Okun’s Law
The Fed of San Francisco just published a note on Okun’s Law and the Unemployment Surprise of 2009.
In the paper they conclude that strong productivity was the main reason employment growth was weaker than the traditional relationship that Okun’s law implied.
Of course, we at Angry Bear have long known this. I have published this chart that shows that roughly before 1974 that a one percentage point growth in real GDP generated a 0.3 percentage point growth in employment. This is what Okun’s law is based on. But during the era of low productivity growth, 1974 to 1995 a percentage point growth in real GDP generated almost a 0.5 percentage growth in employment.
But since productivity growth rebounded in 1995, every percent increase in real GPD was accompanied by almost a 0.9% gain in productivity so that employment barely rose 0.1% — a significantly lower rate than Okun thought.
The data in the chart is the long term trend and ignores the cyclical pattern in productivity where productivity growth peaks in a recession or early recovery period and slows as the expansion continues. That is why productivity growth has long been widely considered a leading indicator. It is also why you get patterns such as the San Francisco Fed found for 2009, and why we now seem to have jobless recoveries.
Arthur Okun. Not Okum.
“[T]his chart … shows that roughly before 1974 that a one percentage point growth in real GDP generated a 0.3 percentage point growth in employment.”
Sorry. I’m confused. Employment growth isn’t charted.
But the increasing trend toward jobless recoveries seems to have begun with the ’81 recession.
http://www.asymptosis.com/wp-content/uploads/2010/01/EmploymentRecessionsDec1-480×311.jpg
I agree that increasing productivity is at the heart of the answer, but…?
seems like their cart is before their horse…it is increasing unemployment that leads to higher productivity…
I’ve had a few posts here on AG about the issue of advancing technology and the likelyhood of increasing unemployment in the future (see http://www.angrybearblog.com/2009/11/could-advancing-job-automation.html).
I suspect the FRBSF economists are picking up some quantitative evidence of this trend. The problem is that analyzing historical data is not going to be an very effective way to forecast future changes that may well occur at an accelerating rate. Most economists, of course, will not accept ANYTHING unless it is backed by analysis of hard historical data. I’ve written more on this at my blog: http://econfuture.wordpress.com.
The divergence between GDP and GDI is considerable right now. Okun’s Law is violated a lot less by GDI, which shows the economy has performed even worse through the recession and early recovery than does GDP. If the economy has, in reality, performed worse that suggested by GDP, then there is less explaining to do.
Further implications of GDI being closer to right than GDP would be that the US fiscal stance may be more out of balance than currently projected through the recovery, and that productivity growth has been less remarkable than has been reported, which means less fuel for a sharp rebound in hiring.
There is some evidence that GDI has been more reliable in recent years. It has require less revision and has fit better with NBER business cycle timing. Doesn’t mean it is right in this case, but GDI is more consistent with labor market data than GDP.
Okun’s Law is, in fact, a relationship that was derived from empirical data of an economy that is very different than the one we have today. In short, it was based on an economy dominated by manufacturing industries, while our economy since the 1970s has become increasingly dominated by service industries.
Today, many of these service industries enjoy regional monopolies/duopolies and are able to charge more and more for less and less even in a down economy. See Anthem/BlueCross/BlueShield (who raised premiums up to 40% in CA), and the TBTF banks (who raised credit card interest rates to as high as 30% when they are borrowing money for free).
By pointing to worker “productivity” to explain why Okun’s Law remains valid and useful, which it isn’t, we’re ignoring a lot of rent-seeking behavior in order to believe that we’ve “recovered” from recessions when, in fact, we haven’t. The consumer has merely paid more for less, and has been left poorer for it.
http://www.tradereform.org/content/view/2413/34/ Defining Okun’s Law also in laymen’s terms
http://www.economicpopulist.org/content/lets-chat-labor-productivity Economic Populist discusses the divergence as well.
Be cautious of the “Economic Populist” discussion. In it, we find references to offshoring leading to wage advantages which drive up productivity estimates. Productivity is measured in unit terms, not $ terms, so in the absence of a further mechanism being laid out to link wages over there with productivity over here, we have reason to doubt that Economic Populist has a firm grasp of the issue.
Outsourcing can work to raise domestic productivity, by steadily subtracting low-productivity jobs from the US total. That would raise measured productivity even if the productivity of remaining US workers was utterly stagnant. Wages may drive that process, but it is not directly because of lower prices that productivity – calculated from real business output – rises.
Kharris,
Thanks for the catch. I read too fast.
Rdan,
You should have asked kharris a simple question: How is unit output determined?
The Economic Populist is on the money with most of its commentary on this matter. I’ve reviewed a number of its recent posts including reference documents cited and find no principal fault with its observations including the various offshoring discussions.
The Economic Populist has provided some BLS references that should be reviewed in order to understand how unit output is determined. I recommend a careful read before accepting anyone’s negative opinion of The Economic Populist pieces.
The Economic Populist has raised the right issues.
I also recommend that you and others read Susan Houseman’s paper which The Economic Populist also cited.