Tuesday, April 17, 2012

Pikkety-Saez Tax Policy

On this income tax day I wanted to discuss the income tax.  Two economists (Thomas Pikkety and Emmanual Saez) produced a paper at the turn of this century which argued that income distribution in the US had deteriorated from a fairness perspective.  Over the last decade their work has been widely publicized including in support of things like the Buffett rule.  Thomas Pikkety, who now resides in Paris, has even argued that the ideal rate for upper income taxpayers would be between 15 and 40 percent higher than that proposed by the president.

There are two flaws in their research.   First, their definition of income is a bit too narrow.  Indeed there has been some movement among higher and lower income people in the last couple of decades (BTW - that has happened in a lot of countries who have very different tax structures).  Alan Reynolds points out many of the oddities of the data set they use.   While there are legitimate disagreements about the national income accounts and how to use them you should not take Pikkety and Saez's assumptions on their face.

But second, the perspective ignores the substitution effect.  The tax code is gargantuan.   That is in part because every little interest has forced a provision into the code to solve their "problem."   Raising rates never generates as much money as supporters suggest it will.   That is because as you move up the income curve you have a lot more discretion about compensation - that is especially true for things like capital gains.   If rates are too high, I simply will sit on the gains I have accumulated in a stock or other asset.

So how do you solve the problem?  There are two ways.  First, one could index the capitalization of capital - so that before you paid taxes on gains (or losses) you would calculate the effects of inflation on the asset.   That was tried once (in 1954) and discussed at the beginning of the 1986 Tax Act.  But it was dismissed as entirely too complicated.

The alternative is to lower rates below what the average taxpayer pays.  That is the wisdom of the capital gains exclusion.  And the evidence is that if you set it correctly, people actually are more willing to liquidate their investments in line with their wishes rather than tax policy considerations.

If the highest income individuals can decide when to take income (including income on capital gains) then raising rates is unlikely to raise much revenue.    The graph above gives you an idea why the Buffett rule or some variation proposed by Pikkety is pure folly.

No comments: