Mankiw relies on “[A] 2006 study [by] the economist William C. Randolph of the Congressional Budget Office [who] estimated who wins and who loses from this tax. He concluded that ‘domestic labor bears slightly more than 70 percent of the burden.'”
Mankiw somehow avoids posting the URL of the Randolph study (it’s here). Perhaps Mankiw was concerned that someone might also read the portion of the same study that concluded: “Replacement [of the corporate income tax] by a tax on the worldwide capital income of domestic residents–a tax that achieves capital export neutrality–shifts the entire amount of domestic labor’s burden toward the domestic owners of capital. That shift increases the domestic resident capital owners’share of the burden by about 68 percent of the tax revenue. Worldwide, both labor and capital owners benefit slightly from an increased investment efficiency under that replacement tax, but the largest changes are in the redistribution of tax burdens toward the domestic owners of capital away from domestic labor, and away from foreign owners of capital toward foreign labor.”
Of course, this would undermine the attractiveness of Mankiw’s preference, a regressive excise tax such as a tax on gasoline.