Do we want to use the tax code to subsidize home ownership? And, if we do, is the mortgage interest deduction the best way to do it? A new paper by my Tax Policy Center colleagues Eric Toder and Katherine Lim, along with Urban Institute researchers Margery Turner and Liza Getsinger, asks these provocative questions, and comes up with some surprising answers.

To even ask seems almost un-American—almost like suggesting we replace barbeque at the Memorial Day picnic with, oh, tofu. But a close look suggests there is much less to the hallowed deduction than meets the eye. Thus, we’d miss it much less than we think.

In 2012, the deduction will reduce federal revenues by $131 billion. In contrast, the entire budget for the Department of Housing and Urban Development is just $48 billion. The conventional wisdom says these tax breaks are important because A) they increase home ownership and B) homeowners are more engaged in their communities than renters.

It turns out that neither of these assumptions is necessarily true. For instance, for a half century--until the recent real estate boom and bust--home ownership rates in the U.S. have barely budged even though the value of the deduction has fluctuated widely. Similarly, there is no clear connection between home ownership and the availability of mortgage deductions in other countries.

The exact relationship between home ownership and other social benefits is just as uncertain. We know home owners are more connected to their communities than renters. But is that because they own a house, or is it merely that the same types of people who are engaged in their communities are also prone to home ownership? We don’t really know.

We do know, however, that the deduction is not a very efficient way to encourage home ownership. Most benefits go to high-income households that would probably buy a house with or without the deduction. Since non-itemizers get no benefit from the deduction, it is not surprising that most of the subsidy goes to upper-bracket taxpayers.

So is there a better way? The paper looks at a half-dozen alternatives, from eliminating the mortgage subsidy entirely to capping its value or turning it into a credit. Not surprisingly, each design has its own set of winners and losers.

To take one example: If the goal is to encourage homeownership among people who otherwise would not buy, what if we replaced the deduction with a credit? Remember that credits are usually a better deal for middle-income households. Simple example: A 20 percent credit on $1,000 of interest is worth $200 no matter what your tax bracket or whether you itemize. But a $1,000 deduction is worth $350 to someone in the 35 percent bracket but only $100 to an itemizer in the 10 percent bracket, and nothing to someone who takes the standard deduction ).

The paper looks at four different credits, each of which provides the same total subsidy amount as the current deduction. For instance, replacing the deduction with a non-refundable credit equal to 20 percent of interest payments would raise average after-tax incomes for households in the lowest 80 percent of the income distribution, with middle-income households getting the biggest average benefit. However, on average those in the top 20 percent would do less well with the credit than with today’s deduction. A non-refundable 100% credit capped at about $2,000 would benefit middle-income households even more but raise taxes more for people in the top 20 percent.   
 
Sadly, the study also explains why the deduction is likely to stay right where it is. The big winners under the current system are upper-middle-class suburbanites who disproportionately own homes, itemize deductions, and spend a relatively large share of their income on mortgage interest. And nobody wants to get them mad, either by cutting their housing subsidies or feeding them tofu.