The Short-Run Shifting of the Corporation Income Tax: A Simultaneous Equation Approach

Abstract

THIS paper centers on an empirical analysis of the short-run incidence of the corporate income tax. The topic is made interesting by the existence of two major conflicts. First, traditional microeconomic theory suggests that firms will exhibit no short-run response to variations in the corporate tax rate; yet, businesses (whose behavior theory is supposed to describe) generally claim that some response does take place. Second, statistical estimates of the degree of short-run tax-shifting conflict. A study by Musgrave and Krzyzaniak (1963) suggests that firms are successful in avoiding the entire burden of the tax; however, Gordon (1967) concludes that firms bear the entire burden in the short run. The Musgrave and Krzyzaniak (M-K) study has been criticized on the grounds that it did not take sufficient account of cyclical variables that may have an impact on profit rates. Goode (1966), Slitor (1966), and Oakland (1972) have attempted to demonstrate that the absence of demand-pressure variables causes an upward bias in the M-K estimate of tax shifting. Nevertheless, Dusansky has used two-stage least squares to estimate a profit equation in which such influences are incorporated, and has obtained results consistent with the M-K conclusion of full shifting. One of the apparent shortcomings of the studies mentioned above is that they seem to be either unaware of or disinterested in the mechanism of tax shifting.1 The studies consist of the specification and estimation of profit equations in which a tax-shift parameter is specified. But this technique yields no information with regard to the tax-effects on the decision variables of the firms in question. Traditional tax literature considers two possible means of shortrun shifting:2 forward-shifting (accomplished through variations in product-prices) and backward-shifting (carried out through changes in factor-prices). Obviously, non-zero overall shifting requires that either forward-shifting or backward-shifting (or both) is carried out. Thus, the Musgrave and Krzyzaniak conclusion of full overall shifting implies that product-prices and/or factor-prices must be sensitive to variations in the tax rate. But, because of the use of the reduced-form estimation procedure, one cannot even begin to translate their findings into a reasonable guess concerning which of these prices are affected. And Gordon's conclusion of zero shifting could conceivably be the result of offsetting changes in product-prices and factorprices. It seems plausible to suggest that one should be concerned with several dimensions of taxresponse, rather than merely the end (net) result. In a later exchange between the authors of the first two studies cited above,3 Musgrave and Krzyzaniak take up this point by suggesting that the most reasonable approach to the controversy would involve the specification and estimation of a structural model in which price, wage, and shifting behavior are (1968, p. 1360). This paper is, in essence, a response to that suggestion. In section II, an estimated model is presented in which the potential vehicles of short-run shifting (prices and wages) are specified as endogenous variables. This treatment allows us to trace through the short-run effects of variations in the tax rate. In sections III and IV the effects of the tax on the endogenous variables are computed from the reduced form of the system, and these results are translated into measures of overall shifting and wageand price-responses. Then, in section V, conclusions are drawn. Received for publication September 18, 1975. Revision accepted for publication April 20, 1978. * San Diego State University. Thanks for financial support are due to the Center for Public Economics at San Diego State. Another objection is that both papers tend to be ambiguous with respect to the rationale behind shifting behavior. This point is taken up in my previous paper (Sebold, 1970). 2 Not all types of response to the tax need be categorized as ,'shifting. This point will be discussed later in this paper. 3 See Gordon (1968) and Musgrave and Krzyzaniak (1968).