A Partial Defense of the Mortgage Interest Deduction
August 20, 2013
The mortgage interest deduction is one of the most frequently-criticized elements of the entire tax code. We have written that it distorts investment towards owner-occupied housing. It is also regressive, in that it doesn’t benefit renters or standard deduction filers, both of whom are generally poor. It adds a small layer of complexity to the tax code, which should generally be avoided. It looks for all the world like a social engineering effort.
All of these accusations, unfortunately, are quite true. And yet, I would stop short of saying the mortgage interest deduction absolutely must be eliminated. A scary truth about our tax system is that owner-occupied housing is one of the few types of capital treated somewhat correctly.
The mortgage interest deduction itself actually resembles a fairly sensible idea. A borrower and a lender don’t create any new income in aggregate by agreeing to a mortgage. They only shuffle income around between them. Therefore, if the lender’s income from interest payments is taxed, the borrower’s interest payments should be deductible as, well, “anti-income,” if you will.
This regime usually makes sense, but there are some problems. The balance fails if the lender is not subject to US taxes – or even if he is, there is an arbitrage if the borrower is subject to a higher tax rate than the lender. Also, other sorts of consumer loans (like car loans) do not get this same treatment.
The other peculiar aspect of taxes on housing is that imputed rent – that is, the value of rent that the occupant would be paying if he were renting the house from someone else – is untaxed. People tend to think their imputed rent should be untaxed, and they are right; the value of that imputed rent is already contained in the price of the house, which they purchased with after-tax money. However, most other capital assets – particularly, financial assets or rental properties – do not get this treatment, with a few exceptions like municipal bonds.
Combine that factor with the mortgage interest deduction, and investment in owner-occupied housing has little to no tax burden. In fact, the first post linked above cites a GAO report that estimates the marginal rate of capital taxation as only 2% for owner-occupied housing, as opposed to 18% for noncorporate investment and 32% for corporate investment.
Taxation on capital is poor economic policy. While the treatment of housing in the tax code is far from perfect, the distortion people see towards housing investment actually comes from the poor treatment of other sorts of capital, not from unduly good treatment of housing. Our macroeconomic model sees the issue this way. It finds that eliminating the mortgage interest deduction would be harmful, unless that revenue was used to pay for better treatment of capital elsewhere.
Nobody designing a sound tax system from the ground up would keep the mortgage interest deduction exactly as is – but the best way to help non-housing investment is to reduce its tax burden, not to increase the tax burden on housing.