Here’s another way of looking at the MID from a purely financial perspective. As I discussed at length in the main series of posts about the MID, the two real economic effects of repealing it would be a ~10% estimated drop in housing values on average, and an increase of about $120 billion in total tax liability for current homeowners. I had the following idea, which I would definitely not advocate as a practical plan for unwinding the MID, but I think it demonstrates how poorly it works as an “investment” in home ownership.
According to an investor presentation from Freddie Mac, the total value of U.S. residential housing stock is about $16 trillion, so we could expect a loss of about $1.6 trillion in equity if the MID were repealed. With this in mind, let’s pretend the government were to take the following drastic steps:
- Completely repeal the MID immediately for future mortgages, but allow current mortgage holders to continue to deduct their interest for the full term of their loans.
- Take out $1.6 trillion in 30-year treasury bonds and mail every current homeowner a check for their loss in home equity.
For current homeowners, their homes would decrease in value but they would be compensated dollar-for-dollar by the government for their loss, while their tax situation would remain the same. For new home buyers, it would be as if the deduction never existed, with correspondingly lower prices. I don’t think anyone would have much of a case that they were being particularly harmed by this plan.
So, how would this affect the government’s finances? The current rate on the 30-year treasury bond is about 2.9%. Selling such a large amount of bonds would surely cause the prices to go down (and the yield to go up), so to simplify the math, let’s just call it 3%. Also to make the math a bit easier, I would assume that the extra tax revenue the government would get (compared to the current system) would increase from 0 to the full $120 billion evenly over the next 30 years, as old mortgages come to term and borrowers take out new mortgages which would not be eligible for the deduction. Let’s suppose that the government would use all the new tax revenue to pay interest and principal on the treasury bonds it buys to compensate homeowners.
During the first several years, the government would be losing money from this move because the interest on the one-time charge of $1.6 trillion would overwhelm the new tax revenue trickling in. But by year 15, the critical point would be reached when the balance would start to go down. By the end of the 30-year term of the bonds, the balance would have declined to $1.35 trillion, and the government would have at least an extra $120 billion in annual cash flow. The $120 billion is actually a very conservative estimate, as population growth and even modest housing price increases following inflation would increase the revenue gained by the MID repeal over that time period. Even with the conservative estimate though, the extra revenue would be about 9% of the remaining debt once the treasury bonds came to term. I have no idea what interest rates will be in 30 years, but treasury yields have only been that high during the stagflation days of the 80’s, so I’ll say it’s pretty likely that the revenue would eventually be able to pay off the debt and, in the long run, help the government’s balance sheet.
If you assume a very modest 2% annual increase in the total cost of keeping the MID, then the debt would be down to $261 billion at the end of 30 years, while the annual tax revenues from the repeal would be $217 billion. We’d be a year and change away from paying off the debt completely and then having hundreds of billions of dollars of extra tax revenue.
So, by my back-of-the-envelope calculations, if you really want to subsidize house prices, you’d be better off having the government take on additional debt and just give money directly to homeowners, rather than doing it through the tax code.