Author
|
Topic: Understanding the NAIRU
|
Stephen Gordon
rabble-rouser
Babbler # 4600
|
posted 22 June 2006 06:45 AM
There have been any number of threads that have touched on the 'NAIRU'. These discussions are problematic for me, since there's a certain amount of confusion about what it is and how it's used to conduct monetary policy. So here it is, as it is taught in an intermediate macroeconomics course. Suppose that firms and workers are able to perfectly forecast prices - denoted by P - and that they have to determine the (nominal) wage W. They have a common frame of reference: the real wage (W/P). Labour demand decreases as real wages increase, labour supply increases as (W/P) goes up, and the labour market equilibrium is characterised by a standard supply-and-demand graph. The equilibrium is the point e, real wage is (W/P)*, and employment will be N*. Figure 1 If contracts could be written in terms of real wages, then that would be the end of the story: if the price level rises, nominal wages will rise with them in order to keep real wages at (W/P)* and employment at N*. Variations in prices will have no effect on employment. But suppose that for some reason, it's not practical to agree on just what the real wage is. There are lots of reasons to expect that this might be the case. For example, the only price that matters to firms is the price of the product that's being sold, while for workers, the relevant price would be the cost of living. Since they can't agree on what the real wage is, they are obliged to write contracts in terms of the nominal wage W. This distinction is important from firm to firm, but at the aggregate level, they wash out. So when they negotiate wages, they'll do so based on their expectations of inflation. Labour supply and demand curves are now drawn conditional on their expectations for future prices. If both firms and workers have the same expectations - and there's little reason to believe that they won't, since they have access to the same information - then they'll try reproduce point e in Figure 1. So if expectations are Pe=P0, the equilibrium will be e0, employment will stay at N*, and the nominal wage W0 will be set so that the real wage (W0/P0) = (W/P)* as before. If instead they expect Pe = P1, then the equilibrium is now e1, nominal wages will rise in proportion so that (W1/P1) = (W/P)* as before, and employment is still at N*. Both e0 and e1 describe the same equilibrium as point e in Figure 1. Figure 2 Suppose we start with prices at P0, and the labour and supply curves are drawn conditional on this. The equilibrum will be e0, and employment is N*. Now suppose that prices increase to P1. Eventually - we'll get to the transition in a minute - we'll get to point e1, where real wages and employment are as before. The long-run relationship between prices and employment is zero. Wages and prices will (eventually) move in proportion so that employment always returns to N*. Suppose now that we're at e0, and that there's an unanticipated increase in prices to P1. Firms will see an immediate rise in their prices, so labour demand shifts right away to ND(Pe=P1). What happens to labour supply?
Figure 3 There are two main versions of the story for what happens next: New-Keynesian: Nominal wages are fixed, and the new equilibrium is determined by labour demand - point ea. As workers realise that their real wages have gone down, they will negotiate higher nominal wages when their current contracts expire. Once everyone has had a chance to react to the price increase, we return to e1, the long-run equilibrium.
New-classical: Nominal wages are flexible, but workers don't yet know if an increase in the price of the widgets that they make will affect their cost of living. The labour market moves to eb: nominal wages rise, but since prices have risen more, real wages fall. Back in the day, there was a lot of debate about whether we went to ea or eb. But we eventually realised that the size of the increase in employment and in nominal wages is an empirical issue, and the field is pretty much dominated by data-based models these days. All we really need to know is that the rise in prices leads to a reduction in real wages, because for some reason or another, workers don't get W1 right away. Because real wages have declined, firms hire more workers, so employment and output go up. Of course, once workers understand what's going on, they'll demand an increase in their nominal wages, real wages go back to their old levels, and employment goes back to N*. The short-run effect of an unanticipated increase in prices is a decrease in real wages and an increase in employment. In the long run, there's no effect on either; all that happens is that nominal wages adjust to keep real wages at (W/P)*. What are the implications for monetary policy? It's clear that if the Bank of Canada can engineer a temporary rise in employment whenever it wants to by creating an inflation 'surprise'. Can this effect be made permanent? No. Remember that the reason employment went up is that inflation reduced real wages; once workers renegotiate their nominal wage, employment goes back down again. The only way you can keep employment at levels greater than N* is to continually increase prices beyond what workers had expected. Workers will of course revise their expections of inflation ever higher, and since the Bank will always try to inflate beyond what was expected, inflation would rise and keep on rising. This is why the unemployment rate associated with N* is called the Non-Accelerating Inflation Rate of Unemployment. It not just that inflation would go up if employment were greater than N* (or, equivalently, if unemployment were less than the NAIRU), it would continue to accelerate. So just what is the NAIRU? If demographics and participation rates were steady, the labour supply curve would be fairly stable - but it's not. If technical change, capacity and trade patterns were stable, labour demand would be stable - but it's not. The NAIRU is therefore some a function of all these factors (and more), and estimates of the NAIRU will evolve over time. We've pretty much lost interest in the question: since employment is a lagging indicator of the business cycles, comparing observed unemployment to an estimate of the NAIRU isn't be much of a signal for where the economy is. Some concluding comments: - The Bank of Canada can't influence N*. - Employment can be above N*, but only if inflation succeeds in reducing real wages. - Attempts to keep employment above N* will lead to runaway inflation. [edited to fix a subscript] [ 22 June 2006: Message edited by: Stephen Gordon ]
From: . | Registered: Oct 2003
| IP: Logged
|
|
Michelle
Moderator
Babbler # 560
|
posted 22 June 2006 08:17 AM
I wish I could understand this, Stephen, but it's way over my head, I guess because I've only taken a first year microeconomics course six years ago. I've been spending the last 15 minutes trying to wrap my head around your first section and I just don't get it.I don't understand in your first graph what the red line and the blue line represents. Also, I don't know what "NAIRU" stands for. What's a "nominal wage"? Sorry to be asking what are probably really basic questions.
From: I've got a fever, and the only prescription is more cowbell. | Registered: May 2001
| IP: Logged
|
|
|
unionist
rabble-rouser
Babbler # 11323
|
posted 22 June 2006 08:43 AM
quote: Originally posted by Michelle: I wish I could understand this, Stephen, but it's way over my head, I guess because I've only taken a first year microeconomics course six years ago. I've been spending the last 15 minutes trying to wrap my head around your first section and I just don't get it.I don't understand in your first graph what the red line and the blue line represents. Also, I don't know what "NAIRU" stands for. What's a "nominal wage"? Sorry to be asking what are probably really basic questions.
NAIRU is Non-Accelerating Inflation rate of Unemployment. "Nominal wage" is the dollar figure of your wage, not taking into account whether the cost-of-living is going up or down (which would tell you what your "real wage" was, or rather your purchasing power). For example, if you earn $18.50 per hour in 2006, that's your "nominal" wage. If you wanted to know how much that bought compared to what you were earning in 1996, it would be the same basket of items as a $15.00 per hour wage would have bought 10 years ago. So in 1996 terms (or you could choose other benchmarks if you wanted), your "real wage" in 2006 would be $15.00 per hour. Source: Bank of Canada Inflation Calculator Having said that, I wouldn't spend too much time on Stephen's essay if I were you. He says employment rises when real wages go down. And vice versa. That's enough to put this in the realm of voodoo economics for me.
From: Vote QS! | Registered: Dec 2005
| IP: Logged
|
|
|
unionist
rabble-rouser
Babbler # 11323
|
posted 22 June 2006 09:13 AM
quote: Originally posted by Stephen Gordon: That's not what the theory says; it says that the only way that monetary policy can engineer an increase in employment is by reducing real wages below what they would otherwise have been.
Anyone interested in reading about this phony rightwing economic theory from a progressive point of view should just check out the many writings of CLC chief economist Andrew Jackson on the subject. Or here's a quick intro from Monthly Review magazine: Challenging Right-Wing Think Tanks' Economics-Lite quote: The argument against the minimum wage is really an argument against decent wages for workers -- just like the Fed's theory of inflation and unemployment...The Fed's theory is referred to as NAIRU, the nonaccelerating inflation rate of unemployment. The belief was that whenever unemployment fell below some number, believed to be roughly 6.2%, inflation would take off. But in the early 1990s, under some pressure from the Clinton administration and with no evidence of inflation, the Fed began to restrain itself from raising interest rates. During this period, unemployment fell and there was little or no evidence of inflation. This left the Fed in a quandary: Federal Reserve Bank of San Francisco, Economic Letter 97-35, "NAIRU: Is It Useful for Monetary Policy?" (November 21, 1997).
From: Vote QS! | Registered: Dec 2005
| IP: Logged
|
|
|
|
|
|
|
|
|
|
unionist
rabble-rouser
Babbler # 11323
|
posted 22 June 2006 10:46 AM
quote: Originally posted by Michelle: Ah. Well, let's not emulate the neo-cons then. I'm sure we can come up with better ways of expressing the same thing.
Honestly, Michelle -- "voodoo economics" is not a term used by the neo-cons. It is a definition of neo-con Reaganite supply-side economic theory! It's used against the neo-cons. And it's widely accepted in that usage. Example: Definition of Voodoo Economics quote: Voo´doo economics n. 1. (Politics) an economic hypothesis, proposed by President Ronald Regan, that large cuts in tax rates would so stimulate the economy that the tax revenue on the increases in business and personal income would offset the anticipated tax revenue losses, so that such tax cuts would not increasing the federal budget deficit. Its believers do not consider the actual massive deficit increases subsequent to the 1982-83 tax cut as being caused by the tax cut itself, but by other governmental policies. This hypothesis was graphically illustrated by the Laffer curve.
From: Vote QS! | Registered: Dec 2005
| IP: Logged
|
|
rici
rabble-rouser
Babbler # 2710
|
posted 22 June 2006 10:50 AM
What Stephen Gordon said, in his last post, was: quote: [The theory] says that the only way that monetary policy can engineer an increase in employment is by reducing real wages below what they would otherwise have been.
Let's accept that for the moment. In that case, if full employment is our goal, then we cannot achieve it with monetary policy. That's fine by me. Let's go search out other ways to do it. But in response to N. Beltov: you seem to favour enforced employment to enforced unemployment, and I agree that it is less bad. But is it optimal? In other words, is the best we can come up for people to force them to work in order to survive?
From: Lima, Perú | Registered: Jun 2002
| IP: Logged
|
|
|
Fidel
rabble-rouser
Babbler # 5594
|
posted 22 June 2006 11:33 AM
quote: Originally posted by N.Beltov:
A common euphemism for NAIRU is the "natural" unemployment rate. But there is nothing "natural" about compulsory unemployment for so many thousands, indeed millions, of people. A theory that asserts that a "too low" unemployment rate is associated with (or will "cause") accelerating "inflation" seems useful for another reason; it demonstrates that capitalism necessitates or requires mass unemployment and provides a theoretical substantiation for getting rid of capitalism as a socio-economic system and replacing it with a system that includes full employment.
Just don't talk about nationalising our stuff. It's an unwritten rule of capitalism that public money should never be used to compete with private enterprise. Yeah, ex-nay on the nationalisation end of it. We'd likely see warplanes buzzing Ottawa. We'd find out suddenly that our feds are corrupt(who knew?) or hiding WMD, and regime change would be in order.
From: Viva La Revolución | Registered: Apr 2004
| IP: Logged
|
|
|
|
Stephen Gordon
rabble-rouser
Babbler # 4600
|
posted 22 June 2006 11:55 AM
quote: Originally posted by Fidel: Stephen, what do you make of Pierre Fortin's comments on NAIRU several years ago in Sudbury?. He said that according to NAIRU, for every two percentage points rise in unemployment, there should be a permanent reduction in inflation by one full percentage point. Fortin said that our experience with high unemployment rates in 1980's Canada, France, Italy and Britain called Friedman's NAIRU theory into serious question, and specifically in the years 1984 to 1987 in Canada with unemployment averaging 10 percent. By 1987, real inflation in Canada had been lowered by just 1.3 percent(if i remember correctly).[ 22 June 2006: Message edited by: Fidel ]
He might have been referring to the sacrifice ratio. This is the return trip of the story I told in the OP. If you start at e1 and you want to reduce inflation, there will be a period where employment will be below N*.
From: . | Registered: Oct 2003
| IP: Logged
|
|
|
|
|
|
|
|
Fidel
rabble-rouser
Babbler # 5594
|
posted 22 June 2006 12:58 PM
quote: Originally posted by Stephen Gordon:
He might have been referring to the sacrifice ratio. This is the return trip of the story I told in the OP. If you start at e1 and you want to reduce inflation, there will be a period where employment will be below N*.
quote: And in Europe, this reverse-NAIRU effect became quite pronounced, and leading some economists to question whether unemployment could be pushed up briefly in the interests of inflation control, but brought back down at some point later. What seemed to be happening was the once unemployment was raised, it became stuck at the higher level. This phenomenon, which was dubbed "hysteresis" by two U.S. economists, Olivier Blanchard(MIT) and Lawrence Summers(Harvard), was the exact inverse of what Milton Friedman had described. (excerpts from "The Cult of Impotence", McQuaig, p.53)
Friedman claimed that attempts to reduce unemployment would push inflation higher still. But hysteresis suggested that attempts to reduce inflation require permanently higher unemployment. So Friedman's short-term pain for long term gain is really long-term pain for the sake of protecting the wealth and assets of the wealthy few in Canada and abroad, is it not, Stephen?. [ 22 June 2006: Message edited by: Fidel ]
From: Viva La Revolución | Registered: Apr 2004
| IP: Logged
|
|
|
|
|