In an earlier piece, I argued that the $8,000 first-time homebuyer tax credit was a poorly targeted subsidy that should be allowed to expire, as planned, at the end of November. Unfortunately, the President and Democratic Congressional leaders are moving toward extending the credit. Senator Dodd has suggested making the credit available to all home buyers (not just first-time buyers), subject to income requirements. Senator Dodd said he is working with Senator Isakson, who previously proposed a $15,000 tax credit to any buyer of a home.
Let’s do a back-of-the-envelope calculation of the tax expenditure and the expected increase in housing sales from a one-year, $15,000 tax credit for homebuyers.
Assume that we can expect 5.5 million home sales in the following year (based on the latest estimate of 5.1 million annual sales of existing homes and 0.4 million annual sales of new homes. Assume also that the median sale price is $200,000. (This source says average – not median – house price in Q1 was about $206,000.) Finally, we need to know the price elasticity for home purchases. There are many estimates in the literature: Todd Sinai has it around -0.5, whereas other studies – such as those reviewed by Harvey Rosen – have it around -1.0). For my calculation, I assume a price elasticity of -0.65.
The credit amounts to a 7.5 percent (= $15,000 / $200,000 * 100) reduction in sale price. With an elasticity of -0.65, this price reduction will lead to a 4.875 percent (= 7.5 * 0.65) increase in housing sales. With a baseline of 5.5 million housing sales, this means about an additional 268,000 (= 5.5 million * 0.04875) sales.
The subsidy is poorly targeted because it would give a credit to 5.5 million homebuyers who would have bought a home anyway. So we are getting about 268,000 additional sales at an estimated tax expenditure of about $86.5 billion (= $15,000 * 5,768,000), implying a cost-per-additional-sale of about $323,000!
Granted, this computation is based on some uncertain parameter assumptions. Perhaps the most uncertain is the price elasticity of housing. So let’s do a robustness check. According to the National Association of Home Builders (NAHB), extending the $8,000 credit for a year and making all buyers eligible will lead to 383,000 additional home sales. NAHB’s news release does not describe how it arrived at this calculation, but if it is assuming the same baseline housing sales and median housing sale price as I am, this implies a price elasticity of housing of approximately -1.7 (rather than the -0.65 I use).
Even using NAHB’s more optimistic estimate, the $15,000 tax credit will result in about 700,000 (= $15,000 / $200,000 * 1.7 * 5.5 million) additional sales at an estimated tax expenditure of about $93 billion (= $15,000 * 6.2 million), implying a cost-per-additional-sale of about $133,000. Even under NAHB’s optimistic assumptions, this still amounts to a very expensive and poorly targeted subsidy.
The estimates above assume that every homebuyer will be eligible for the $15,000 tax credit. While this is what Senator Isakson proposes, Senator Dodd says that high earners will not be eligible for the tax credit. I don’t know of any data on the income distribution of homebuyers, but according to the Joint Committee on Taxation (Table 6), about 11 percent of recipients of the mortgage interest deduction in 2007 made $200,000 or more in adjusted gross income.
So let’s assume that 15 percent of homebuyers will not receive the new tax credit, so there is no change in baseline sales for these people. Of the remaining 85 percent, let’s assume the median house sale price is $180,000 (compared to the $200,000 median price including the high earning buyers). Assuming a price elasticity of -0.65, this implies an increase of about 253,000 (= $15,000 / $180,000 * 0.65 * 0.85 * 5.5 million) additional sales at an estimated tax expenditure of $73.9 billion (= $15,000 * 4.928 million), implying a cost-per-additional-sale of about $292,000. Using NAHB’s implied price elasticity of -1.7, we get about 662,000 additional sales at an estimated tax expenditure of $80.0 billion, implying a cost-per-additional-sale of about $121,000.
If the goal of extending and expanding the homebuyer tax credit is to spur housing sales, an optimistic assessment is that we will spend $121,000 in tax revenue per additional housing sale. Again, this is a very poorly targeted subsidy.
Putting aside the high cost, the goal of increasing housing sales is misguided. The core problem with our housing market is that the bubble led to too many homes being built, and the recession led to a decline in household formation. Even if the tax credit spurs house sales, many of these transactions will just shift renters into buyers, which does not address our excess inventory of houses. Indeed, one unintended consequence of the first-time homebuyer tax credit could be the worsening of conditions in the rental market.
Finally, there are some larger points we must not lose sight of. First, tax expenditures are not a free lunch. Everything spent on this program will ultimately have to be paid for later through higher, economically harmful, taxes. Second, we should heed Ken Rogoff’s warning that “the rate at which government debt is piling up could easily lead to a second wave of financial crises within a few years.” Finally, we ought to learn from our mistakes of the past: government policies to encourage people to become homeowners (or, perhaps more accurately, home-borrowers) partially contributed to our housing and credit market problems. Ultimately, we need to decrease the government’s housing incentives, including the mortgage finance subsidies, the mortgage interest deduction, and the favorable capital gains treatment for housing. A good place to start cutting the cord would be by not extending, let alone expanding, the homebuyer tax credit.