The Welfare Cost of Monopoly

Harberger's first contribution to applied welfare economics estimated that the efficiency loss from monopoly—or, more precisely, from prices exceeding marginal cost—in US manufacturing was unlikely to exceed 0.1% of GDP. In making this estimate, Harberger assumed that the extent of market power of the firms in an industry is reflected in the degree of above-normal profits in that industry. The efficiency loss from monopoly arises from the fact that high-profit industries produce too little output, and low-profit industries produce too much output. The net welfare cost is the sum of the gains that would be obtained if output were reallocated in such a way as to bring the rate of profit in each industry to the economy-wide average. To do this requires an estimate of the elasticity of demand for the output of each industry in his sample. He argued that a value of -1 for this elasticity was on the high end of plausible values, because each industry consisted of more than one firm. (The elasticity of demand facing a particular firm is inversely related to its market share. Harberger's assumption clearly implies that individual firms face elasticities far greater than -1, and is thereby fully consistent with the theory of pricing by firms with a degree of market power.) For Harberger's assumed value of demand elasticities, the cost of the resources that would be required to achieve the efficient change in an industry's rate of production is equal to the calculated excess profits in that industry.

Using the formula developed by Harold Hotelling,^[8] Harberger measured the efficiency loss from noncompetitive pricing in each industry as the excess of the implied willingness to pay by consumers for each increment to output over the estimated cost of that increment. His conclusion was that monopoly power was not pervasive in the U.S. economy, so that the manufacturing sector could be treated as nearly competitive.

Over a decade later, Gordon Tullock argued^[9] that Harberger's estimate of the welfare cost of monopoly was low because some resources would be used to compete for or protect monopoly rents, but his argument applies principally to instances of non-price competition for protected status, as in the case of tariffs or restrictions on entry into an industry through government regulation.

The Incidence of Taxation

Harberger's seminal paper on the corporate income tax pioneered the use of general-equilibrium modeling by recasting the classic Heckscher-Ohlin model of international trade as a model of one country with two sectors, one comprising incorporated firms subject to a tax on their net incomes and the other made up of unincorporated firms. Harberger's approach laid the groundwork for the later use of computable general equilibrium analysis of the impact of taxes on an entire economy.

In Harberger's model, the output of both sectors is produced under conditions of constant returns to scale, using homogeneous labor and capital. Labor is perfectly mobile, so wages are equalized between the two sectors. The corporate income tax is a tax on the return to corporate capital. There is no integration of the corporate and personal income taxes so any dividends are taxed twice. In the long run, capital is also fully mobile between sectors.

Harberger's key insight was that the mobility of capital between sectors means that in long-run equilibrium the after-tax rate of return to capital is equalized between the two sectors. In this way, the corporate income tax lowers the after-tax real rate of return to all owners of capital equally. When the income of capital owners in the corporate sector is taxed, the initial after-tax rate of return on corporate capital falls, prompting a shift of capital to the untaxed sector. This shift continues until the rate of return in the untaxed sector falls, and the before-tax rate of return in the corporate sector rises, by amounts sufficient to equalize the after-tax rate of return to capital in all uses. The contraction of the corporate sector leads to a reduction in that sector's demand for labor and a fall in the output of that sector. Some labor is thereby induced to move from the corporate to the non-corporate sector. The flow of labor and capital into the non-corporate sector results in an increase in that sector's output. The change in the relative quantities of output in the two sectors leads to an increase in the price of the taxed sector's output relative to the output of the untaxed sector. The response of wages to the shifts in production depends on several factors: the relative substitutability of labor for capital in both sectors, the relative labor intensity of both sectors, and the relative sizes of the two sectors. If the net effect of the intersectoral shifts is to reduce the total demand for labor, then the wage falls and workers bear a part of the burden of the tax. If the aggregate demand for labor is unchanged, then capital bears the full burden of the tax. If the aggregate demand for labor (and so the wage) rises, then capital's income falls by more than the revenue raised by the tax. This latter case, which Harberger showed to be not at all unlikely, completely overturned the previous consensus within the profession, which had been that the burden of the corporate tax was distributed among the workers and capital owners in the taxed sector, and the consumers of the goods produced in that sector, in amounts that were individually between zero and 100 percent of the tax revenue.

Some years later, Harberger extended his analysis to the case of an economy that buys and sells goods, and imports or exports capital, in worldwide markets.^[10] The results from this specification differ markedly from those of the closed-economy version. In particular, if externally provided capital is in perfectly elastic supply to a country, then capital cannot bear any part of the tax burden in the long run, and it is quite possible for labor to bear more than 100 percent of the burden of the corporate income tax. He views the open-economy case as relevant for the analysis of a single country's choice of tax rate, with the closed-economy model as applying to the effects of a general movement of worldwide corporate tax rates in the same direction, such as the decline that has been observed over the past several decades. A summary of his views on the practical policy considerations regarding the corporate income tax in particular countries can be found in his 2008 paper.^[11]