Okun’s Law suggested that an increase in output would imply a fall of unemployment, and the relation was often assumed to be a 2:1 ratio in more recent times. Jon Hilsenrath suggested recently, in a WSJ blog, that the relation is broken. He argues that the average growth between 2007 and 2009, the recession years, was around 2.5% below its trend (the actual growth was slightly negative, around –0.22% in the three years), and that Okun’s Law would have implied and increase of unemployment of approximately 5% (he says little more than 5%, from 4.4% to 8.9%” [sic]), but it increased more, reaching 10.1% in October 2009. That is the basis for the claim that the law is broken. In fact, the average annual unemployment increased from 4.6% in 2007 to 9.6% in 2010, precisely 5%. His fears about the breakdown in the recovery are also exaggerated. What is really surprising is that he thinks that: “unemployment seems to be falling sharply.” I need to find these rosy goggles WSJ bloggers use.
Subscribe to:
Post Comments (Atom)
Inflation, real wages, and the election results
Almost everybody these days accepts at face value that the result of the election was heavily determined by negative perceptions about Biden...
-
"Where is Everybody?" The blog will continue here for announcements, messages and links to more substantive pieces. But those will...
-
There are Gold Bugs and there are Bitcoin Bugs. They all oppose fiat money (hate the Fed and other monetary authorities) and follow some s...
-
By Sergio Cesaratto (Guest Blogger) “The fact that individual countries no longer have their own currencies and central banks will put n...
I love it when Matias shows his sometimes latent quantitative nature. Well done!
ReplyDeleteWhich brings up a point...if you are anywhere near the U of Utah campus on April 22, show up at the HESA sponsored panel on Heterodox Econometrics -- why we should love and embrace statistical methods and how we should do so.
We plan to cover the space several ways...hearing first from those who are using econometrics in heterodox topics, including Matias.
Also, I plan to make the case for data-first methodology. Or at least taking your deductive model to the data...no anti-Popperianiasm here.
And Richard Fowles will have some provocative things to say about statistical inference...things you really should know if you have ever invoked a t or p.
So plan to come and express your quantitative self -- we may even toast Okun and Verdoorn.