U.S. housing starts rose to a 2.088 million annual rate last month( strongest year since 1978)

Bloomberg ^ | 01/21/04 | Bloomberg

[Pikamax]

U.S. December Housing Starts Rise to 2.088 Million (Update1) Jan. 21 (Bloomberg) -- U.S. housing starts unexpectedly rose to a 2.088 million annual rate last month, capping the strongest year for the industry since 1978, a government report showed.

The December pace was the fastest since February 1984 and brought the total number of homes started in 2003 to 1.848 million, the Commerce Department said in Washington. In November, starts increased to a 2.054 million annual rate.

[Starwind]

Not surprisingly, the undajusted numbers tell the opposite story.

Unadjusted Starts are lower (though permits are slightly higher than the drop in November, both unadjusted Starts & Permits are below October and earlier in the year. This is also consistent with the construction layoffs.

See http://www.census.gov/pub/indicator/www/newresconst.pdf

New mortgage apps spiked last weeked, so some growth new home construction may be in the pipeline, depending on interest rates & constructions costs (inflating).

[dangus]

No, the unadjusted numbers simply reflect the fact that it is December, a terrible month to start building homes. Actually, the fact that December was colder than normal across most of the US (with predictions of cold throughout the winter), probably means that housing construction growth could have been still higher. (My guess is the cold is why experts had expected a decline.)

[Starwind]

No, the unadjusted numbers simply reflect the fact that it is December, a terrible month to start building homes.

What they reflect, actually, is reality. The actual number of homes started/permitted, due to whatever conditions you care to imagine. And that number has been declining since July, whereas the adjusted numbers (the altered by historical assumption numbers) were rising.

That's why they're called 'unadjusted' - they haven't been altered by anyones' assumptions.

If people actually bought adjusted homes and construction workers had jobs to build adjusted homes and adjusted furniture were bought to fill those adjusted homes, it would all be quite encouraging, now wouldn't it?

How is it that the 'adjusted' numbers have consistently risen, while the 'unadjusted' (the actual numbers) have fallen?

What do you suppose forms the basis for the adjustements and has changed that the actual numbers for several months would actually move in the opposite direction of the adjusted numbers?

[MissAmericanPie]

Followed by the housing crash in 85.

[dangus]

Yes, but the *significance* of the housing starts numbers is what they mean for the economy and the future of the housing market. People don't start houses as cold weather approaches.

If you want real numbers that demonstrate the strength of the economy, compare December '03 to December '02. This is "year-to-year" growth.

Now. maybe all of 2002 was lousy, and December was the one good month. Or, vice versa. Rather than compare December '03 to a single past December, they consider what share of the year's total starts December *normally* gets. That's where they get the adjusted figures from. No-one's *assuming* anything: They're looking at HISTORY, not the future when they adjust.

>> How is it that the 'adjusted' numbers have consistently risen, while the 'unadjusted' (the actual numbers) have fallen? >>

No-one wants to start to build when cold weather is coming. Less daylight, hard ground, it takes longer, and you get higher labor costs. If you start a home in September, you won't finish it (on average), the crews will be there through January and February. So. most people start building a home as soon as the ground softens in March. This happens every year, and no-one is changing any assumptions

>> If people actually bought adjusted homes and construction workers had jobs to build adjusted homes and adjusted furniture were bought to fill those adjusted homes, it would all be quite encouraging, now wouldn't it? >>

It's true that people buy fewer sofas in December than they do in June. On the other hand, *overall* retail (i.e., not including homes) sales are stronger in December than in June. People buy more coats, and toys, and food and appliances, and, jewelry, and, surprisingly, even car sales are strong. But retail sales are adjusted seasonally, also. SO when you hear that December retail sales weren't as strong as many had hoped, keep in mind that they still are up very strongly over the summer. This "seasonal adjustment" stuff works both ways.

THE MORAL: Economists don't always have your best interest in mind; they think like the corporations that pay them. SO, we can feel like they're full of sh!t when they cheer "stable labor costs." (IOW: they say, "Yay! we can underpay you!") But they are not stupid, either.

[Starwind]

No-one wants to start to build when cold weather is coming. Less daylight, hard ground, it takes longer, and you get higher labor costs. If you start a home in September, you won't finish it (on average), the crews will be there through January and February.

Then why does the adjusted data show more starts in Jan '03 than in Dec '02?

So. most people start building a home as soon as the ground softens in March. This happens every year, and no-one is changing any assumptions

If no one is changing assumptions, then on what physical or economic "constant" is the assumption based, that it never changes?

Actually, if you read and compare the adjusted and unadjusted data, you would see that in fact, adjusted (assumed) starts grew steadily from Dec '02 thru Dec '03, dipping one time in Feb '02.

Actual Starts however, Declined Dec '02 thru Feb'03, rose to June July, and declined to Dec '03.

So there is an actual divergence. Adjusted starts rising consistenly the entire year (including the cold months), but actual starts declining since mid-year.

[dangus]

>> Then why does the adjusted data show more starts in Jan '03 than in Dec '02? >>
Ummm... the economy improved? That's the point of adjusting the data: to make sure it reflects nothing other than the economic trend.

>> If no one is changing assumptions, then on what physical or economic "constant" is the assumption based, that it never changes? >>

I tried to explain this to you. The (relative) constant they are based on is what share of the year's total typically happens within that period. Look at it like this: December is 8.5% of a year, but (let's say) housing starts in December are normally only half of what they are for an average month. So they "adjust" December's housing starts by doubling them, so they can be compared evenly with, say, August. This is not based on predictions, but on seeing that December sales are *always* less than August sales.

>> Actually, if you read and compare the adjusted and unadjusted data, you would see that in fact, adjusted (assumed) starts grew steadily from Dec '02 thru Dec '03, dipping one time in Feb '02>>

Lose the idea that "adjusted" means "assumed." I don't know where you got it from, but it's outright false. And that enroumous falsehood is blocking you from comprehending anything else, apparently.

>> Actual Starts however, Declined Dec '02 thru Feb'03, rose to June July, and declined to Dec '03. >>

Yes, just as I said: people don't like to build homes in the winter. In March, the ground softens and the promise of warm weather approaches, and housing takes off, and housing starts take off until July. But September, or even August, people cannot finish what they start before winter, and housing starts head downhill. EVERY YEAR. LIke I said, this isn't prediction, it's history.

[Starwind]

In March, the ground softens and the promise of warm weather approaches, and housing takes off, and housing starts take off until July. But September, or even August, people cannot finish what they start before winter, and housing starts head downhill.

Then why didn't the adjusted numbers head downhill? Why did they instead increase?

[dangus]

>> Then why didn't the adjusted numbers head downhill? Why did they instead increase?>>

Because the decrease happens every year. But this year, the decrease was very small.

[r9etb]

You know, I'd be a lot happier about these numbers if the results of these "starts" were not so damned ugly....

[Starwind]

You stated:

But September, or even August, people cannot finish what they start before winter, and housing starts head downhill.

And I asked:

[>> Then why didn't the adjusted numbers head downhill? Why did they instead increase?>> ]

And you answered

Because the decrease happens every year. But this year, the decrease was very small.

To re-cap; I asked why the adjusted numbers increased and your answer was because they [I assume by 'they' you mean unadjusted numbers] decrease every year.

The adjusted numbers didn't decrease. They increased. I asked why. You answered because the decrease this year was very small.

So what kind of 'adjustements' cause an 'actual small decrease this year' to be continually adjusted upwards?

Seriously, think about what assumptions are being made in the adjustments for the last 6 months to report the opposite trend? Which is reality?

[dangus]

UNadjusted housing starts head downhill every year. When UNadjsuted housing starts declined less than most years, the result was that ADJUSTED housing starts increased.

Look at it this way.

Say a farmer gets a few crops a year out of his farm. His fall crop is always smaller than his summer crop.

But this fall was a great year. He grew almost as much in the fall as he did in the summer. He excitedly tells his creditor, "We've got a great crop! our yield is up!"

Understand this: His *adjusted* yield is up. After a good, but unexceptional summer crop, he grew far more this fall than he normally did. On the other hand, he still grew less than he did during the summer. Every year, his fall crop is smaller than his summer crop, but this year he did so well that his fall crop is almost as big as his summer crop.

If his summer crop was 10% above his normal summer crop, but his fall crop was 40% above his normal fall crop, he will tell his creditor that his productivity is up 30% since the summer, even though his "unadjusted" fall crop is still smaller than his "unadjusted" summer crop. The point is: It's good news.

And like the housing starts "increase" is good news.

[dangus]

>>You know, I'd be a lot happier about these numbers if the results of these "starts" were not so damned ugly....>>

Blame that, and the construction layoffs on the illegal immigration. No seriously... the construction industry has been flooded with low-pay Mexican immigrants. Those that pay good wages are losing ground to the cheap-o, Mexican-labor outfits.

The cheap-o outfits are lower quality, but few people understand why they should pay $100,000 more from a home; they don't appreciate the difference in housing construction quality.

[r9etb]

Blame that, and the construction layoffs on the illegal immigration.

You can't blame ugly, packed-together houses on illegal immigrants. Ugly comes from the architects who design them. Packed together comes from the developers putting as many houses as possible on a building site, because doing that maximizes their profits.

[dangus]

Hehehe. If they're *built* ugly and falling apart, blame the work crews. If they're *designed* ugly, blame overcrowding.

O yeah, that's because of immigration too!

(Please note the presence of my tongue firmly in my cheek.)