The headline and opening sentence in Derek Kravitz's Associated Press report this morning on the Census Bureau's homebulding industry data release gives readers the impression that industry activity increased impressively during September. It increased a tiny bit, but certainly not by the percentage indicated.
The headline ignorantly assumes that a double-digit increase in housing starts is the same as an increase in "home building." It isn't. That headline, the first four paragraphs from Kravitz's report, and some other indicators of housing market progress -- and the stunning lack thereof, three full years after the politicians promised that the Troubled Asset Relief Program would right the ship -- follow the jump (bolds are mine):
Home building jumps 15 percent in September
Homes were built in September at the fastest pace in 17 months, a hopeful sign for the economy.
Most of the gain was driven by a surge in volatile apartment construction. That should help create jobs and boost economic growth, but it doesn't signal a comeback for the depressed housing market.
Single-family home construction, which represents nearly 70 percent of homes built, rose only slightly. And building permits, a gauge of future construction, fell to a five-month low.
Overall, builders began work in September on a seasonally adjusted 658,000 homes, the Commerce Department said Wednesday. While that's a 15 percent increase from August, it's roughly half the 1.2 million that economists say is consistent with healthy housing markets.
Do I really need to explain that starting work on a home isn't the same thing as building (i.e., completing) one? I guess I do.
Kravitz's report obviously got better after the first paragraph, but of course the headline and the first paragraph will be the items on which news snips and broadcast news reports will focus. The headline also conveys an implied assumption that the (mythical) activity increase is in single-family homes, when it isn't. Excuse me for contending that the placement and phrasing are hardly accidents.
Here's a rundown of activity noted in the Census Bureau's release and the underlying data:
- Starts, seasonally adjusted (SA) -- As AP noted, up 15%, with a heavy skew towards apartments. The SA single-family component of an annualized 425,000 is less than three other months this year, 11 out of 12 months in 2010, and, with the exception of December 2008 through May 2009, every other month on records going back to 1959.
- Starts, not seasonally adjusted (NSA) -- The overall raw number of 60,100 is the best September since 2008. The single-family number of 37,700 is the lowest of any September recorded.
- Building permits (SA) -- Down 5% from September, as AP noted, to an annualized 594,000. That's lower than each of the four previous months.
- Building permits (NSA) -- The actual figure of 52,200 is the lowest of any September on records going back to 1959 except 2010.
- Completions (SA) -- The annualized figure of 634,000 is 2.1% better than both last month and September 2010, and is the best result since June 2010.
- Completions (NSA) -- The actual figure of 57,900 is 2% higher than September 2010, but the lowest of any September on records going back to 1968.
- Housing units under construction (SA) -- 412,000, the second-lowest on records going back to 1970. Last month, at 409,000, was the lowest ever.
- Housing units under construction (NSA) -- The actual number of 425,400 is the lowest of any September recorded.
The above data clearly refute the notion that "home building" increased 15%, and make mincemeat of Derek Kravitz's contention that "Homes were built in September at the fastest pace in 17 months." Kravitz's "hopeful sign for the economy" isn't really there.
Geez guys, why don't you just tell us what happened: "Housing starts up 15% in September, permits down"? Sadly at what should be renamed "The Administration's Press," the question answers itself.
Cross-posted at BizzyBlog.com.