May 24, 2007

Bubble, Schmubble: Median Home Sale Price Decline Almost Entirely Due to Regional Mix

That’s right. Bubble, shmubble, despite this picture from Matt Drudge, who got snookered on this one:


Fire sales, schmire sales.

The Chief Snookerer in the latest search for the elusive housing bubble is Martin Crutsinger of the Associated Press, with a significant assist from the Commerce Department (link is to a PDF), which inexplicably did not, and apparently does not, report the regional sales data needed for a more detailed look.

Crutsinger took Commerce’s housing report showing a significant decline in the nationwide median selling price of a new home, both in the past month and year over year, and ran with it at an all-out sprint (bold is mine; BizzyBlog copy for future reference):

….. the median price of a new home sold last month fell to $229,100, a record 11.1 percent decline from the previous month. The big price decline indicated that builders are slashing prices in an effort to move a huge overhang of unsold homes.

The drop in median prices in April compared to March was a record one-month decline. If the April sales price was compared to the sales price a year ago, the decline was 10.9 percent, the biggest year-over-year drop since 1970.

So should we all start looking for cardboard boxes and bridges as our home-value declines swallow up our equity?

Uh, no.

Most obviously, the Commerce report itself shows that the inventory of unsold homes dropped from 8.1 months in March to 6.5 in April — which makes sense, given April’s huge increase in home sales. The fact is, the April “overhang” hung way lower than it did in March.

Less obviously, because Commerce didn’t include regional sales prices, about 90% of the “big price decline” can be explained by big changes in the mix of where homes were sold in April, compared to March and compared to April 2006. In both cases, sales as a percentage of total sales have declined significantly in the West, which is by far the most expensive region in the country. At the same time, sales in the South, the second-least expensive region which comprises more than half of all sales, have increased their percentage share.

The following pictures illustrate what I’m referring to. I used Commerce Data in the report just issued for unit sales, and National Association of Realtor data for regional median sales prices at a spreadsheet that is available at the second link at this NAR page. The sale-price data from NAR is from March 2007 and April 2006, as their April report has not yet been released:


Even without doing the math, you can clearly see that if you mix in a higher proportion of sales from a cheaper part of the country with fewer sales from the most expensive one, the medians will naturally come down, even when no one region is suffering a steep decline.

I estimate that after changes in mix are taken into account, the overall change in home-sales prices from March to April was about -1.2% (vs. the overall -11.1% reported), and about -1.5% (vs. the overall -10.9% reported) from April 2006. The details supporting those estimates are at a separate post here.

Crutsinger also clearly erred by giving readers the impression in the third and fourth paragraphs of this excerpt that the changes described were month-over-month instead of year-over-year (bold is mine):

The jump in sales was the biggest increase since a 16.4 percent surge in new home sales that occurred in April 1993.

However, analysts cautioned against reading too much into the big gain, especially in light of other surveys showing that builder confidence has sunk in recent months over worries that troubles in the subprime mortgage market will further crimp demand in coming months.

There was also concern because all of the strength in sales came in one region of the country, the Northeast, which saw a surge of 43.1 percent.

Sales were down 28.1 percent in the Midwest and 25.4 percent in the West. Sales fell a smaller 3.4 percent in the South.

The drop in median prices in April compared to March was a record one-month decline. If the April sales price was compared to the sales price a year ago, the decline was 10.9 percent, the biggest year-over-year drop since 1970.

The 16.4% in the first paragraph was the one-month increase from March to April, but the percentage changes in the third and fourth paragraphs are from April 2006 to April 2007. Crutsinger never told his readers he was switching gears (and switching yet again in the fifth excerpted paragraph). The fact is that in April, instead of having a “fire sale,” the South’s housing market was “on fire,” with sales increasing a whopping 28% over March. Far from showing a big “strength in sales,” the Northeast was up only about 4% in April over March, while the West and Midwest were up 8.5% and down about 4%, respectively.

All of the above renders the inevitable straw-grasp by Crutsinger look really lame as he, like so many other business reporters during at least the last four years, rolls out the R-word yet another time (bold is mine):

Analysts are hoping that spending by consumers and businesses will be able to overcome the weakness in housing and keep the country out of a recession.

Sorry, Martin, if you’re looking for a recession, it would appear that you’re going to have to go somewhere besides the housing market to make a your case.

Cross-posted at


UPDATE: Zheesh, I soooooo despise the way Yahoo handles AP’s coverage. It continually updates existing links instead of preserving older versions (how convenient). Crutsinger’s later report as of 6:34 PM that is now at Yahoo has the correct info for month-over-month sales volume (perhaps our favorite Martin is a NewsBusters/BizzyBlog reader/fan?). The original link above has been replaced by one at that (for now, even though it says 6:37 PM) has Crutsinger’s original reporting and errors. The link replacement was also done at NewsBusters, but without further comment.


No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.