One of the fundamental decisions in tax policy is the choice of the tax base, with the broadest decision being whether to tax income or tax consumption. Income taxes are levied on what people contribute to the economy – their skills, time, and overall labor, as well as their savings to finance investments in innovations, skills, and physical capital in the economy. My preference is to not punish these contributions, but rather to tax consumption. Consumption is what people take out of the economy, and raising the price of consumption is preferable as a tax policy framework.

That does not mean just any consumption tax is perfectly fine. Indeed, one of the worst-designed taxes to rear its ugly head is the FairTax, which has been around for over 20 years. The FairTax would replace the existing federal revenue system – individual income tax, corporation income tax, payroll taxes, and estate and gift taxes – with a single, national sales tax (which is one form of a consumption tax). This has two immediately appealing features: its apparent simplicity and potential economic efficiency. A sales tax is one form of consumption taxation; moving the United States toward a reliance on consumption taxation would improve long-run incentives for saving, capital accumulation, output growth, and productivity growth.

But, as my co-author Gordon Gray and I note in our new study of the FairTax: “The FairTax also comes with a list of immediate concerns: Sales taxes are notoriously regressive (although advocates will argue that FairTax design features address this), a federal sales tax would poach on a tax base important to many states, the tax rate would have to be quite high to replace existing federal revenue, there would be strong incentives to evade the tax, and when introduced prices would rise enough to accommodate the tax (following the standard assumption that the Federal Reserve accommodates FairTax by increasing the money supply). In an economy currently enduring the highest inflation in four decades, this latter feature is especially problematic at this moment.”

Why study something that looks like a non-starter? There are news reports that the most recent legislative version of the FairTax, the FairTax Act of 2023 (H.R. 25), may be taken up for consideration by the House of Representatives. Accordingly, the American Action Forum sought to identify better the empirical magnitude of the impacts. It retained EY’s Quantitative Economics and Statistics Group to conduct the analysis.

The study and detailed appendices supplied by EY are chock-full of important findings. Here are the highlights. First, the (only) good news is that if (and I emphasis if) the economy, budget, and tax system survive long enough, the shift from consuming to greater saving and investment will produce a larger economy (more gross domestic product). A consumption tax can’t turn out any other way.

The bad news is that it takes a really large tax to generate enough revenue. If one manages to keep the broadest tax base and suffer minimal non-compliance (15 percent “cheating”) the rate has to be 49 percent of the price. Ouch! But if Congress acts like states do when running their sales taxes, the base will narrow and the rate could be as high as 387 percent.

Even worse, when the tax is first introduced, sellers will have to raise their prices by enough to cover the tax. How does a 49 percent price rise sound right now? 387 percent? Simultaneously, households will scale back their consumption sharply and it will stay depressed for as much as the subsequent 10 years. At least to my eye, it seems unlikely that the FairTax will survive long enough to generate the economic benefits sought by consumption tax advocates.

The modeling is quite sophisticated and the study walks through many issues in tax policy. But this summary gives you the flavor of the bottom line. No thanks.