To access the full paper, download the PDF on the left-hand sidebar.

Introduction

Consumption taxes are the third largest source of government revenue in the U.S, generating nearly as much revenue as corporate income and property taxes combined and exceeded only by personal income and Social Security taxes. In 2016, consumption taxes accounted for 17 percent of all federal, state, and local government taxes collected, providing roughly 3 percent of total U.S. GDP in revenue. Each day consumption taxes are paid millions of times with the purchase of most consumer goods. Though its impact on equity has been extensively disputed and discussed, much is left to discover regarding the progressivity of consumption taxes.

One difficulty in evaluating consumption taxes, such as sales and excise taxes, results from the lack of data matching the taxpayer with the tax payment. Usually, buyers pay the tax, while sellers collect it for the government. Although the incidence of the tax may fall more heavily on either the buyer or the seller, equity of the buyer-borne tax burden is difficult to assess because very little is ever known about the buyer. In contrast, when individuals pay income taxes to the federal government, extensive paperwork documenting individuals’ unique economic circumstances are documented in submissions to the IRS, allowing researchers to measure certain features, including tax progressivity.

Since the progressivity of the sales tax is difficult to directly measure, this paper introduces an indirect approach combining simulated household income with realizations of consumption behavior from survey data. This approach circumvents two issues regarding the measurement of consumption tax progressivity. First, the simulated approach overcomes the anonymity of buyer information by using granular consumption expenditure data. The data pairs uniquely identified individuals with a constructed consumption tax base that can be modified to replicate the composition of any existing tax jurisdiction’s specific consumption tax base. Second, as discussed below, annual consumption tax progressivity provides a poor indication of an individual’s actual tax burden. To overcome this limitation, the quantitative strategy involves simulating both lifetime income and lifetime consumption to provide a measure of lifetime consumption tax progressivity.

Economist James Poterba first proposed measuring consumption tax progressivity from a lifetime prospective as he evaluated lifetime excise tax burden. This paper contributes to that concept by presenting a lifetime sales tax progressivity simulator, which is applied to a sales tax base constructed for illustrative purposes. The results of the simulation applied to this tax base show that the total lifetime sales tax bill increases with lifetime income, but lifetime sales taxes are slightly regressive.