Evan Soltas
Aug 15, 2013

An Interesting Case

Home Mortgages
I don't like the tax deduction on mortgage interest. I don't know anyone that does -- besides those people whose salary depends on their not understanding the arguments against it, to paraphrase Upton Sinclair. And I think it's particularly egregious given the extant, and probably unfixable, subsidy for owner-occupied housing -- that is, nobody taxes imputed rent -- and since deadweight losses scale quadratically with the per-unit subsidy, the combined effect is particularly ugly.

But someone asked me a while ago if there could be a good argument for a mortgage-interest subsidy. This is a different question from weighing the arguments, or even if there is a good argument -- it's whether, given some reasonable-but-not-necessarily-true assumptions, there exists a reasonable-but-not-necessarily-winning case for such a deduction, in some form.

In case you're wondering: No, this person is not from the National Association of Realtors. He or she does competitive debate and was wondering if there existed any possible counter-argument (as if none exists, then the team that proposes that the debate be over the mortgage-interest deduction is at fault and loses).

I told him or her I'd think of one. I've been turning it over in my head for a month or so now. Here goes. I think you can make an interesting argument for some form of mortgage-interest deduction on two grounds: behavioral economics and imperfect financial markets.

The behavioral argument is that people hate to save -- and they really, really hate saving when it's super abstract, like sticking dollars in a retirement account, buying a mix of stock and bond index funds, and waiting 40 years. They hate it so much that they just don't do it.

In fact, when you take away Social Security and defined-benefit pensions -- Social Security because it's forced saving and defined-benefit pensions because they're of another era when employers covered retirements -- you'll notice only two sources of retirement savings that are of any significance. They are home equity and private investment accounts. And then you'll notice that the private investment accounts have $42,000 on average in them.

When you consider the lifetime income of the average worker, what that number means is that most Americans don't save anywhere near enough on their own for retirement. But where Americans do have some money stashed away is their home -- twice the amount in the investment accounts.

Now why is that? Or, more pointedly, would that saving shift into private investment accounts, on the margin, if the mortgage-interest subsidy went away? I think there's a reasonable claim that it wouldn't, and there's some pretty solid ground for a mortgage-interest subsidy.

The house has a unique power to incentivize household savings. Whether it's because there's a massive social pressure to homeownership, or physical proof that you're a homeowner, or a day-to-day reminder of the importance of saving doesn't really matter. What matters for this argument to matter is that saving is behavioral.

Argument Two is that financial markets are imperfect in such a way that relatively-low-savings savers are locked out, or at least face fixed costs and frictions that impede investment. This feeds off Argument One, because if the only way to get people to build savings is to put it in a home, and if the only way to get savings to be productively invested is in a home, then you're kind of stuck with encouraging people to save through homeownership.

And I think the underlying claim of fixed costs to investment is somewhat true, albeit weaker than the first claim. But the point is that to make any case to defend the mortgage-interest deduction, you really have to show that there is no other choice. Notice how that shuts out some arguments that the subsidy is inefficient on revealed-preference grounds, or that it's a waste because it just inflates the value of homes.

The reason you have to shut out these points is that telling people to hold basically two-thirds of their private savings in a single asset class makes absolutely no sense. They should diversify, and a disincentive against diversification is simply dumb policy at work. So to make an nth-best policy optimal, you have to say that all of the n-1 policies before it really don't work in real life.

An effective counterpoint to the two stretches of arguments I've put out here is that I've misrepresented who benefits mortgage-interest deduction. It's people who already have the money to buy big homes -- see the CBO's May report or the Tax Policy Center's distributional analyses for more on that. The solution, of course, is just to advocate for some more progressive version of the mortgage-interest deduction that functions as a no-alternatives, de-facto subsidy for saving.

* Oh, and I'll add in one more argument after the fact. There could be an external social benefit associated with homeownership. You're aligning the incentives of residents with the interests of future residents, who want the community maintained nicely.

I don't believe the arguments. But that's not point of the question. It's "can you find a foothold." Not "can you find an argument." I believe in this type of exercise -- given confirmation bias, you'll never know how strong (or weak) a position is until you force yourself to make the best arguments against it.

Photo credit: Jeremy Brooks, Flickr.