal combustion engine industry in order to examine its performance on an alternative industry. The model consists of an equation for the industry's output, an investment and a manhour equation for the industry's input, and an equation for the shipments of the industries which use the industry's outputs as an intermediate input. There are also two identity equations: one estimates demand for the industry's shipments by utilizing the input-output technical coefficients, and the other calculates the industry's capital stock by aggregating the annual capital expenditures.;An historical simulation is conducted for both industries on a microcomputer to test the accuracy of the model over the estimation periods. The model is used to answer counterfactual questions such as: what would have happened to the semiconductor industry in case of reduction in government defense purchases? What would be the effect of an economic slowdown on the industry? What would be the effect of subsidized loans on the industry? The results indicate that government defense purchases have significant impact on the information industries, which in turn influence the semiconductor industry. Government defense purchases, however, do not show any significant effect on the shipments of the primary purchasers of the internal combustion engines. It is also found that the semiconductor industry, in recent years, is less dependent on the government defense purchases compared to earlier years, and the industry sensitivity to the non-government sectors of the economy has declined. It is further shown that subsidized loans stimulate investments in this industry.