New Homes Report Reveals Old Problems with Government Statistics

by: Daryl Montgomery

The U.S. New Home Sales report released on December 23rd indicated a drop of 11.3% in November. Analysts had expected a gain. According to the previous Commerce Department reports, new homes sales had risen every month since April. They were expected to rise again in November because a government tax credit for new home buyers was originally scheduled to end in the beginning of the month (it was extended to June 2010), so analysts had assumed that there would be a rush of last minute buyers. There may have been and without them the drop might have been much greater than 11%.

New Home Sales is almost certainly the most inaccurate of the economic reports issued by the U.S. government. I can say this with some certainty because it would be almost impossible to produce something more error ridden. One of the major news services stated in their coverage of the November report, "Government statisticians have low confidence in the monthly report, which is subject to large revisions and large sampling and other statistical errors. In most months the government isn't sure whether sales rose or fell."

Read that last sentence again and then consider that if the U.S. government is willing to issue an official report on housing that is about as accurate as picking numbers randomly out of a hat, how much can you trust the GDP, CPI, PPI (the two major inflation reports) and Non-Farms Payroll reports. Also note that the mainstream financial media seems to be well aware of the lack of reliability, but doesn't mention it except on very rare occasions when the news is particularly bad.

If the New Homes Sales report is so prone to inaccuracy, why not just fix the problem? This is indeed a good question. The statistical tools to make this report better have been known for decades and yet the U.S. Commerce department doesn't seem to be able to apply them. It can be assumed that this isn't done because they don't want to do it. Statistically, sloppy work is extremely prone to manipulation after all, solidly done work is not.

When confronted with this problem, you will get a more accurate picture of what is taking place by looking at many months of data in aggregate and comparing it to the previous year (the errors will cancel out at least to some extent). In the first 11 months of 2009, new home sales are down 24% from the first 11 months in 2008. Inventories have been falling throughout 2009 and are now at 38 year lows. The number of homes under construction or planned for construction have fallen to a record low.

If new home sales were rising between April to October as the Commerce Department reported, why are home builders building fewer and fewer homes? That doesn't look like an industry in recovery as the public has repeatedly been told. For some reason, we seem to have gotten a glimpse of the true state of the housing market in the November New Home Sales report. Perhaps the guy in charge of producing cheerful statistics was on vacation? Somehow, I'm sure he'll be back soon.

Disclosure: Not relevant.

About this article:

Expand
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here