The Rivlin-Domenici Deficit Reduction Plan Is Not As Progressive As It Appears

Our guest blogger is Michael Linden, Associate Director for Tax and Budget Policy at the Center for American Progress Action Fund.

The New Republic’s Jonathan Chait writes this morning about how he might have been a little too quick to jump into the seemingly open arms of Erskine Bowles and Alan Simpson, the co-chairmen of President Obama’s deficit commission. But now there is a new bipartisan deficit reduction plan out from Commission member Alice Rivlin and Pete Domenici at the Bipartisan Policy Center, and Chait thinks this could be the one. He seems to especially like Rivlin and Domenici’s approach to the tax code, saying, “the tax reform, while lowering the corporate and top income tax rates to 27%…also makes the overall burden more progressive.”

Well, I take no pleasure in being the bearer of bad news. The Tax Policy Center’s analysis of the BPC plan does suggest that their reforms would result in a more progressive system, but there is one really big problem –- the VAT. Rivlin and Domenici rely on a 6.5 percent value-added tax which they call a “debt reduction sales tax.”

Now, generally speaking, VATs are thought to be very regressive –- that is poorer people pay more in taxes, as a proportion of their income, than rich people do. The reason is simple. Poor people tend to spend all of their income in any given year, whereas rich people do not. If you pay a 6.5 percent tax on every dollar you spend, and poor people spend every dollar they earn, then they are paying the tax on 100 percent of their income.

Rich people, by contrast, only spend a portion of what they earn, and so they only pay the tax on a fraction of their income. But, that’s not how the Tax Policy Center models value-added taxes.

Instead of treating a VAT as a tax that is paid as income is spent, they treat it as a tax that is paid as income is earned (technically, they treat it as a tax on wages and existing capital). The idea here is that, after the VAT is introduced, every new dollar you earn will be worth a little less, because the consumption that you’ll eventually use that dollar for is going to be 6.5 percent more expensive. In other words, economically speaking, the VAT reduces the value of income at the time of earnings, because eventually those earnings will be used to pay the VAT.

The consequence of treating the VAT this way is a much less regressive-looking distributional analysis. Instead of showing poor people paying the tax on all their income, and rich people paying the tax on just some of their income, the burden of the VAT essentially falls on everyone’s total income each year.

To see what a huge difference this makes in terms of distributional analysis, take a look at this chart from a paper by Len Burman (who’s on the Rivlin/Domenici commission), Jane Gravelle and Jeff Rohaly (who actually helped produce TPC’s estimate of the commission’s plan):

The chart shows the average tax rate, by income percentile, of a 20 percent VAT, but using several different distributional methodologies. The first column, titled “Consumption” is how people usually think about a VAT, and the second column, “Wages and Equity” is essentially how TPC models a VAT. Under the first method, the VAT looks really regressive, with tax rates declining precipitously as income rises. But the second method makes the VAT actually look a little bit progressive! Big big difference.

The problem is that while the TPC method is may make sense academically, that is not how people will actually experience the VAT. That’s not to say that the consumption method is perfect, but the fact is that in any given year, poorer and middle class people will be paying a much higher average tax rate than rich people will. True, over the course of a lifetime, things may even out somewhat, but that is very cold comfort to those who will be paying the tax year-to-year.

The bottom line is that the way people will actually experience the effects of the Rivilin/Domenici plan in any particular year is very different from that suggested by TPC’s distributional analysis. My suspicion is that if they produced an analysis in which they treated the VAT as a tax on consumption rather than as a tax on income, the overall skew would be much less progressive.