Thursday, June 24, 2010

Bounding the Price Impact of the New Home Buyer Credit

The IRS says that $19 billion of tax credits have been paid. Assume, for the moment, that:

  • The credit was well targeted in the sense that (as NAR claims) it reached 1,000,000 persons who would not have purchased a house otherwise,
  • These 1,000,000 persons received a credit of $7,000 (close to the maximum of $8k, and equal to the average among those found to be cheating),
  • All owner-occupied homes are perfect substitutes with each other,
  • Owner-occupied homes are in fixed supply, and
  • The persons receiving the credit previously lived in dwellings that are not substitutes for owner occupied homes.
Of the 1,000,000 who supposedly bought homes because of the credit, some of them would have been induced to buy with a much smaller credit (say, $100) and others needed the entire credit. Note that housing prices had to go up LESS than $7000 because the price increase would have just canceled the credit and nobody would have been induced to buy. $3500 is a good guess for the price increase (given the assumptions above), and would be exactly right if marginal potential new buyers were uniformly distributed in terms of their willingness to pay.

The average home in 2009 sold for $217,000, of which $3500 would be 1.6 percent.

1.6 percent is still less than the housing price increase reported by Case-Shiller over the most recent twelve months. So, even if a 1.6 percent impact were not exaggerated, housing prices would have been stable or increasing over the past year even without the credit.

Moreover, the 1.6 percent impact is exaggerated. First, homes were sold to buyers who did not get the credit, and those buyers did not compete directly with buyers who did (for example, vacation homes or larger homes that are typically sold to buyers who already own a home). Those homes count in the housing price index, and had a price impact of much less than $3500.

Second, owner-occupied homes are not in fixed supply -- indeed much of the construction activity over the past year has been attributed to the credit (this is an exaggeration too). To the degree that the supply of homes was impacted by the credit, housing prices increased less than $3500.

Third, the persons receiving the credit were living somewhere before they purchased their credit-eligible house. Their former dwelling is on market, pushing down housing prices.

One back-of-the-envelope way to consider all of these effects is that each cuts the price impact in half, so that the price impact is $438 ($3500/8) or 0.2 percent. In this case, the main effect of the credit was to cause a few new houses to be built and a couple of million families to switch swap residences with each other.

Regardless of whether the price impact was $400 or $4000, the fact is that housing prices and construction activity stopped falling a year ago, and it is a wild exaggeration to claim that the pattern would have been much different without the new home buyer tax credit.

1 comment:

Richard Gabbert said...

I agree with the most of the assumptions you make of your post (and I think the credit was a bad idea), but I'm not sure the point about the effect of the tax credit on home prices is quite right. Given the highly leveraged nature of home purchases (where the typical buyer can finance 90% of the purchase), I suspect that probably the majority of homebuyers saw the tax credit as affecting the amount they were willing to put down for a home. In other words, in the typical financing situation, an extra $7,000 might make the typical buyer willing to pay an additional $70,000 for the home (or, more likely, purchase a different, more expensive - say $70,000 more expensive - home than they otherwise would have done). Heck, that's what I would have done, had I been in the market for a home, as a cash-strapped but high-income young attorney.

I'm conservative financially, and I know that this is how I think about leveraged debt: Over 30 years, $70,000 (or $35,000) is for many people a manageable increase in one's monthly payment, and the temptation to use cash available in the short term to finance more leverage would be extremely hard for many to resist.

Further evidence of this is the behavior of investors: Give investors $7,000, and many of them will use it to buy shares on margin, rather than just plop the cash into a mutual fund. The effect on share prices, I would guess (though I'm not an economist), is probably a function of that leveraged amount, not just of the amount of the subsidy.

If I'm right, then with your assumptions the price could have increased by up to $35,000 (likely adjusted downward to some extent as not all buyers would have qualified for the loan with the higher monthly payment). Under your approach, this still nets out to a price increase of $4,380, still much lower than the amount of the average credit, but the picture may not be quite as grim as it's painted here.