Yesterday I remarked that Okun's Law seems to be broken. I think I should provide more background on this. Of course "Okun's Law" isn't a scientific law at all - it's just a rule of thumb. But it' a consistent rule of thumb that is worth remarking on when it breaks down. It states that for every two percent decrease of output from potential output, we can expect to see the unemployment rate rise by one percent. We're not seeing this now (and I should note - people have been remarking on this for a while now, it's just gotten new attention recently, even if not by name from all commenters) - unemployment is a lot higher than Okun's Law would predict. It's just a rule of thumb - there's nothing deeply existentially troubling for the economist to see that it's not working. But it is grounded in enough sensible economics that it's worth asking "why is this time so different?"
It is useful to note that this break-down is happening in a lot of places. Greg Mankiw suggests that the Phillip's Curve might be exhibiting unusual behavior too (this is not new misbehavior for the Phillip's Curve, a more regularly misbehaving rule of thumb).
Mark Thoma and David Altig note another misbehaving rule - the Beveridge Curve, which relates unemployment to job vacancies. Like the Phillip's Curve, it's not particularly surprising that the Beveridge Curve is misbehaving. It shifts up and down with the matching efficiency of the labor market. So this change is the least mysterious.
Russ Roberts remarks on the same disconnect between employment and output, and he seems to think that this is troubling for Keynesians. He circles around a couple (IMO) relatively unimportant issues for a while, and then he writes this:
"I’m not saying Keynes (or Obama or Larry Summers) is wrong because all spending does is bid up prices and wages. I’m not trying to prove that the stimulus failed. I’m challenging the standard macro textbook story that says aggregate demand leads to output that leads to employment."
So we have a situation where Russ is not trying to say that employment isn't going up because stimulus is channelled into wage and price changes (a little redundant, but nevertheless...). He makes it clear that he's not making that argument. The stimulus-to-output chain may be weak for him, but it's plausible. His problem is how "the standard macro textbook" explains the output-employment link. Okun's Law, in other words. Roberts goes on:
"The textbook aggregate demand story ignores how people view the future– the animal spirits–the confidence employers, investors, and consumers have about the future. If employers are anxious about the future (partly because the government is running up debt and because the regulatory environment is highly uncertain and partly because we’ve had a run of bad times), then spending doesn’t obviously create jobs."
I did a double-take when I read that, because it was easily the most Keynesian thing I've ever read on Cafe Hayek. Concern about the future means that money may not go to prices or employment - it may go to liquidity which ultimately has it's roots in fear of "the dark forces of time and ignorance that envelope our future". Of course Roberts doesn't seem to recognize that this is a very Keynesian insight. Perhaps he does, but there's no indication of it.
What does this have to do with all the other rule-breaking? The best explanation I've seen so far for why changes in the Beveridge Curve, the Phillip's Curve, and Okun's Law all relate to liquidity preference is Roger Farmer. I really, really, really need to get this man's new book.
So the obvious question in my mind is "did Keynes forsee this". Unfortunately, I can't say right now but I'm going to check up on it this weekend. Hicks clearly didn't. For Hicks, liquidity preference broke the link between money and output, not between output and employment. In other words, in the IS-LM model, Okun's Law should work fine because all the disruption of liquidity preference happens before output even comes into the picture. I know Hayek did critique Keynes for not presenting a detailed enough explanation of the relationship between employment and output (does anyone know where he said this... I'm very, very curious now). What I don't know is whether Hayek's critique was valid or not. Did Keynes have as simple of a view of the output-employment relationship as Hicks (in which case the critique is valid), or didn't he?
Page 25 of the General Theory gives his equations on employment and demand. If that's not silly, I don't know what is.
ReplyDeleteFor those interested in Mattheus's reference you can look here: http://www.marxists.org/reference/subject/economics/keynes/general-theory/ch03.htm
ReplyDeleteWhat Mattheus calls an "equation" is really just an undefined function - we're not really sure what the functional form is, which makes sense because it's not likely to be simple in real life.
I was thinking of discussion later in the book, because the essence of my question is "Russ Roberts seems to hit on the fact that liquidity preference is not just going to affect output - corprorate liquidity preference is going to effect labor demand as well. Hicks doesn't (I don't think) incorproate liquidity preference in the labor demand function. Does Keynes?
Here he certainly talks about labor demand as a function of investment demand and demand for consumer goods. This seems to reflect the Hicksian framework - that employment will be impacted by liquidity preference through the effect on output. Doesn't it? I think it's probably pretty likely that in forming labor contracts, though, corporate liquidity preference is going to be a factor in the willingness to strike new contracts.
That seems to be under-emphasized here, but again this is the introductory chapter only. Which is why I say I think we'll probably need to look deeper than just page 25.
I'm sure you will have to look deeper than page 25; Keynes' equations on labor and investment demand seem to be relevant to this discussion though. But I'm not a Keynesian, so who knows.
ReplyDeletePS: What do you mean by 'output?' Are you measuring output simply by the totality of goods and services produced or is there any question of how relevant those goods are. For instance, are we talking about nominal output or value output?
Not nominal output, but market-valued output. So, real output.
ReplyDeleteWhat do you mean by "value output"? It is "value output" insofar as real price changes represent real changes in value, I suppose. But I'm not sure if that's what you mean by "value output".
I'm talking about real production, real GDP.
I have recently run a series of Okun's Law calculations for a forthcoming paper and I found that, for the US, it runs around 3:1 on a national level, but they are still cointegrating variables. I looked at it at the state level and did not get anything that high suggesting a labor mobility story. Productivity has also played a role in reducing the link between output and unemployment. This paper by Daly and Hobjin details this idea.
ReplyDeletehttp://www.frbsf.org/publications/economics/letter/2010/el2010-07.html
I'd maybe like to do a paper looking at Okun's Law on a cross-country basis, taking into account productivity and k/l intensity differences. I know G7 all run around 3 these days too.
That's interesting.
ReplyDeleteOne of the most common interpretations of the Great Depression at the time was "technological unemployment" - essentially a rising productivity story. I think these kinds of explanations, at the very least as aggravating factors, are underappreciated because they are more historical explanations than theoretical ones.
One way to think of a productivity story is that it's just about the only really plausible RBCT! Shifts in real factors that actually involve ad hoc technological progress rather than technological regress!
I'm talking about real production, real GDP.
ReplyDeleteThis doesn't really answer my question. GDP simply measures aggregate production; that's not quite what I'm asking.
During the low interest rates of the business cycle, GDP soars because consumption and investment are both increasing. According to you that is "increased output" but not to me. The market is creating malinvestments; it is creating plenty of items that have "market-value" (and in that sense are real goods), but they do not represent true tastes of consumers. Can you spot what I mean by value output?
Mattheus -
ReplyDeleteOK - now I see what you're talking about. But there are major problems with that sort of approach. How are you going to identify value if not with prices? To me, it seems safer to use real GDP measures and say "there are business cycles and bubbles in the economy" then to try and guess at what "value output" is. How would you know?
Besides, the fact that even malinvestments have "market value" is a reflection of the "true tastes of consumers". The problem isn't that it doesn't reflect the "true tastes of consumers" - the problem is that those tastes are temporary, distorted, volatile, etc.
the problem is that those tastes are temporary, distorted, volatile, etc.
ReplyDeleteI should have been more specific then. Clearly, the temporary and distorted tastes are not the true tastes that would exist in an unhampered market economy. And if we take an unhampered market (and the consumer demand that would go with it) as a kind of Weber-esque "ideal type", then the production during bubble economies doesn't measure up.
You wouldn't know what the value of production is given a bubble economy because tastes and production are distorted - that's the point. You can't use a yardstick for production in a bubble and "subtract the bubble" to arrive at a real value of output. The yardstick is useless at this point. You would have to concede that production during bubbles (and any government manipulation) is value-destroying and therefore GDP is not an accurate measure of the value of production.
I could potentially agree with you if I thought monetary distortion by a central bank was the only cause bubbles and the business cycle. I really don't see how that can be true, and if there are other causes I don't see how, say, speculation, animal spirits, etc. doesn't count as real valuation by real market actors. It shifts, it's volatile - but it's as a real as non-shifting, non-volatile valuations.
ReplyDeleteSo in spirit I suppose I'm with you - I just don't think it's that clean.
Daniel,
ReplyDeleteThe following post on the "Organizations and Markets" blog (an ostensibly Austrian-themed blog) might point you in the direction of Hayek's critique. I myself haven't read the work to which it refers, although, like many Austrian critiques of Keynesianism, it seems to be a restatement of well-established classical arguments.
http://organizationsandmarkets.com/2009/02/10/demand-for-commodities-is-not-demand-for-labor/
"John Stuart Mill’s profound insight that demand for commodities is not demand for labor, which Leslie Stephen could in 1878 still describe as the doctrine whose “complete apprehension is, perhaps, the best test of a sound economist,” remained for Keynes an incomprehensible absurdity (Collected Works, vol. 9., p. 249)
[I believe this to be the Contra Keynes and Cambridge volume, which is fairly easy to track down]
In essence I believe it's a restatement of JS Mill's Fourth Fundamental Principle - Garrison has a brief discussion of it in Time and Money on page 248-249.