Another month, another terrible job by the business media in reporting housing data released by the U.S. government. As usual, the headline mills out there are carrying misleading information, either because they can't understand the statistics they're reading, or because they don't care to report the numbers truthfully.

For instance, this story began the day with the patently false title of "New Home Construction Up Last Month," and carried a blurb reading, "Construction of new homes and apartments rebounded in October by the largest amount in eight months..."

That's utterly, irresponsibly wrong, and writer Martin Crutsinger and the AP should be ashamed. No journalist who actually understands what he's talking about would ever utter such garbage. (Indeed, judging from the extensively revised article now replacing the original, Crutsinger and the AP seem to have belatedly decided to reverse their conclusion.) Here's why.

October housing completions, as reported by the Census Bureau here, came in at a "seasonally adjusted annual rate" that was 1.9% "above" the revised September estimate. Here's the problem: That monthly number is subject to a relative standard error of +/- 8.9%, meaning it's actually not possible to determine whether there was an increase or a decrease from last month. Moreover, we all know by now that month-to-month comparisons are bogus -- they often illustrate typical, seasonal variations. Year-over-year comparisons really matter, and on that basis, things look terrible.

Turning to the October 2006 comparison, we see just how bad things are for builders like Hovnanian Enterprises (NYSE:HOV), D.R. Horton (NYSE:DHI), Toll Brothers (NYSE:TOL), Ryland Group (NYSE:RYL), Beazer Homes (NYSE:BZH), and Pulte Homes (NYSE:PHM). The October estimate is an incredible 25.2% below the prior-year period, subject to an error of only +/- 6.7%. That means we're as certain as we need to be (more than 90% confident, to the stats geeks) that there was a drop in completions from this time in 2006.

Housing is in terrible shape for builders, and it's likely to get worse. How do we know? Take a look at the rest of that release. Building permits were down 24.5% (+/- 1.4%) from last year's level. Starts were 16.4% (+/- 8%) below the 2006 level.

Over the long term, this is good news, because it indicates that builders are slowing the number of units they're dumping onto an already-glutted housing market. However, it also means that, over the shorter term, many of those builders will have a very hard time paying their bills, because they don't collect cash if they don't complete and sell homes.

And because so much of our economy depended on both homebuilding's halo effect (construction wages, equipment purchases, etc.) and home-flipping consumers, a continued downswing means a continued drag on the economy. That may not be pretty, but it's the truth. Good luck finding it reported correctly by the rest of the press.

For more housing Foolishness, check out these new kids on the block: