Monday, August 4, 2014

The Natural Rate Does Not Equal NAIRU: A Reminder

I am not keen on poking at a post appearing on Angry Bear, a friendly blog that often links favorably to posts here by many of us and with which I generally agree.  But maybe it is precisely because of this usual agreement that I am posting here an extension I made of a comment there on an August 1 post by Edward Lambert arguing that not only is 6.2% the US natural rate of unemployment, but that it is also the NAIRU (Non-Accelerating Inflation Rate of Unemployment), because, hey, the two are equal to each other, period.

Much of the post is quite good and interesting.  Lambert has a model that focuses on such things as capital stock capacity utilization and such measures for land aalso, as well as the unemployment rate.  Based on his model, levels of capital stock capacity utilization are now in ranges that look like full employment of capital stock, at least.  This may reflect the low rate of capital stock formation in recent years, a supply-side damage coming from insufficient aggregate demand over an extended period of time, but Lambert may well be on to something. 

Let me also state that I am not necessarily opposed to the idea of a "natural rate of unemployment," although I think that one should not make too much of it, which Lambert does (along with many others).  I think it is more useful to think of a "natural rate of employment," more relevant to the current state of low labor force particpation.  Indeed, Lambert trumpets that he beat out the Brookings Institution in forecasting the most recent unemployment rate.  They (and "most economists") said the UR would go down while he said it would go up, and, whoop de doo!, he was right.  However, it turns out that this was not due to some failure of job growth, but due to a favorable response on the labor force participation side.  There are plenty of reasons for employment to continue rising without any inflationary presssure in the near term, something Lambert actually admits, but what happens to the unemployment rate per se may really mean nearly nothing.

As it is, the original definition of the natural rate of unemployment by Milton Friedman back in 1968 was simply a rate that the economy would tend to go to if left to itself with "neutral" policy (if that can be defined) at any point in time.  Such may well exist. The issue is whether this natural rate has any signfiicance for policy, and for that things are not at all convincing.  Indeed, from the very beginning, others who were pushing similar ideas, notably Edmund Phelps, were heavily caveating the concept with the fact this natural rate is highly endogenous to the recent rate, particularly for labor force participation reasons, many of those related to how labor market skills (if not labor skills themselves) can deterioriate during lengthy periods of unemployment, an issue most relevant to the current situation.

My big complaint here has to do with identifying this natural rate with NAIRU, a highly problematic concept.  Friedman vaguely suggested such a link in his original talk, but he never really made a case for why inflation would not only increase, but that the rise would accelerate.  Indeed, aside from some handwaving about expectations, the case for why such an acceleration should occur at URs below the natural rate has never ever been made anywhere in any rigorous or coherent way by anybody.  That said, the subsequent inflation of the 70s was blamed by many on the supposedly excessively stimulative policies of the 60s, even as the acceleration of that inflation coincided with rising unemployment (and unemployment rates) in the notorious staglation.  But this somehow came to be identified with Friedman being right, and the identification of the natural rate with NAIRU entered macro textbooks at nearly all levels despite the lack of clear or rigorous or even empirically validated arguments supporting it.  It became something "everybody knows," (or should know) with even progressive economists who should know better such as Lambert falling for this empty nonsense.

Quite aside from the warnings from Day One about endogeneity of the natural rate to the recent actual rate, the famous episode of the 90s when Greenspan, urged on by Janet Yellen btw, allowed continued stimulative monetary policy even as the UR went below its then widely advertised natural rate and we saw falling inflation rather than rising, much less accelerating.  Many considered this to be the death knell of this identity, with some such as James Galbraith declaring it publicly in print in the Journal of Economic Perspectives in 1997 in "Time to Ditch the NAIRU" (although Jamie made it clear then that he had never thought much of the concept, and I don't think he has ever thought much of the natural rate concept either).

Now Lambert does recognize the Greenspan episode, but elides the issue  by saying that it set up the economy for the next recession, which took a good half a decade to show up.  What caused that recession in most peoples' books was the collapse of the dot.com bubble.  Now, maybe it was the stimulative monetary policy that goosed that bubble to ridiculous levels, but it had been going on for some time and was associated with real capital investment in that leading sector that did generate a lot of non-inflationary growth with lowered unemployment and reduced inequality.  Not too bad.  And the end of it was not associated with any increase in inflation.

Lambert's position is given in his final remarks:

"...when unemployment pushes below its natural level, the economy tends to overheat and become unstable.  At which point, the economy would like to return to its natural level.  The paths of that process can involve inflation, reduced output, falling capacity utilization and even tighter monetary policy."

Well, just two final comments on this.  One is that he leaves out what happened last month and also what happened during the three year period in the mid-80s when employment and output grew but the UR remained stuck at 7.0%, namely that labor force participation can increase, particularly if it has been "unnaturally" suppressed for an extended period.  The"natural rate of unemployment" may not even be tied to any particular level of the economy, much less its "natural" level.

The other comment is that in fact those declines in output and capacity utilization that have occurred in the past when "overheating" has occurred have usually been triggered by the tighter monetary policy, which has usually been a result of either actual inflation picking up or in many cases the Fed (or foreign central banks such as the ECB in recent years) getting nervous that inflation might pick up, in some cases spooked into doing so by taking too seriously all these fantasy stories so embedded in so much literature about the natural rate equaling the NAIRU.

This is a fantasy that needs to be exposed once and for all and banished from the textbooks once and for all, and indeed from publich discourse more generally, particularly by economists.

Barkley Rosser

7 comments:

Bill H aka run75441 said...

Hi Prof. Rosser:

I was just reading your response and what stopped me was the comment on Fed Rates and Greenspan. In the nineties the Fed Rate pre-2001 was around 4-5% from 94 to 99. It was only slightly before the recession (8/1/99) the Fed Rate started to climb and went to ~6.5% in 8/1/2000. I am not sure if such of an increase in 13 months is enough to push us over the edge; but, it appears it might have assisted in doing so as Greenspan took every action possible to drop it down to 1.75% 4/1/2002. You have to tell me as I do not hold the same credentials as you do Prof. Rosser.

1996-01-01 5.56
1996-02-01 5.22
1996-03-01 5.31
1996-04-01 5.22
1996-05-01 5.24
1996-06-01 5.27
1996-07-01 5.40
1996-08-01 5.22
1996-09-01 5.30
1996-10-01 5.24
1996-11-01 5.31
1996-12-01 5.29
1997-01-01 5.25
1997-02-01 5.19
1997-03-01 5.39
1997-04-01 5.51
1997-05-01 5.50
1997-06-01 5.56
1997-07-01 5.52
1997-08-01 5.54
1997-09-01 5.54
1997-10-01 5.50
1997-11-01 5.52
1997-12-01 5.50
1998-01-01 5.56
1998-02-01 5.51
1998-03-01 5.49
1998-04-01 5.45
1998-05-01 5.49
1998-06-01 5.56
1998-07-01 5.54
1998-08-01 5.55
1998-09-01 5.51
1998-10-01 5.07
1998-11-01 4.83
1998-12-01 4.68
1999-01-01 4.63
1999-02-01 4.76
1999-03-01 4.81
1999-04-01 4.74
1999-05-01 4.74
1999-06-01 4.76
1999-07-01 4.99
1999-08-01 5.07
1999-09-01 5.22
1999-10-01 5.20
1999-11-01 5.42
1999-12-01 5.30
2000-01-01 5.45
2000-02-01 5.73
2000-03-01 5.85
2000-04-01 6.02
2000-05-01 6.27
2000-06-01 6.53
2000-07-01 6.54
2000-08-01 6.50
2000-09-01 6.52
2000-10-01 6.51
2000-11-01 6.51
2000-12-01 6.40
2001-01-01 5.98
2001-02-01 5.49
2001-03-01 5.31
2001-04-01 4.80
2001-05-01 4.21
2001-06-01 3.97
2001-07-01 3.77
2001-08-01 3.65
2001-09-01 3.07
2001-10-01 2.49
2001-11-01 2.09
2001-12-01 1.82
2002-01-01 1.73

Pre-recession if I remember correctly, Participation Rate climbed to 67.1% with a U3 of ~4.0%. After the recession, Participation Rate was around 66.7/8 and a U3 of 4.7% more than likely because of PR. From Fall of 2001 till now we had a downward trend.

If we go back to 79 (BLS data), there has not been a period where U3 was in the ~4% area. It was always higher. It was in 97 we had our first taste of it at 4.9%and it lasted to 2001 (4.7%). In 2000, we hit 4%.

Where were you New Years Day of 2000. In Michigan we had ice on the highway. I saw one accident with a guy laying on the highway and the police tending to him. I was motoring into Yazaki NA to check out the Y2K issues for my systems as I ran a 100,000 sq. foot, $200 million/year component shipping warehouse. Y2K never materialized as they claimed and later that year, business started to decline.

So is Edward wrong in assuming Greenspan had a hand in this? I do not think so. Neither do I believe it took 5 years for the 2001 recession to happen when I look at the data. It appears the collapse/recession occurred in 13 months.

My $.02.

rosserjb@jmu.edu said...

Bill,

Actually I said nothing at all about what Fed funds rate or targets were or are specifically. I expect the numbers you provide are correct. So, they were at a low at the beginning of 99 at 4.63%, which was lower than previous years, when the rates were mostly between 5 and 6%. The rate rose by more than a percent to 5.85% at the beginning of March, 2000, just before the stock market peaked and then crashed. I am a bit mystified that they continued to raise the rate for a few more months, peaking out at 6.54% four months in July, staying up near there until after the election. That the rates were lowered sharply later reflected trying to offset the recession that had started in 2000, with them becoming more aggressive about pushing down rates after 9/11/01 because of fear that event would seriously damage the US economy. Mostly rates were pretty stable from the mid to late 90s, which shows that Greenspan did not tighten then, although apparently did somewhat around 88 and 00.

Regarding labor force participation, those rates remained remarkably stable during that period, around 66 to 67%. The rate was 66% with almost no variation until late in 2008, when it began to drop and is now at 63%.

Again, the bottom line is that it was 96 when Greenspan was pressured to tighten monetary policy because of the unemployment rate declining. That the fed funds rate did not change much during that period is the sign of this lack of change in policy, which was supported by Yellen. I would note that from the series you show, from the beginning of 96 until late 98 the rate never left the zone of 5.21 to 5.56%, which are the first two observations shown, a very small variation. Policy was basically constant during that period.

Myrtle Blackwood said...

Lambert's position is given in his final remarks:

"...when unemployment pushes below its natural level, the economy tends to overheat and become unstable. At which point, the economy would like to return to its natural level. The paths of that process can involve inflation, reduced output, falling capacity utilization and even tighter monetary policy."

Hmmm... A simple solution to too low a level of unemployment is to change the definition of UE. Currently, if you're getting one hour of (paid) work per week in Australia you are defined as 'employed'. So, to increase the national levels of UE simply change the definition of 'employed' to 2 hours of work per week, or even more.

Honestly, things could be so simple!

Myrtle Blackwood said...

Errr...I meant change the definition of 'employed'.

Bill H aka run75441 said...

Prof. Rosser:

"Now Lambert does recognize the Greenspan episode, but elides the issue by saying that it set up the economy for the next recession, which took a good half a decade to show up. What caused that recession in most peoples' books was the collapse of the dot.com bubble. Now, maybe it was the stimulative monetary policy that goosed that bubble to ridiculous levels, but it had been going on for some time and was associated with real capital investment in that leading sector that did generate a lot of non-inflationary growth with lowered unemployment and reduced inequality."

There are several coinciding events going on here; one being the prep for Y2K which involved reprogramming major data bases in a short period of time preceding 2000, the purchase of new computers to replace old ones, and the redesign and testing of old programs to meet the requirements of using the date 2000; secondly, the increase in Fed Rates one year prior to the recession to slowdown a perceived over heating of the economy; and thirdly the non-event of Y2K which pretty much let the air out of the big Y2K balloon shortly after January 1, 2000. My story of going into work New Year's day was repeated globally to make sure programs worked.

We saw 67% PR in 96 and flirted with it going into 2001. It hit a high of 67.3% January of 2000. The Y2K event preparation was late 98 and all of 99 and it was estimated corporations spent ~$300 billion to prepare for it. Greenspan started to increase Fed Rates 8/1/99 in what I believe to be an attempt to slow what he perceived to be an overheated economy.

Three events taking place over a short period of time. Of course, we will never know what would have happened if Greenspan had let the economy play out. Y2K would have still been a nonevent. Thousands of programmers, etc. would have hit the streets lowering PR and increasing U3.

Did Greenspan have a hand in the 2001 Recession as Lambert suggests? I think Greenspan did as he over reacted and did not recognize the short term events (Y2K, business expenditures to meet it, hiring, etc.) taking place late 98 through 2000. Again, this is just my opinion on looking at what happened.

rosserjb@jmu.edu said...

Bill,

I think the only way one can argue that Greenspan had a hand in the 2000 recession was that the failure to noticeably raise interest rates back around 96 when the issue realy went down aggravated the dot.com bubble, which it probably did to some extent, and whose collapse is widely thought to be the main culprit in that later recession. I think all the Y2K stuff played a relatively minor role in what went down.

Myrtle,
Clearly one can play with various definitions of the unemployment rate. As it is for the US, I think using the employment rate (percent of working age population working) would be preferable to the measures currently used.

Myrtle Blackwood said...

America's very sharply rising dependence on imported energy is likely to be a more significant factor in the 2000 recession. Bubbles of money chasing vacuums of real collateral.

* Diminishing returns on debt [the lifecylcle of financialisation): The expansion of debt was paralleled by rising earnings until the late 1990s. After that, household earnings continued rising, but the rate of growth was outstripped by debt, which more than doubled from 2000 to 2008.
http://www.zerohedge.com/news/2013-11-09/guest-post-our-era%E2%80%99s-definitive-dynamic-diminishing-returns

*US oil production versus US oil imports: a crossover in the early 1990s where the level of imports met and from then on exceeded oil production.
http://en.wikipedia.org/wiki/File:US_Oil_Production_and_Imports_1920_to_2005.png

Around the same time the GDP of China began to rise very sharply ( a way for capital to chase cheaper sources of energy and labour supply?):
http://www.businessinsider.com/15-more-mindblowing-facts-about-china-economy-2010-4#chinas-economy-grew-by-119-in-the-past-year-and-its-gdp-chart-has-gone-vertical-6