Turning the Upside-Down Benefit Rightside-Up

President Obama's budget plan includes a dramatic change in the tax code that -- because it would change the way tax deductions are handled -- seems anything but dramatic. Even so, it is a partial version of an extremely good idea that should be extended still further. The idea is to make the tax benefit that one receives from deductions completely independent of the person's tax bracket. Asleep yet? Stay with me. This could be really big.

Two people each give $100 to their favorite charities. Anwar has taxable income of $425,000, putting him in the 35% tax bracket, while Zoe has taxable income of $25,000, putting her in the 15% tax bracket. This means that the $100 deduction for charitable contributions reduces Anwar's taxes by $35, allowing him to give his favored charity $100 at a personal cost of only $65. Zoe only saves $15 and pays $85 to give $100 to her favored charity. (Actually, the chances are pretty good that Zoe does not even itemize her deductions, in which case she saves nothing on taxes and pays the full freight for her charitable donation.)

This unappealing result has long been known in the tax world as the "upside-down benefit" problem, because it gives the largest tax savings to the highest-income taxpayers. It might seem to be impossible to fix this problem, however, because it appears to be the inevitable result of having a progressive rate structure. If you want to have higher tax rates on higher incomes (and I do), then you have to accept the upside-down benefit, right? Actually, it is well known that the answer is no.

The Obama budget takes an approach that is only a partial fix to the upside-down benefit problem, but it is still a huge step forward. The plan would limit the tax benefit for deductions to 28% of the amount deducted. Even though a single person with more than $164,550 in income (in 2008) pays tax at a 33% rate on income up to $357,700 and 35% on amounts above that, the Obama plan would limit the tax benefit to 28% of the deducted amount. The benefit is still upside-down for lower brackets (and for non-itemizers), but it is no longer upside-down for the highest brackets.

This approach might seem to hit higher earners coming and going. They pay a 33% or 35% rate when they earn income, yet they only get a 28% benefit when they offset income. True enough, but why should there be any connection between the rate that one pays on income and the benefit that one receives from deductions? To stay with charitable deductions as our example, why should my tax bracket have anything to do with how much it costs me to give to charity? The charitable deduction is a method by which the rest of society agrees to share in the cost of a decision that an individual taxpayer makes. You pay (1-x)% of something, and we pay x% of it. There is no coherent reason that Anwar's and Zoe's tax brackets should have anything to do with how much of their deductible activities are subsidized by other taxpayers, with taxpayers currently contributing $35 to Anwar's charity but $15 to Zoe's. It is simply an accident of history that we have used the tax code to provide this subsidy, and it is another accident of history that we have tied the amount of the subsidy to the structure of the tax brackets by making this a deduction. The Obama plan partially fixes those historical accidents.

There is no good reason why we should continue to provide these benefits through tax deductions. In fact, we have perfectly good examples of how we can provide tax advantages to certain activities without creating an upside-down benefit. For example, when I lived in Wisconsin in 1998-99, I discovered that the state completely de-couples tax deductions from tax brackets. At that time, the state's income tax brackets ran from 3-7%, but all deductions were multiplied by 5% to determine the tax savings from the favored activities. Instead of deducting dollars from taxable income, taxpayers would simply reduce their taxes by 5% of their total deductions. This turns a tax deduction (which interacts with tax brackets) into a tax credit (which does not). This could be done at the federal level by making deductions worth 28% in tax savings for all taxpayers, not just for the upper-middle and upper classes.

To take it a step further, there is no reason why there has to be a single rate to determine the tax credit, nor do the credit rates have to be tied to the tax brackets at all. We could set up credit brackets that decrease with the amount of deductions. For example, the credit rate could be 25% for the first $20,000 of deductions, 15% for the next $20,000, and 5% for anything more than that. This would be no more complicated than having a table for tax credits that would work much like the table for tax brackets.

Even stopping short of my preference for progressive credit rates, though, Congress should embrace Obama's idea and extend it to all income levels. This would not only increase the overall progressivity of the tax code, but it would do so by separating two things that never should have been linked in the first place. It would, in other words, be a good idea even without the increase in progressivity. This is hardly the sexiest policy idea on the table, but its effects could profoundly affect the way people think about taxes, charity, and all of the other things in our lives that we currently subsidize through deductions.

[As a side note, it is obviously true that any change in the tax code changes people's incentives and thus might change their behavior. Decreasing the tax benefits of charitable donations for higher income earners might cause them to reduce the size of their gifts; but will it? And if so, by how much? The best evidence indicates that the effect on charitable giving will be minimal to non-existent. See various sources linked here.]

-- Posted by Neil H. Buchanan