Veteran readers know one of my favorite charts is to juxtapose housing starts vs. unemployment. I forgot how I ran into this particular chart or what compelled me to contrast the two, but many years ago I did, and what I found is that housing starts typical forecast unemployment by 6-18 months. Ergo, if housing starts tank, you can expect unemployment to spike within the next year and 1/2.
Though I could speculate, I do not know the precise reason why housing starts forecast unemployment so well. I do know, whether the correlation is spurious or not, it's been a pretty good predictor in the past which is why I wanted to revisit and update the chart. I also wanted to revisit the chart because this recession is different than most others (duuhhhh!) in that unemployment has remained stubbornly high. Instead of a traditional recession where (cough cough, WHEEZE WHEEZE) FREE MARKET FORCES are brought about to take advantage of lower labor costs and lower cost opportunities presented by recessions and bring about a SHARP drop in unemployment, this time around it has not happened as unemployment stalls above 9%.
The question is of course, "when will unemployment come back down?" and this chart tells us the answer;
Not any time soon.
Notice that housing starts have also stalled. Unlike recessions before, when housing tanks, but then quickly recuperates, this time around it has not only tanked, but stayed tanked. Housing is still down and out, and assuming the correlation between housing starts and unemployment holds, you can expect unemployment to be hovering around 9% for some time to come.
Now this is GOOD NEWS in some regards because many political pundits (lacking the super awesome economic genius I am gifted with) are worried that Obama will be like Ronald Reagan. Reagan during his first term had very low approval figures just like Obama, but then an economic boom occurred in the third year of his first term, vaulting him into a second term. Those of the right side of the political spectrum are worried this could happen with Barry.
However, that is unlikely to happen. If you look at housing starts under Reagan in both parts of the double dip recession/Volcker Recession, housing starts recovered immediately. Additionally, they never went below 800,000. In Obama's case, housing starts tanked to 500,000 and have consistently stayed down there for over 2 years.
Of course, I don't know why housing starts track unemployment so well. And I am nowhere near as "educated" as the likes of Christine Romer and Paul Krugman. But so far this chart is batting about .925 while Keynesian-addicts such as Romer and Krugman are batting at about .000.
So those of you on the political right need not worry. The economy is not going to recover in time and strongly enough to the point it will save Obama. More importantly though however, is that we should not be cheering on a moribund economy so we can defeat somebody for political gain. We should be cheering this on because it's simply going to be another metric ton of empirical evidence that socialism and Keynesianism does not work, and free markets and a free people do. Of course "tell them that."
"Speculation"
ReplyDeleteNew house:
Employs people to build it
Buys lots of products and materials ( from concrete to dishwashers)
Creates a New tax base
Needs infrastructure - roads, sewers
There someone buys it, there is a profit, etc.. etc...
So you have whole industrial base behind this. Which means it's kind of a nothing succeeds like success thing the faster you go the more benefits the more benefits the faster you can go...
Hence the immoral aspect of the those that promoted and abetted the artificial extention of the housing boom - a huge malinvestment of resources extending throughout the world economy that got cut off long after it should have tapered off or busted more gently.
Sigh.
"Whether the correlation is spurious or not, it has been a pretty good predictor" is an amazingly ignorant comment.
ReplyDeleteI like your graphs. But I feed these data into SAS, and, there's no correlation. There just isn't. Sorry. One thing does not predict the other. If I show a human brain two plots of random noisy data it will become convinced that one plot predicts the other. That's all this is.
"many political pundits are worried that Obama will be like Ronald Reagan."
ReplyDeleteMany political pundits smoke crack. Bambi won't be like Reagan no matter what he does. I guess there might be a Clinton analogy, since Clinton's numbers were pretty bad in 1993-94 and he got reelected anyway.
I disagree with your thesis. Obama is not an ordinary president. He is a religious/messianic figurine. A realization of the progressive ideal (affirmative action). Ergo, he is evidence of a perfecting world; ergo, he is a conduit to an a left-wing eschatology. He could have just pushed a pencil around his desk for four years (the country would have been better off he had) and he would still have a strong contingency ready and willing to vote - dare I say die - for him.
ReplyDeleteDo you think that if Jesus ran for president his policies would matter?
It's rational self-interest versus religious eschatology - guess who wins?
I question the wisdom of using housing starts as an indicator of overall economic health when the housing sector is so utterly screwed up right now.
ReplyDeleteIt's like calculating the purchasing power of the dollar using a basket of precious metal commodities. It'll be a pretty good indicator most of the time, but in the middle of a gold rush it becomes much less accurate.
I agree with your conclusion, just not with your use of housing starts to arrive at it. I think we're not recovering because of taxation, regulation, and fear of more of the same. Obama is an avatar of left wing populist bullshit, which does very little to encourage confidence and investment from the entrepreneurial types.
I'm finding it fitting that the word verification I'm being asked to type in is "spited."
See Austrian business cycle theory. Credit boom goes into mortgages, to some degree, therefore housing is very sensitive to credit expansion. When credit expansion ends, two things happen:
ReplyDelete(1) housing tanks as do all industries that are sensitive to credit changes (usually earlier stages of production, but thanks to the wide usage of mortgages housing too) and
(2) a recession starts, as a clearing, equlibrium-establishing market process after previous unhealthy credit expansion that caused malinvestment (in housing, among others). And recession = unemployment.
OK, so if its clear that SAS says there's no significant correlation, and if Ortega is right and Austrian Business cycle predicts a correlation, then we can conclude...
ReplyDeleteYeah, and SAS never led a bunch of econometrician automatonic brainwashed idiots into a ton of Black Swans that completely negated their "triple doctorate in statistics and econometrics" now did it?
ReplyDeleteI mean, SAS is almighty and 100% encompassing. Right? I mean it's SAS.
How could a finite mathematical model POSSIBLY be wrong!
in the meantime SAS/model freaks, why don't you program a model to tell you what to do to find a date and swoon women. I mean, if SAS can predict entire economies, then it certainly can tell you how to get a date.
But that's also not smart, and articles on quants and the Gaussian copula document that. The math was fine. People who didn't understand it used it wrongly. The modellers, in many cases, objected. A couple of 'em -- see The Big Shortparticular pattern in housing starts presages a fall in employment. Also possible.
ReplyDeleteBut unemployment does not appear to be correlated with housing starts in any way we normally mean "correlated," and if you want to talk about some freaky-deaky "economics" that has no math in it, fine, but it's not very useful for much.