What the Case-Shiller and OFHEO House Price Indices Are—and Are Not—Telling Us

The Case-Shiller indices paint a more ominous story. The latest numbers, updated through April, implied that prices were falling just about everywhere at accelerated rates. Year-over-year (y/y) prices retreated in all 20 cities covered; adjusted for inflation, prices declined even more, since the CPI was rising 3.9% y/y. In 10 cities, prices were dropping at double-digit rates; in 13 cities, they were falling at record year-on-year rates; in seven cities, the drop was 20% or more. Las Vegas (down 26.8%), followed by Miami (down 26.7%) and Phoenix (down 25.0%), reported the steepest declines. The 10-City Composite Index dropped 16.3% y/y (a record decline), while the 20-City Composite was down 15.3% y/y (also a record).

The government's OFHEO purchase index showed prices falling, but hardly in a freefall. The national monthly index was down 4.6% from a year earlier. The Pacific division (Alaska, California, Hawaii, Oregon, and Washington) had the largest year-on-year price decline (14.9%). In the other eight divisions, however, prices were dropping 5% or less. Moreover, prices in the East South Central division (Alabama, Kentucky, Mississippi, and Tennessee) and in the West South Central division (Arkansas, Louisiana, Oklahoma, and Texas) were up slightly from a year earlier.

Both indices are same-sales indices, which means they track homes that have sold more than once. Why, then, are they telling such different stories about how much house prices are dropping?

The main difference is that they track a different set of transactions. The OFHEO indices only track single-family homes financed by conforming, conventional mortgages that Freddie Mac or Fannie Mae purchase or securitize. A conforming loan is one that meets Fannie Mae’s and Freddie Mac's underwriting standards, which, until recently, included a $417,000 loan ceiling. A conventional loan is one that is not backed by the FHA or other government agencies.

Prices of homes financed with subprime and jumbo loans rose more, and are now falling more, than homes financed with conforming conventional loans. This is why the OFHEO indices are dropping much less than other house price measures.

The Case-Shiller indices track the prices of homes financed by all types of mortgages—which would appear to give it an advantage over other yardsticks. However, these indices have other shortcomings. The 20-city and 10-city indices, for example, are hardly representative of what is happening to prices in the rest of the United States. The same might be said about its national index, which excludes 13 states (Maine, Indiana, Wisconsin, North Dakota, South Dakota, South Carolina, West Virginia, Alabama, Mississippi, Idaho, Montana, Wyoming, and Alaska) and only partially covers another 29 states.

Is One Index a Better Measure of House Prices Than the Other?

Unfortunately, all housing price statistics have two shortcomings that cannot be quantified or overcome. First, they cannot account for the non-price incentives widely used today. These incentives—which, for example, might include the seller paying for a new kitchen or bathroom—effectively lower the selling price of a home. Second, the samples are not random. Homes that are selling are not representative of the U.S. housing stock, since they contain a disproportionate number of homes that are selling at distressed prices by banks, and homeowners trying to avoid a foreclosure. These shortcomings make it impossible to determine how much the price of a typical house in the United States is falling.

Despite these shortcomings, the two indices are saying a lot about housing prices. For example, both agree that adjusted for inflation, prices are falling just about everywhere. Both agree that Florida and California are the epicenters of the housing downturn. Both agree that prices are falling at a faster rate now than six months ago. And both imply that housing prices have further to fall, since both remain far above their 2000 values.

In addition to this, the Case-Shiller tiered indices show that the prices of low-end homes are falling more than those of high-end homes. For example, in San Francisco, low-tiered homes (under $474,000) have plummeted 37% y/y, while high-tiered homes (over $722,000) have fallen 9%. The aggregate index is down 22.1%. A corollary to this is that in most cities (but not all), the further one moves away from the center, the more prices are dropping. Now, with gasoline prices permanently higher, prices in outlying suburbs are falling even more.

The two organizations that publish these indices are also adding to our state of knowledge by periodically issuing research reports on the latest numbers. In June, for example, S&P released a short study by Karl Case that discussed why house prices, which are normally sticky, "have fallen sharply and quickly." Professor Case pointed out that according to a questionnaire he has been conducting since 1980, homeowners are as reluctant as ever to budge on price. The difference between this downturn and previous ones is the number of foreclosed properties—which banks are pricing to sell.

He concludes that, "As long as the foreclosure auctions continue to account for a significant portion of the observed transactions, prices will continue to fall until the market clears. When it does, the traditional stickiness will return and prices will eventually stabilize."