When Christopher J. Artur was president of the Realtors Association of Greater Philadelphia during the housing-market downturn of the mid-1990s, many of his colleagues urged him not to publicize sales and price data.
"I argued successfully that, if we didn't give out that information in bad times, how would anyone believe us in good times?" Artur said.
Today, even though such data are available from numerous sources, the truth about the state of the housing market is an elusive quarry.
For example, two major collectors of housing statistics - the National Association of Realtors and Standard & Poor's - acknowledged recently that the data they have gathered might be overstating price declines.
Lawrence Yun, chief economist for the Realtors group, said the first-quarter median-price declines it reported were linked to a drop in jumbo loans and a resulting increase in the sales of lower-priced homes.
David Blitzer of Standard & Poor's, which produces the three Case-Shiller Indexes, told the financial Web site Marketwatch. com that its monthly index painted "an incomplete picture" of home values.
The 20-city monthly index does not include Philadelphia. Many of the markets it does include - San Diego, Las Vegas, Miami, Phoenix, Los Angeles and San Francisco - experienced huge price increases, followed by spectacular declines.
The data issues come as no surprise to many observers of the real estate market.
"This is not news to most housing economists who work with these price indices," said Kevin Gillen, a research fellow at Wharton and vice president of Econsult, a Philadelphia economic-forecasting service. "What I find ironic is that the NAR does not point out that if these indices overestimate the downside, then they are also prone to overestimating the upside.
"During the boom, I heard no such criticism from the NAR that overestimation of house-price appreciation was leading people to be excessively bullish about the housing market," Gillen said. "The stats can also make things look rosier than they are."
Commerce Bancorp Inc. chief economist Joel Naroff asked: "Do we really know that there has been a greater impact from the jumbo [loan] problems than from the complete withdrawal from the subprime market, which has tended to be for lower-priced homes?
"There is always an issue when using medians or means when the structure of the data, in this case the housing market, changes," Naroff said. "But it is also important to understand what bias the change creates, and I am not as certain it fosters a greater price decline right now."
A median price is the middle value: Half the houses sold for more, half for less.
Mark Zandi, chief economist for Moody's Economy. com, said that using median prices was "OK, but can be very volatile due to a changing mix of sales, particularly quarter-to-quarter."
Zandi puts more faith in the quarterly Case-Shiller national index, which tracks all sales in 100-plus metro markets around the country, including Philadelphia, than in the quarterly report produced by the Office of Federal Housing Enterprise Oversight.
"The key reason is that OFHEO covers homes with conforming mortgage loans, and there aren't many in places like California," Zandi said.
Case-Shiller's national index is more heavily weighted to coastal markets, he said, while the federal agency is weighted toward interior markets, and that would make price declines greater in Case-Shiller.
(Conforming loans have established limits, such as $417,000 for a one-family home, thus higher-priced markets such as California tend to have fewer of them. In the market recently, there have been more foreclosures of nonconforming loans nationally.)
The state of the new-home market isn't much clearer. The Census Bureau's monthly national sales numbers "are hardly ever statistically significant," said Patrick Newport, housing economist at Global Insight Inc., of Lexington, Mass. "Because of the small sample, the monthly numbers are volatile."
In addition, Newport said, rising cancellations for new-home contracts have increased the unreliability of the estimates. So, to gauge trends, he uses a four-month "moving average" showing average value over the period.
Given all the caveats, why follow the data at all?
Center City Realtor and mortgage broker Fred Glick said he could see no value to the Realtors' reporting national sales and prices, which he considered virtually meaningless to individual markets.
"All housing is local, and it just does not matter to people in Philly what the market is in Waco, Texas," Glick said.
"They're just numbers," Artur said. "They don't mean anything to me. What is going on in Center City doesn't mean anything to me. I just care about the Northeast."
Wayne Norris, regional director of Hanley Wood Market Intelligence, which tracks new-home sales, said it was important for homeowners in the Philadelphia region to understand that just because a national or regional median figure is down 5 percent, "it does not mean their home has lost the same value."
"Many home sellers will get less for their home today than in 2005, but that reduction will be substantially less than the numbers they see and hear from the NAR and S&P's," Norris said.
Peter Morici, a professor of business at the University of Maryland, said the Realtors' numbers do have some value - for comparison purposes.
"Look at the NAR data and then go look in your neighborhood," Morici said. While the data from Case-Shiller "are likely over the top," the Realtors have "an interest in saying prices are not falling."
Zandi said Case-Shiller and median prices were much more useful than the federal oversight agency's index in gauging the market.
Still, he noted, "no measure is perfect."