A few days ago I linked to a couple of articles on VAT tax proposals that have been circulating, including an attack by Daniel Mitchell at the Cato site, and a short blog post from Greg Mankiw explaining why, as a replacement for the rest of the tax system, he thought it was a better tax mechanism, as well as being the functional equivalent of certain versions of the flat tax.
I received several interesting emails from tax professors in response. One pointed to a paper very much on the point of the post by Brian D. Galle, Hidden Taxes, upcoming in Washington University Law Journal; the SSRN abstract says:
The idea of hidden taxes is as old as John Stuart Mill, but convincing evidence of their existence is new. In this Article, I survey and critique recent studies that claim to show that there are some taxes that can go unnoticed by those who pay them. I also develop the array of unanswered theoretical questions and policy implications that potentially follow from the studies’ results.
Probably the central question for hidden taxes is whether they might enable government to raise revenue without also distorting the economy. If so, I argue, they have the potential to radically refashion the architecture of redistributive government. But, as I also show, whether that is true turns on the cognitive mechanisms that might permit taxes to go unnoticed. For example, if hidden taxes are caused not by rational ignorance but by cognitive shortcomings, then it is likely that the burden of a hidden tax will be borne disproportionately by poorer taxpayers, and vice-versa. Thus, I attempt to integrate with the tax literature some recent developments in our understanding of bounded rationality in consumers more generally.
But I also received an email from a friend and colleague on my own Washington College of Law faculty, tax professor Benjamin Leff. (Ben is a junior – ie, untenured – professor, and he had some hesitation about putting out views on a blog. I assured him that people understand that this is informal, first draft discussion, not a final academic or scholarly product, and moreover, it is a space to think about the political ramifications of various policy positions, in ways that one might not think appropriate for a scholarly paper.) I’m delighted to say that Ben is letting me put up his remarks as an embedded guest post, and my thanks to him for taking this up:
Your post on why the “hiddeness” of a VAT tax is “a bug, not a feature” was very interesting. Basically, if I understand correctly, you’re dipping into an old argument about the relevance of tax “salience.” The argument you’re making is that a less salient tax (a more “hidden” one) creates a public choice problem, because it enables policy-makers to tax more with less protest from the taxed. The implication is that if the people fully felt the sting of the taxes they pay, they would do the hard work of cutting government spending down to optimal levels, rather than overspending as they currently do (if they do). Thus, if a VAT were passed, and if it had the benefit of being less salient than current taxes, it would permit additional spending by the government.
I think the most common answer given to your argument currently is that the discovery of limitless deficit spending put an end to plausibility of the “starve the beast” argument you’re making. In other words, the link between taxing, voting, and spending that you propose is broken by the option of neither taxing nor reducing expenditures. That seems convincing to me, but I have no special knowledge about it. In recent memory, tax cutting has not generally been accompanied by reduced government spending, but obviously that doesn’t really prove anything. At root, it’s an empirical question: if a VAT were introduced, would the government use the revenue generated to (1) reduce non-VAT federal taxes (keeping overall revenue neutral); (2) reduce the deficit; or (3) increase federal spending. That question may or may not have anything to do with how “hidden” a VAT is. It may have more to do with the public debate that supports the imposition of a VAT, the intentions of Congress in enacting it, and the continuing commitment in Congress to whatever choices made with regards to spending and deficits.
But, more important than whether the argument is convincing or not, I think, is some context for it. You describe “hiddeness” as a feature that makes a VAT “particularly special” among tax mechanisms, but that’s not really true. We currently have a broad range of “salience” in the federal taxes that are imposed. What is especially problematic from a public choice perspective in the current system is that the “hidden” taxes appear to be disproportionately borne by wage earners. Therefore, as wealth increases (generally), one’s sense of being taxed increases more sharply than one’s actual tax burden. That is, (again, generally) rich people think they’re more taxed than they are and working people think they’re less taxed than they are. That’s a distributional public choice problem, and I think it should be clear why a distributional public choice problem would do more damage to the political process than a general one.
The “hiddeness” of wage-earners’ taxes is generally caused by two phenomena. First, wage-earners pay flat payroll taxes (generally social security and medicare), which is 7.65 percent of their income right out of their paycheck. This largely invisible (as evidenced by the fact that some of your commenters said that the bottom half of the country pays no federal tax, when in fact the vast majority of that bottom half pay a relatively steep flat tax on their very first dollar earned). But it’s not completely invisible, because at least it shows up on their paychecks and decreases their stated wage. But payroll taxes are even more invisible than that because employers are paying an additional 7.65% on their employee’s wages that doesn’t even get reported to them. There may be some argument about what the incidence of that tax is, but the consensus is that it falls at least substantially on labor. In other words, every employee in the country pays a flat 15.3% tax on their first dollar earned. Commentators often ignore it or are confused about it (especially conservative commentators who want to claim, falsely, that working people don’t pay federal taxes). Oops, I forgot to mention that it’s not a flat tax. It’s a regressive tax, because (at least for the social security component) once you earn above a ceiling ($106,800) the tax disappears.
The second factor that “hides” the taxes paid by working people is withholding. Because of withholding, wage earners often experience taxes as a refund, rather than an expense. When it was introduced, the biggest argument against withholding was exactly the argument you are making – that it’s a public choice problem to hide taxes. Many would argue that the biggest reason why our tax system can work at all is because of the withholding system. So, if you’re afraid of hidden taxes, the game has already been played (at least for the vast majority of Americans who are employees). Then the question if you’re still committed to “visible” taxes is whether a VAT is more or less hidden that withheld wages.
Thus, there’s currently a distributional problem with the federal tax system, because high-income taxpayers generally pay visible taxes, while low/middle-income taxpayers generally pay invisible ones. Because a VAT taxes only consumption, and exempts income from capital, it is yet another “hidden” tax primarily on wage earners, exacerbating the distributional salience problem that already exists in our current system. But if you think that the point of an income tax is to roughly measure “ability to pay” (as I and other liberals generally do), then you will be unhappy with a VAT not primarily because it is more or less “hidden” than current federal taxes, but because it actually increases the tax burden on wage earners while decreasing the burden on the wealthy. My view is that because a VAT excludes from tax income from investments, an income tax does a better job of tracking “ability to pay,” which is the cornerstone of an equitable tax system. But that discussion is well beyond the scope of your post.
All that to one side, though, what I think makes your post interesting to think about has to be the sharpness of the opposition of the “public choice” argument to the “economic” one, because your readers may be drawn to both. The way you cast it, the public choice argument seems infuriating to economists (or anyone who cares at all about economics), as you point out, because (generally) everything that makes a tax “efficient” also makes it less visible. And so, under the public choice argument – assuming that you thought that government spending is bad – the best tax would be the least efficient one. The more a tax changed market choices, the more it would “sting” (by thwarting one’s desires) and therefore, the more likely it would be to encourage the populace to reduce taxes. That should be true of spending programs too, by the way. If you’re opposed to government spending, then the worst possible thing is efficient government spending that really makes people’s lives better. You should be promoting wasteful spending that messes people up as much as possible. That is to say, if you’re looking for a revolution, make the current system work as badly as possible.
But on reflection, the economic argument and the public choice argument are not actually so opposed. Because, according to the economic argument (at least the welfarist economic argument), the government should do what it is most efficient for it to do to provide for the greatest happiness. So, if many things are public goods, for example, which are likely to be undersupplied by the market, then as an economist, you should not be for reduced government spending, but for spending sufficient to supply those public goods. You should be for exactly the right amount of government spending. It’s not a foregone conclusion, then, that we have excessive government spending, though we may be spending on the wrong things. Then you have a much more nuanced “salience” question. Taxes should be exactly “hidden” enough to permit people to make the right choices about how much government spending there ought to be. You have a problem of baseline, though. What is “the right” amount of hiddeness?
(Corrected, and thanks to commenter for pointing it out, to shift the last paragraph from Galle’s abstract from Ben’s discussion, where it wound up accidentally, back to Galle’s abstract.)