Here's a scary thought about the housing market: Things may be far worse than what's already being revealed by the troubling government and industry statistics.

At issue is what goes into sales price data and what does not. When those numbers are crunched, many of the incentives that sellers are using to lure buyers – including cash rebates – aren't being included. That suggests prices may be falling faster in many markets than is now being reported.

The same goes for how the mortgage-application indexes don't account for the implosion of lenders. That could have the effect of masking a slowdown in demand, which is why the housing market could be in for rough sailing much longer than most anyone anticipates.

Not long ago, the housing market was leading the economy's growth. But the business has slumped over the last two years in many parts of the country, with sales of new and existing homes plunging. The mortgage business also is under intense stress as borrowers with weak credit increasingly default on their home loans.

Since what happens in housing has a far-reaching effect – it has rattled financial markets, caused dramatic tightening of lending standards and shaken consumer confidence – every bit of data is scrutinized for hints of whether a recovery is near or more trouble lies ahead.

There certainly has been plenty of bad news, but it might not even be giving a full picture of how difficult things really are.

For instance, the Commerce Department reported last week that the median sales price of new homes fell 0.9 percent in May from a year ago, after tumbling 10.9 percent in April.

But those numbers don't include the thousands of dollars in lavish incentives like plasma televisions, pool installation and closing costs that sellers are increasingly using to woo buyers. That means a home selling for $600,000 gets reported for that price even though all those extras technically are reducing the net sale price.

Sales incentives at Lennar Corp., one of the nation's biggest builders, averaged $43,700 a home in its fiscal second quarter, up from $24,700 in the same quarter last year. And it isn't just builders piling on the incentives – it's spilling over to the existing-home and foreclosure market, too.

“In effect, they are reducing the new sales price but that is not showing up anywhere in the actual sales data,” said Peter Schiff, who runs the investment firm Euro Pacific Capital Inc. in Darien, Conn.

Another report with some distortions is the Mortgage Bankers Association's weekly survey of mortgage applications, which has long been considered a reliable indicator for new and existing home sales, and a gauge of housing activity in general.

Over the last decade, the year-on-year change in the MBA's “purchase applications index” has shown a 76 percent correlation with the year-on-year growth rate of the combined new and existing single-family home sales, lagged by one month, according to Goldman Sachs.

That might not be holding true now, since the index fails to account for some of the turmoil in the mortgage finance business. The purchase applications index is now showing a 9.5 percent rise for June from year-ago levels.

For one, the MBA survey refers to mortgage applications, not originations, which legally bind borrowers to the mortgage. That is not a problem when lending standards are stable, but in times when lending standards are tightening – as they have been in recent months – more applications are rejected, and therefore, originations go down, according to the new Goldman report.

Another problem is the survey represents about half of all mortgage applications, but Goldman says that subprime is slightly underrepresented. Lenders who have gone out of business – a regular occurrence in today's markets, with dozens of mortgage lenders having imploded in recent months – might not be included in the sample. That could overstate its results.

The MBA acknowledges that its survey includes only lenders offering mortgages through retail branches, and that most lenders that have gone out of business aren't in its sample because they were primarily subprime lenders relying on mortgage brokers. They also weren't included when applications were soaring a few years ago, notes Jay Brinkmann, vice president of research and economics at the Washington-based MBA.

He also points out the six-week to two-month lag time between a mortgage application filing and a home closing means the jump in May's applications might not be reflected until the June home sales data are reported in July. “We have to wait until then to see if there is really a decoupling,” Brinkmann said.

How that plays out could tell much about the state of housing today, and whether the problems in that market are a lot more troubling than appears on the surface.