Wednesday, September 7, 2011

Tax Code Has Upside-Down Rewards for Good Behavior

Tax Code Has Upside-Down Rewards for Good Behavior: Peter Orszag

Tax Code Rewards

Illustration by Matthew Hollister

About Peter R Orszag

Peter R. Orszag, vice chairman of global banking at Citigroup and an adjunct senior fellow at the Council on Foreign Relations, was President Obama's director of the Office of Management and Budget.

More about Peter R Orszag

As the supercommittee created by Congress to cut $1.5 trillion from the federal deficit over the coming decade begins its work, one thing that is said to be on the agenda is tax reform.

Given the complexities involved in a broad-scale overhaul of the tax system, don’t hold your breath that we’ll see anything comprehensive enacted this year.

If the committee wants to bring about constructive tax reform that is a bit less ambitious, however, here’s an idea: Change the tax breaks that are meant to encourage people to do good things -- such as save for retirement, buy real estate, get health insurance or give to charity -- into flat-rate credits that aren’t affected by the taxpayer’s income. After all, it makes no sense, in terms of economic efficiency or simple fairness, to have the size of such incentives depend on earnings.

Consider the current deductions for retirement saving. If a person with a high income contributes $1 to his 401(k) plan, he saves 35 cents in income taxes. The contribution is excluded from his taxable income, so his tax is reduced by the amount of the contribution multiplied by the marginal tax rate. And, this year, 35 percent is the marginal rate on taxable incomes of more than $379,150.

A middle-income person contributing the same amount to a 401(k) would save only 15 cents in income taxes. So the system provides a smaller incentive for the middle-income person to save for retirement.

It would be better if both people got the same tax break per dollar. This is an idea that Gene Sperling, President Barack Obama’s national economic adviser, has previously embraced -- as did the bipartisan deficit commission led by former Senator Pete Domenici and Alice Rivlin, who was President Bill Clinton’s budget director.

Matching Contribution

In a 2006 paper, Jonathan Gruber of the Massachusetts Institute of Technology, William Gale of the Brookings Institution and I laid out a specific way to make it work. We proposed replacing the current deduction for retirement saving with a 30 percent matching contribution from the federal government for every dollar put into a tax- preferred retirement account. Under such a system, if the high-income person saved $1, the government would add 30 cents to the account, and if the middle-income person made the same $1 contribution, the government match would also be 30 cents. (Our proposal had some additional bells and whistles, but that was the gist.)

Here’s the kicker: The available evidence suggests that when high-income households make a contribution to a tax-preferred retirement account like a 401(k), the dollars deposited are very likely to have been saved anyway (in a taxable account). The dollars contributed by lower- and middle-income families, by contrast, are less likely to have been saved absent the tax benefit.

So if our goal is to use the tax code to encourage retirement saving, rather than just push people to shift dollars from one account to another, the current system is precisely upside-down. It offers larger per-dollar tax incentives to big earners, who disproportionately respond by shifting savings they would have kept anyway into a tax- advantaged account. To encourage more saving, per-dollar tax incentives should be larger at the bottom than at the top.

A flat credit rate doesn’t quite go that far, but it moves in the right direction.

What should the credit rate be? Estimates from the Tax Policy Center, a joint venture of Brookings and the Urban Institute, suggest that a 30 percent credit would be about equivalent, in terms of tax revenue, with the current deduction. The differences show up when you look at the changes by income level. Roughly three-quarters of American households face a marginal tax rate of 15 percent or less, and for them, a 30 percent match rate would represent a significantly bigger incentive to save. Households in the 35 percent marginal tax bracket, on the other hand, would face a diminished impetus to save for retirement.

As a result, the Tax Policy Center estimates that about a third of households in the middle fifth of the income distribution would enjoy an average benefit of almost $500. And about 40 percent of those in the top 1 percent would suffer an average loss of about $5,500.

Given the need for revenue, it might be even better to set the credit rate lower, at a level that still holds harmless those in the 15 percent marginal tax bracket. For this, the credit rate would have to be about 18 percent. To see why, consider how much it would cost, after taxes, for a household in the 15 percent marginal tax bracket to accumulate $1.18 in a 401(k). Under the current system, that $1.18 would cost $1 after tax, since 15 percent of $1.18 is 18 cents. Under the new system, the household could also deposit $1, which wouldn’t reduce income taxes but would trigger an 18 cent match from the government, so again the after-tax cost of the $1.18 in the account would be $1.

Larger Savings Incentive

The Tax Policy Center estimates suggest an 18 percent match rate for retirement savings would raise about $400 billion in revenue over the next decade. And the result wouldn’t increase marginal tax rates. Furthermore, it would make the tax break more economically sensible while providing an equal or larger savings incentive for three- quarters of households. We could do worse.

This type of reform could easily be extended to tax incentives for other things. The tax code provides more than $5 trillion a decade in incentives intended to encourage socially valued activities, including not only retirement saving but also homeownership, health insurance and charitable contributions. Most of these are also linked to a household’s marginal tax bracket, so that higher- income taxpayers receive larger per-dollar incentives than lower-income taxpayers.

This approach is justifiable when a tax break is aimed at accurate measurement of income -- as is the case for deducting business expenses. For deductions that promote socially valued activities, though, it makes little sense. It is not only unfair, but also economically inefficient -- absent some evidence that high-income people are more responsive to the tax break or generate larger social benefits when they do respond. In most cases, neither of these conditions passes the laugh test. All such tax breaks should be made into flat credits that don’t depend on income.

Since the change would make our tax system both more efficient and fairer, Republicans and Democrats alike should support it. Sorry to say, though, in this era of hyper-polarized politics, I wouldn’t count on it.

No comments:

BLOG ARCHIVE