AP's Rugaber Fails to See Past Seasonally Adjusted Numbers, Misses Troubling 9% June Drop in Single-Family Permits

July 22nd, 2013 9:18 PM

I was going to leave this alone because the original item involved goes back to last week. But Christopher Rugaber at the Associated Press brought it up again in his report today on existing home sales, so it's fair game again.

The final sentence of his dispatch refers to last week's Census Bureau data in the new home market, and claims that "In June, they (builders) applied for permits to build single-family homes at the fastest pace in five years." Not really -- in fact, not at all -- as will be seen after the jump.


Below are the past 6-1/2 years of seasonally adjusted (expressed as an annual rate) and not seasonally adjusted (i.e., actual) results for building permits:

SAandNSAbuildingPermits0107to0613

Actual building permits issued fell by almost 9% in June to 56,900 from May's 62,400. Yet somehow, after seasonal adjustment, June's result annualized came out as a slight improvement.

That doesn't appear to make sense in the context of the past five years which are supposed to be the basis for seasonal calculations. Last year's May-June change was downward, but by less than half as much percentagewise, and it led to a smaller seasonally adjusted gain compared to May. The 2011 May-June change was a positive 4%-plus, and again the seasonally adjusted May-June gain was smaller than this year.

Looking back further, there is no clear trend of actual (i.e., not seasonally adjusted) increases or decreases in May-June time periods going back to 1959. There were 29 increases and 25 decreases before this year. A big actual May-June decrease would thus be expected to generate a seasonally adjusted decrease. This time it didn't.

I'm not saying that the seasonally adjusted calculations are wrong; I'm saying, as I have been for years, that seasonally adjusted figures are not the be-all, end-all in analyzing economic data, especially in a completely abnormal economy like the one we've had for at least five years.

The significant drop in actual permits from May to June makes the following statement from Rugaber's report on housing starts and building permits last Wednesday more than a little shaky:

But applications for permits to build single-family houses rose to the highest level in five years, suggesting the housing recovery will continue.

They only rose on a seasonally adjusted basis because of an inexplicable quirk. Actual permits fell significant, suggesting that the housing recovery, such as it is, might not continue.

Today's report on existing-home sales supports the premise that the housing market is softening. They fell 1.2% from May to June (seasonally adjusted). They had been expected to rise by 1.5%.

Additionally, besides observing that today's miss versus expectations was the biggest in 12 months, Zero Hedge noted that "this is for a period that reflects closings with mortgage rates from the April/May period - before the spike in rates really accelerated." Higher interest rates will certainly not be helpful.

The irony in all of this for those of us old enough to remember is that the current historically weak, barely noticeable "recovery" in housing is occuring even with extremely accomodating mortgage rates, while in the 1980s, housing roared back after that decade's recession ended in November 1982 despite mortgage rates which ranged from just under 10% to over 14% during the entire rest of the decade.

Unlike now, families in the 1980s had a belief that the economy would continue to expand and create opportunity. Few believe that now, which is why the housing industry, after four-plus years of Keynesian policies and failed industry-specific government interventions, is still a mere shadow of what it was (especially after adjusting for population growth) several decades ago.

Cross-posted at BizzyBlog.com.