Here is the full text of my 63d Quartz column, “VAT: Help the poor and strengthen the economy by changing the way the US collects tax,” now brought home to supplysideliberal.com. It was first published on June 8, 2015. Links to all my other columns can be found here.

If you want to mirror the content of this post on another site, that is possible for a limited time if you read the legal notice at this link and include both a link to the original Quartz column and the following copyright notice:

© June 8, 2015: Miles Kimball, as first published on Quartz. Used by permission according to a temporary nonexclusive license expiring June 30, 2020. All rights reserved.

Despite a hard first quarter, someday soon the US economy may be in a position where it needs people to save more instead of spend more. Economists will start talking about the importance of having people save to provide funds for investment instead of the importance of having people spend to generate enough demand that investment is worthwhile. A few weeks ago, I pointed out in “How Increasing Retirement Saving Could Give America More Balanced Trade” how increasing the saving rate can also raise net exports, in a way that has a long-lasting positive effect on jobs. And I pointed out how a regulation making saving something automatic people have to opt out of, instead of something they have to opt into, could dramatically raise the saving rate.

A change in our tax system could also raise the national saving rate: shifting from taxing income to taxing only the part of income that is consumed, while exempting the part of income that is saved. The clean, well-tested way to tax consumption rather than income is to use a value added tax or VAT.

A VAT tax is like a sales tax on final sales to households that is collected gradually all along the way as goods and services are produced. It is structured so that the buyers in business-to-business transactions are motivated to check that the sellers are paying the tax—which makes it much harder to cheat on than other taxes.

Most rich, well-run countries other than the US use a value added tax. And though it is much harder to compare saving rates accurately across countries than one might think, most of those other countries seem to have higher saving rates than the US. In saying all of this about a VAT tax, I am only repeating the conventional wisdom, which I think in this regard is by and large on target.

But there is another aspect of the conventional wisdom about a VAT tax that I think is totally off target: the idea that a VAT tax hits the poor especially hard. Part of this misconception comes from the simple fact that measuring how progressive or regressive a tax is by the fraction of income paid in taxes at different levels of income is already sneaking in the idea that income is the right basis for taxation—exactly the question that is at issue. Measuring how progressive or regressive a tax is by the fraction of consumption paid in taxes at different levels of consumption would give a different picture.

The other part of this misconception is that typical measures of progressivity or regressivity ignore the other side of the ledger: the government assistance that people are given at different levels of income or consumption. Ignoring that side of the ledger slips in an assumption about how government assistance would be affected by a VAT tax that seems wrong to me. (Here I am counting Social Security and Medicare as government assistance since people don’t have individual Social Security or Medicare accounts and the government can change the level of benefits at any time.)

To think about how taxing consumption taxes rich and poor consider this question: “Who can afford to spend more than they earn from their job?” People who absolutely can’t afford a given level of spending will ultimately fall so deeply in debt that outside forces will stop them from spending so much. So they may pay more taxes on their consumption now, but will pay fewer taxes later. So it is the consumption people can afford that matters for consumption taxes over the long haul.

There are two basic ways you and your immediate family might be able to afford to spend more than you earn from jobs: have a pile of your own wealth to draw on for consumption or have someone else give you money to spend. Other things equal, having a pile of your own wealth to draw on for consumption makes you richer; so that side of things makes a consumption tax–such as a VAT–progressive. If someone else is giving you money to spend when you don’t really need it, that counts as being rich in a spongeing sort of way. Or to put it better, although ignoble, the ability to convince other people to let you sponge off of them is its own form of wealth.

On the other hand, if someone is giving you money to spend because you really do need it, they should realize that the amount of money you need has to be grossed up enough to get the same amount of goods and services even after paying the VAT. In particular, if the someone giving you money to spend is the taxpayer, through the government, whatever dire need motivated the taxpayer to help you out is a need for a given amount of goods and services, and the dollar value of the government assistance should of course be grossed up enough to buy the same amount of goods and services even after paying for the VAT. Since consumer price indices are usually calculated including VAT taxes, this could happen through the standard process of cost-of-living adjustments.

Notice that the government receives the VAT taxes paid on goods purchased with money from government assistance. So grossing up the government assistance to pay the VAT tax is just shuffling money from one government account to another and isn’t an unsustainable drain on the government budget.

Of course, the shift to a VAT tax could be used as an excuse to cut the amount of goods and services provided as government assistance. That would be regressive. But analytically that should be considered a shift to a VAT tax in the more neutral assistance-preserving way described above plus a cut in government assistance. The VAT tax itself should not be blamed for this effective cut in government assistance. But in the flawed accounting all too often used, the VAT tax is blamed for this unmotivated and far from inevitable cut in the effective level of government assistance.

What I have laid out is not the end of the story; there are many other issues in the transition to a value-added tax—for example, while lowering income taxes on 401(k) distributions would keep those who saved that way under the old tax system whole, something needs to be done to honor at least in spirit the promise to those who saved in a Roth plan that after paying taxes on that saving up front, they wouldn’t be taxed later on. But the basic story is that a value-added tax is progressive when the accounting is done right and the shift to a VAT tax is not used as an unwarranted excuse to cut the effective level of government assistance. This shouldn’t really come as any surprise, since governments in many countries that intend to do a lot more redistribution than the US use a value-added tax.