Wednesday, December 16, 2009

Tax-side vs. spending-side progressivity

Bruce Bartlett has a new post in which he says the following:

"I think we should simply give up trying to redistribute income on the tax side and accept that it can only be done meaningfully on the spending side. This would require both the right and left to give up some of their pet ideas. The left would accept that the only purpose of the tax system is to raise revenue and the right would accept that a fairly extensive social welfare state is here to stay. In essence, conservatives would raise the revenue and liberals would spend it. That's more or less the way it works in Europe, where conservatives accepted the welfare state in return for having it financed conservatively through a value-added tax. Liberals accepted this regressive form of taxation in return for conservatives accepting the legitimacy and permanence of the welfare state.

"Over the years, I have asked a number of liberal friends if they would take this deal."

Subject to some quibbles, such as concerning what it means for a program to be on the tax side vs. the spending side, count me in. For starters, I agree with Bartlett (per his comments elsewhere in the same post) that the biggest progressivity issue is helping people on the bottom, not leveling down at the top. This tends to follow, by the way, from a utilitarian social welfare norm if one uses what I believe are relatively plausible assumptions about the rate of declining marginal utility. Educated people on the left sometimes think more about the top than the bottom because they are acting out their own resentments and self-pity more than focusing on general beneficence.

As a further point, suppose we have the choice between a flat tax, which is somewhat progressive given its zero bracket, and a VAT that hits everyone at the same tax rate. The latter, however, would be paired with a demogrant (i.e., a uniform cash grant to each individual or household). If we keep the parameters otherwise sufficiently the same, the VAT plus demogrant is clearly more progressive.

To illustrate: Suppose we have a choice between (a) a 25 percent flat tax with a $30,000 zero bracket and (b) a 30 percent VAT plus a $9,000 demogrant. If you earn $30,000, you pay a net of zero either way. Nothing paid or received under the flat tax; $9,000 tax and offsetting $9,000 grant under the VAT + demogrant. But if you earn less than $30,000, you do better under the VAT + demogrant (reaping a net gain from their combined application). And if you earn more than $30,000, you do worse under it (since the tax now exceeds the grant).

The example helps to show that one could think of the flat tax's zero bracket as a variant of a demogrant in which grants are, for some odd reason, phased down for low-income individuals. The motivation for such a design feature is hard to fathom, once one starts thinking about it that way.

To be sure, graduated rates that go beyond the flat tax's two-rate structure (e.g., the David Bradford X-tax with higher brackets) provide more progressivity at the top than you can get via the VAT + demogrant, at least if you don't want the latter feature to grow too great. But this returns to the point of progressivity's relative importance at the top versus the bottom.

If you don't like the demogrant (most people don't, though often for confused and incoherent reasons), substitute for it some form of in-kind government spending that happens to have relatively uniform per-person benefits. The analysis remains pretty much the same. And government supply of goods and services can actually be designed to provide greater benefits at the bottom than the top (e.g., public schooling, if spending per enrolled student is equalized across districts and higher-income people opt out more for private school).

Anyway, I'm not a standard liberal (e.g., ask me about Social Security, the minimum wage, or my generally dark view of collective / political decision processes), but if offered the deal I would certainly take it.

No comments: