Friday, February 17, 2012

The Potential Output Debate And the Older Natural Rate and NAIRU Debates

So, Mark Thoma has Tim Duy on again, http://economistsview.typepad.com/economistsview/2012/02/fed-watch-again-with-potential-output.html , with Krugman chiming in with his semi-goofy "Duy on Bullard on Duy on Bullard on Tinker to Evers to Chance." I agree with Duy that Bullard understates the various ways that the bubble crash messed up the financial system, but I want to note the link between this debate and an older one that we have heard less of lately. Some of the links are curious. In particular, although he avoids the language of those debates, I see Bullard in effect supporting views of people who disagree with the old conventional consensus, although in this he may have been following Alan Greenspan as well as Jamie Galbraith circa 1996 analytically, even if not in terms of policy.

So, Arthur Okun coined the idea of NAIRU, which arguably was linked to the old textbook Keynesian story of a clearly defined potential output, with AS curves flat out to that level of output and then suddenly going vertical, in contrast to the one described by Keynes in the GT chapter on prices that had bottlenecks setting in well before then and "reducing elasticity" as one approached the "classical" zone where that elasticity went to zero. While I have never seen a coherent argument why NAIRU should coincide with this, when Milton Friedman posed the idea of the natural rate of unemployment, which simply swept the profession, most observers tied the NAIRU level of output to the level of output associated with Friedman's natural rate of unemployment, a level of output the economy supposedly goes to if there are no particular shocks and policy is more or less neutral, presumably an equilibrium level of frictional unemployment, hence only voluntary. Some would go further and link this with the old Wicksellian (picked up by some of the Austrians) natural rate of interest as well holding, a nice across the board general equilibrium level of output that could be identified in some sense with the supposedly clearly defined potential output of the old Keynesian textbook stories.

But then a funny thing happened on the way to Broadway, namely the mid-to-late 90s. As unemployment fell through what many thought was its natural rate and no acceleration of inflation appeared, indeed the opposite happened, Greenspan sent his minions to the basement of the Fed and decided that productivity was improving sufficiently rapidly that we did not need to tighten monetary policy to avoid crashing into NAIRU. While this did not fundamentally upend the concept, it coincided with critiques coming from people such as Galbraith who argued that the concept was profoundly flawed and fundamentally useless. And Greenspan's continued loosening of policy without any inflation happening burnished his image.

Indeed, from the very beginning of the natural rate discussion there had been people such as Phelps and later Summers who annoyingly pointed out the substantial endogeneity of the natural rate to past unemployment, how long spells of unemployment can make it harder for people to operate in the labor market, something that we are hearing again as voices are now being raised to claim that the natural rate has gone back up, although it remains unclear that if output were to rise sufficiently to push unemployment below that rate we would see an acceleration of inflation.

Now, Bullard's position on this seems to have two contradictory parts. On the one hand, while eschewing the language, he seems to recognize the very weakness and fuzziness of the whole natural rate/NAIRU argument. This is an implication of his valid arguments regarding the fuzziness of the concept of potential output. As with Friedman, it is supposed to be the outcome of a general equilibrium of the economy. However, what the general equilibrium is or should be depends on a bunch of things such as the number of effective sectors in the economy and the degree of price stickiness, and so on. It is not well-defined, implying that these other concepts are also not well defined. OTOH, it sort of appears that he is using this fuzziness, along with the legitimate concern that potential output, however measured or defined, will be growing at a lower rate due to the outcome of the bubble crash (for my part due to reduced capital investment, even if that is not his argument), that then we may need to worry about crashing up against it and should avoid doing so just in case it might also turn out to be NAIRU, despite the lack of any apparent inflationary pressure in the economy, not to mention still low levels of employment relative to working age population and substantial measured excess capacity of the capital stock.

For my part, to the extent the problem is indeed ultimately one of insufficient aggregate demand holding down real capital investment and thus the growth of the natural rate of unemployment output/potential output, etc., then the answer is most certainly not to tighten up any time soon on the macro policy levers, unless somehow Bullard is one of those folks who thinks that AD might actually be stimulated by higher interest rates. But I have not seen him claim that and doubt that this is what he is advocating.

8 comments:

Ralph Musgrave said...

Would that improvement in NAIRU in the late 1990s be due to Clinton’s welfare reforms: i.e. making it more difficult to live on welfare?

Bruce Webb said...

Barkley although I agree with your theoretical explication, I find a different explanation in praxis.

For my entire lifetime economic theories that posit NAIRU have at the same time served the material interests of what were known 50 years ago as the 'Coupon Clippers' but now perhaps as the 'Invisible Bond Vigilantes'.

That is while there are theoretical arguments while policies that lead to higher levels of employment and real wage would lead to capital flight, there doesn't seem to be a lot of evidence that this works in practice. In particular violations of NAIRU haven't always led to inflation, and persistent fears that such inflation is in the cards and will inevitably lead to such capital flight (eeeek, the Chinese will liquidate Treasuries!!!!!!) haven't necessarily led to effects visible in actual capital markets. For example the TV told me today with amazement that both France and SPAIN had successful bond offerings YESTERDAY.

Yet in the face of a couple decades of historical data and a persistent Faith in various forms of EMH there is still a demand for monetary and fiscal policy that I the short term serves the interest of the Bond market.

Gosh it is almost like modern capitalism was organized around exploiting market power relations to maximize individual and class interest. Could it really be that capitalists are not committed to Utilitarian Greatest Good outcomes? Who knew??

Bruce Webb said...

Ralph seems to me that it would be a heavy argumentative load to suggest that the productivity improvements and rel wage increases experienced by the broader middle class in Clinton's second term were the result of kicking lazy t-bone eating bucks off of welfare.

(Offensive) snark aside it is difficult to see why a policy that would serve to drive up labor participation (in the form of people seeking jobs via welfare reform) would serve to drive down unemployment and drive up real wage. Seems to me that Econ 101 appeals to 'supply and demand' suggest the effect should be in the opposite direction, more competition at the lowest wage level serving to drag down compensation at immediately higher levels.

(BTW Barkley, the particular 'captcha' style used by this site is ridiculous, not everyone has 18 year old eyes or superlative powers of pattern recognition, sheesh)

Barkley Rosser said...

Ralph,

No, I do not think it had anything to do with the "welfare reform" of the Clinton presidency.

Bruce,

I thought I made it clear that the events of the mid-90s undid NAIRU, even if many continued to argue for it, more recently in the guise of DSGE model equilibria, although, as I pointed out, there has never been a sound argument why hitting up against any actual potential output limit should generate accelerating inflation (I can buy some inflationary pressure, but the accelerating inflation story has always required much more than just that for it to fly, although it gained a lot of ground during the stagflationary 70s).

I would hope that for you and Paine and others that while I think Bullard is right to point to the fuzziness of potential output/natural rate measures, this does not imply that we need a near term tightening of monetary policy to avoid running into a very unlikely NAIRU.

Bruce Webb said...

Barkley what I meant to suggest is that in this case praxis has trumped theory and that this may have explanatory power.

For example a great part of the economically inclined left blogosphere (Duncan, dKos and Hullaballoo posters etc) are gob-smacked that the austerity regime world-wide goes on unchecked by actual evidence validating the theory underlying it. But as some of those same commenters have noted the end result is all good for the bankers.

The NAIRU debate fell out along the same lines, although data and theory were at best in uneven alignment, at least Central Bank policy served the interest of that part of the bond market that were more concerned with bond PRICE as opposed to savers looking for YIELD.

You could sum this philosophy up as "First Do No Harm (to Billionaire's Bond Portfolios)". And you don't have to posit straight out dishonesty and still less evil, instead a lot of conventional economic theory seems designed to care for and feed capital stock. As if national wealth always and everywhere translated to national well-being.

And it is not just NAIRU and Austerity, you see the same dynamic working among Free Trade absolutists. I mean who could quarrel with more wealth worldwide? (Except the 90% who aren't sharing much in those gains).

It is less evil than it is a fundamental blind spot.

Structuredsettlement said...

In the long run, unemployment tends toward a "natural" rate that reflects basic ... The next big inflation occurred with the War of 1812. ..... who would go on to examine and debate what effect a currency devaluation ..... the exact level of potential output (and of the NAIRU) is generally unknown and tends to change over time.sell structured settlement

Ralph Musgrave said...

Bruce, I just suggested that welfare reforms might have reduced NAIRU. As you rightly say, there is no reason to suppose welfare reform will have much effect on productivity, but I didn’t suggest it would.

Bruce Webb said...

Ralph we may be talking past each other because starting from opposite ends of the question.

Clinton's second term was marked by high productivity, relatively high real wage increases, and yet by something unanticipated by NAIRU-lower unemployment and lower inflation. Whether or not welfare reform incentivized higher rates of labor participation it can't explain why a combination of low unemployment and relatively higher real wage didn't drive inflation up in and of themselves. And it seems to me that in classical theory reforms that served to drive up labor participation (which arguably welfare reform did) by increasing labor supply should have driven down labor price. In isolation anyway. Now you might be Abe to rescue monetary policy based on NAIRU if you could make the argument that classical NAIRU built in the assumption of a fixed supply of labor and that forcing more workers into the market and so suppressing wages simply offset the inflation effects of more people working. Theoretically of you could drive average labor compensation right down to subsistence levels you could achieve full employment while not driving inflation on goods actually consumed by capital. (BTW approximately the situation in Early Modern Britain when only the poor-and rich peoples horses-ate oats while wheat and meat were priced out of their diets altogether). But that theoretical out falls to pieces if you have more labor participation AND higher real wage and yet repressed inflation. Clearly there must be a missing piece.

And in retrospect that missing piece is productivity. That is the way you beat goods inflation in the face of more people working at higher real wage is by inflating the goods supply at a faster rate than the combined effects of new employment and increased real wage.

So no your argument did not assume productivity improvements. On the other hand it advanced no mechanism by which greater labor supply via welfare reform didn't result in either lower wage or higher inflation.

Plus we could complicate everything by pointing out that standard measures of unemployment (e.g. U-3) mostly exclude welfare recipients to begin with. That is people who don't have jobs and in the absense of reform wouldn't be looking for work are not technically 'unemployed'. This welfare reform should have driven up U-3 while suppressing real wage. Which if it would have happened might have served to rescue NAIRU. But since by the data it didn't have that actual net effect leaves me scratching my head.