The Bush Tax Cuts: How are the distributional effects measured?
A number of methodological choices have to be made in estimating the distributional impact of a tax cut. These decisions include determining who actually bears the burden of a tax, how to count the income of households, and which metric of the distributional impact itself to use. Different choices can lead to quite different results.
- Identifying who is made worse off by a tax is more complicated than it may initially appear, since the person or entity who pays the tax to the government may not be the one who ultimately bears the burden. For example, taxes on a business eventually burden some individuals, but that may be the customers, workers, or suppliers rather than the business owners, if the business can shift the cost of the tax on to them.
- Economists have developed some reasonable conventions on these issues. For example, it is assumed that the burden of the individual income tax is borne by the payer and is not shifted to anyone else. By contrast, it is assumed that the corporate income tax is borne by recipients of capital income generally, not just shareholders in the corporations subject to the tax. The reason is that investors can shift their funds to the most profitable venture with relative ease; a tax on one form of capital thus affects the after-tax return to all forms of capital, because investors constantly compare such after-tax returns and reallocate their investments accordingly. The estate tax is assumed to be borne by those whose estates actually have to pay the tax. Workers are assumed to bear the burden of the payroll tax, including the share nominally paid by the employer, since firms are typically able to shift the cost onto workers by depressing wages.
- As a standard, comprehensive measure of income, and thus of a household’s well-being, the Urban-Brookings Tax Policy Center (TPC) uses a measure called "cash income," which comprises adjusted gross income (AGI, a measure frequently used for tax purposes) plus a variety of additional forms of income including cash benefits from public programs, interest on tax-exempt bonds, contributions to pension plans, the component of Social Security benefits not included in AGI, and some other items. It is important that the measure of income include not just wages and salaries, but also capital gains, dividends, interest, and business income, because these categories account for far more of the incomes of high-income than of low-income households. Excluding such capital income could thus lead to a biased result.
- Another important choice is the metric used to assess the distributional impact of a tax change. TPC’s preferred measure is the percentage change in after-tax income. If a tax cut changes everyone’s after-tax income by the same percentage, the distribution of after-tax income remains the same as before.
- Another issue is which taxes to include. TPC’s estimates include a wide range of federal taxes, including taxes on individual and corporate income, payroll, and estates. It is important to use a broad measure of federal taxes, because different taxes are distributed differently across income levels. The estate tax and the corporate tax, for example, tend to fall more heavily on those with high incomes - that is, they are more progressive. Conversely, the payroll tax that finances Social Security is only paid up to a certain level of income, so it tends to fall more heavily on those with low and moderate incomes and is therefore less progressive than the income tax.