Posted 02.01.2010
The Economist
Those confusing home-price statistics
By Tucker Hart AdamsThere are a lot of confusing, indeed conflicting, economic data released by various organizations and reported with breathless headlines in the popular press. One of the most confusing, even to economists, is the monthly data on housing prices.
Pick a month - say September, when we read that prices fell 7.1 percent, rose 0.2 percent, fell 3.8 percent, rose 1.8 percent and fell 8.9 percent. Did home prices rise or fall? Can both be true?
In a sense, the answer is yes. The problem is various methodologies covering different components of the housing market.
The simplest figure to understand and the one that gets the most attention is the monthly news release by the National Association of Realtors covering existing home sales and median prices the previous month. The November release showed that the median price declined 7.1 percent from October 2008 to October 2009.
There are two problems with Board of Realtors data. First, they only cover existing homes that sold, and second, the overall price can actually increase even though every home price in the data base fell.
How can that be? Consider a community comprised of five homes and look at the median price. Suppose in Period 1, five homes sell and in Period 2, three of those home are resold.
Period 1 Period 2 Price Change
House 1 $150,000 $140,000 -6.7%
House 2 $200,000 No sale n/a
House 3 $300,000 No sale n/a
House 4 $350,000 $310,000 -11.4%
House 5 $600,000 $525,000 -12.5%
Median Price $300,000 $310,000 3.3%
Average Price $320,000 $325.000 1.6%
Each house sells for less in the second period than it did in the first, yet the median and average price increase. Although this is a trivial example, it illustrates what happens when the market softens for lower priced homes, but more expensive homes continue to sell, albeit at a loss. The same thing can happen in the opposite direction if expensive homes sell in the first period and moderately priced homes in the second.
There are two other price reports using different methodologies that give a more accurate picture of what may have happened to the price of your home. Both utilize what is called the repeat sales method, using only data on properties that have sold at least twice in order to capture the true appreciation.
The Office of Federal Housing Enterprise Oversight uses a database of homes insured by Fannie Mae or Freddie Mac. This excludes the sale of any home priced above $720,750 ($417,000 prior to 2009). Data are provided quarterly and showed a 3.8 percent decline between the third quarter of 2008 and the third quarter of 2009, but a 0.2 percent increase from the previous quarter and no change relative to August. That in itself is confusing enough for the average person trying to understand what is happening to residential real estate values.
The S&P/Case-Shiller Home Price Index, named for the two economists who developed the methodology it uses, provides the broadest coverage of the three, since it includes the larger homes not picked up the OFHEO data base. Home price data are gathered from sales reported by local recording offices across the U.S. A search is then conducted to find information regarding any previous sale of the same home. If an earlier transaction is found, the two transactions are paired and considered a repeat sale.
Data are compiled in this way for 20 Metropolitan Statistical Areas, including Denver, and are also aggregated to give a national index. A rolling three-month moving average (i.e., the September figure was an average of July, August and September) is used to offset delays in data reporting and to keep the sample large enough to create meaningful price-change averages. This is the broadest coverage of the three and showed a 1.8 percent increase in prices in the third quarter 2009 relative to the second quarter, but an annual decline of 8.9 percent.
Each series has its shortcomings. Only the NAR gives actual home prices; the other two provide indexes that enable you to look at percentage changes. All exclude new homes and homes that don't sell.
As we've said before, "Figures don't lie, but liars can figure." The many different ways housing prices are reported, with significantly different results, provide a lot of leeway to use the data to prove a point.
Tucker Hart Adams, president of the Adams Group, monitored and analyzed the Colorado economy for 30 years. She can be reached via her website, coloradoeconomy.com.
Readers Respond
Leave a comment
Commenting is not available in this section entry.