Both the House and Senate have put forth tax reform plans that appear likely to succeed. At this writing, the Senate plan has passed, with several last-minute amendments made to win over key holdouts on the GOP side. Tax reform will now go to the House, and differences between the two proposals have to be ironed out to produce a final bill that the President would sign, presumably by Christmas.
Again at this writing, both plans appear to preserve the deductibility of property taxes up to $10,000, so this should not affect the vast majority of homeowners. (In a state with the national average property tax rate of around 1.00%, the $10,000 cap would only affect homeowners whose property value is around $1 million or higher, thus this cap would primarily affect high-tax states with high real estate values, such as New Jersey, which has the highest property tax rate in the U.S.) For purposes of this analysis, we will assume this provision remains intact, and that it will have little impact on housing.
Both proposals have implications for the deductibility of mortgage interest, which has been part of the tax code since its inception, in some form. Currently, interest can be deducted on mortgages with balances as high as $1.1 million. The Senate bill shaves that limit to $1 million, while the House version would more than halve it, to $500,000.
Many pundits – including the National Association of Realtors (NAR), which obviously has some bias regarding this topic – argue that limiting the mortgage interest deduction will reduce the homeownership rate and cause property values to decline. This article will analyze the likely impact of the more restrictive House plan on housing activity and property values. My perspective is politically agnostic, focusing solely on the economic implications of the House proposal with respect to housing, and not on the merits of either proposal.
I conducted an informal survey of current and prospective homeowners. While the sample was small, it cut across geographic, political, economic and demographic lines. The survey consisted of one question with four possible answers:
“For those of you who own (or are planning to purchase in the near future) a home, what is the PRIMARY reason?”
- Mortgage interest tax deduction
2. Would rather own than rent
3. My piece of the American Dream
4. Some other reason”
Responses were limited to those four, but comments were allowed. The results were as follows:
- Mortgage interest tax deduction – 3%
2. Would rather own than rent – 84%
3. My piece of the American Dream – 6%
4. Some other reason – 6%
The primary “other reason” listed was the desire to acquire the exact property the homebuyer wanted, presumably through building a home. Also, 13% of the respondents listed #2 as their primary reason, but mentioned that the tax deduction is “nice,” “helpful,” etc.
Clearly, the overwhelming reason these (and, I believe, most) homeowners decide to purchase a home is the ownership vs. rental decision, and that’s one factor that supports the argument that reducing the maximum tax deduction would not impact homeownership. The example below further illustrates the relative economic benefit of owning vs. renting compared to the deductibility of mortgage interest.
Using Zillow, I found a home in suburban Johnson County, Kansas (where I live) with an estimated value of approximately $426,000. Assuming a conforming mortgage with a 20% down payment, the initial mortgage balance on that property would be approximately $341,000.
Further assuming a 30-year fixed-rate mortgage at prevailing rates, the monthly principal and interest payment on this mortgage would be $1,591, and the average interest payment during the first year would be $1,072 (note that the interest portion of the payment declines as the mortgage amortizes, a point I’ll elaborate on further).
Finally, assuming the average household income for the community in which this property is located, the effective tax rate under the current tax code (Federal only) would be 16%. Thus the average monthly tax benefit during the first year of the mortgage would be $173.
The Zillow rent estimate for the same property was $2,512 per month. Thus the rent vs. own differential, using the pre-tax mortgage payment, would be $921 per month. This is more than five times the tax benefit, which explains why the rent vs. buy decision is dominant in the survey I conducted.
The ownership benefit increases over time, as rents typically increase, while the mortgage payment (assuming no refinance) remains fixed for the 30-year term, after which the homeowner occupies the home free of any monthly mortgage or rent payment. However, the tax benefit diminishes over time (assuming no change in the effective tax rate and no refinance), as the interest portion of the mortgage payment declines with amortization, as noted above. In year 2 of the mortgage, the tax benefit declines 2%. By year 7, the annual decline increases to 3%. The decline in the tax benefit accelerates over the life of the mortgage; assuming the borrower takes the full 30 years to pay off the mortgage, the annual decline in the tax benefit has increased to 64% in the last year of the mortgage term, providing just a $5 average monthly tax benefit in that year.
Economic data also supports the argument that reducing (or even eliminating) the mortgage interest deduction would not materially impact homeownership. The first piece of data is the homeownership rate, which has averaged about 65% since it was first recorded in 1965. (If we exclude the period from 1997 through 2011, when the proliferation and securitization of sub-prime mortgages artificially inflated the homeownership rate – with disastrous results – that average falls to about 64%). The lowest homeownership rate during that period was 62.9%, the highest (during the housing bubble) was 69.2%. The highest rate excluding the bubble years was 65.9%, and the current rate is 63.9%, near the ex-bubble mean.
If the tax incentive meaningfully influenced homeownership, we would expect to see higher homeownership rates when marginal tax rates are higher, and vice-versa. However, that isn’t evident in the data. Interest rates have far greater influence than tax rates, and yet the homeownership rate increased from 64.8% to 65.8% between 1977 and 1980, when mortgage rates shot up from about 8.7% to 16.4%.
Clearly, other factors are more influential when it comes to homeownership: consumer confidence, demographics, mortgage securitization, and more. The only thing that meaningfully increased the homeownership rate was the aforementioned proliferation and securitization of sub-prime mortgages, which again had disastrous results. Hopefully that experiment won’t be repeated.
The other piece of data that supports the argument that the tax benefit has little influence on homeownership is total consumer credit outstanding. Prior to the Tax Reform Act (TRA) of 1986, consumer interest was also deductible. The TRA phased out consumer interest deductibility over a three-year period. At the time it was proposed, retailers, automakers and auto lenders argued that it would choke off consumer credit and cause a recession.
That didn’t happen. Total consumer debt outstanding grew 21% over the three-year phase-out of the deduction. It was relatively flat during and following the 1990-91 credit-driven recession, but then accelerated from 1993 through 2008. It declined during and after the last recession, but has since accelerated even further (in large part due to student loans, the interest on which is tax-deductible – at least for now).
As for property values, if the tax incentive were an influence on them, we would expect to see values decline when taxes are cut, as the incentive becomes less valuable. Again, that hasn’t happened historically. Property value fluctuations are primarily a function of supply and demand.
The combination of this economic data and the economics of homeownership vs. renting, along with the declining tax benefit over a mortgage’s life and the increasing own vs. rent benefit, all argue that the NAR has little to be concerned about. So long as the economy remains strong and mortgage liquidity (through securitization of high-quality conforming mortgages) remains sufficient, reducing the cap on the mortgage interest deduction will have no impact on housing activity or home prices.