August 1, 2006
William C. Randolph
This study applies a simple two-country, five-sector, general equilibrium model based on Harberger (1995, 2006) to examine the long-run incidence of a corporate income tax in an open economy. In equilibrium, capital is assumed to be perfectly mobile internationally in the sense that the country in which a real investment is located does not matter to the marginal investor. In addition, each country is assumed to produce at least some tradable corporate goods for which the country cannot affect world output prices. Like the original Harberger (1962) model, the worldwide stock of capital and the supply of labor in each country are fixed. Under those assumptions, the model provides closed form solutions and easily understood predictions about its comparative static equilibria. As with any simplified model, the analysis is silent about some potentially important issues – such as the effect of the corporate tax on savings, growth and other dynamics – that may also have important effects on corporate tax incidence.
The analysis shows how the domestic owners of capital can escape most of the corporate income tax burden when capital is reallocated abroad in response to the tax. But, as in Bradford (1978), capital owners worldwide cannot escape the tax. Reallocation of capital abroad drives down the personal return to investment so that capital owners worldwide bear approximately the full burden of the domestic corporate income tax. Foreign workers benefit because an increased foreign stock of capital raises their productivity and their wages. Domestic workers lose because their productivity falls and they cannot emigrate to take advantage of higher foreign wages. Under basic assumptions of the numerical application, the outcome is also similar to the implications of the simpler model of Bradford in that the full worldwide burden falls on domestic owners of productive inputs. That outcome changes, however, under alternative assumptions.
Burdens are measured in a numerical example by substituting factor shares and output shares that are reasonable for the U.S. economy. Given those values, domestic labor bears slightly more than 70 percent of the burden of the corporate income tax. The domestic owners of capital bear slightly more than 30 percent of the burden. Domestic landowners receive a small benefit. At the same time, the foreign owners of capital bear slightly more than 70 percent of the burden, but their burden is exactly offset by the benefits received by foreign workers and landowners. To the extent that capital is less mobile internationally, domestic labor’s burden would be lower and domestic capital’s burden would be higher. Burdens can also be affected by the domestic country’s ability to influence the world prices of some traded corporate outputs. But the signs and magnitudes of those effects on burden depend upon the relative capital intensities of production in the corporate sectors that produce internationally tradable goods.