Tuesday, February 4, 2014

Will Future Fed Policy Be Driven By Behavioral Macroeconomics?

Arguably it already is, although this has not been commented on particularly by many, but with the accession of Janet Yellen to be Fed Chair, the question becomes more pressing.  Yes, she will be surrounded by many who do not agree with her views on the FOMC and among those attending meetings of it not on it, but as Chair she will have more influence than in the past, and her husband George has declared that there is almost no difference in their views on macroeconomics.  While both of them have been labeled as "Keynesians," and some of their work has influenced the New Keynesian DSGE models now dominant in central bank research shops, increasingly Akerlof has identified himself as being  behavioral macroeconomist, with this even in the title of his Nobel Prize address published in the AER in 2002, "Behavioral Macroeconomics and Macroeconomic Behavior," reinforced a few years later when he argued that norms are the "The Missing Motivation in Macroeconomics" in his 2007 Presidential address to the AEA, these justifying a somewhat more Keynesian view of macroeconomic policy.

Of course, despite his claims about the closeness of their views, we must be careful not to impute too much from his views to hers, given that she will even more so be in an intense environment dealing with difficult situations and other decisionmakers who do not necessarily agree with her views, with the stock market decline on her first day in office and the apparent crises in foreign nations such as Turkey and Argentina putting her on the firing line immediately as she enters office. We should not expect her always to be following his line or views.  Nevertheless, looking particularly at their joint work, but also to some extent at his, may provide some orientation to her deeper views, whatever may come out when decisions must be made.  And, again, I suspect that the answer is that Fed policy already reflects behavioral macro views and arguments more than most realize, even if her moving to the top does not necessarily incease this by much.

Obviously foundational to all of this is his Nobel Prize winning paper on the market for lemons in 1970, formalizing the model of asymmetric information that had been batted around for some time by the likes of Hayek, Stigler,and Vickrey.  This would underpin some of his earliest and most important work with Yellen in such papers as theirs on efficiency wages in 1984 in the AER, and their famous "near rational" model of business cycles from the QJE in 1985.  Both of these would provide the underpinnings for sticky wage-price models, including "Calvo fairy" type gradual adjustment models that many hard core Minnesota type DSGEers ridicule for their lack of rigorous foundation.  However, in the original asymmetric information models the agents are rational, but boundedly so due to their imperfect asymmetric information.  But the key is that they know that they have such imperfectly asymmetric information.  Thus the buyer of a used car will pay less because s/he knows that s/he cannot know all the defects of the car, and employers will willingly pay current workers more than their MP so as to keep them because they know that they know more about the abilities (and lack thereof) of their current workers than they will of any new hires, at least up front.  The use of rules of thumb to deal with informational limits was shown to have first order effects on output with a monetary shock, even when such rules only have second order effects on the firms using them.

However, from a fairly early point, more behavioral aspects crept into their joint work, particularly as they recognized the role of norms and of how agents view each other, particularly in labor market settings.  One of the most important of these insights was in the paper on fairness and unemployment in 1988 in the AER.  This recognized what is going on with the widespread empirical phenomenon of the downward stickiness of nominal wages, with survey evidence gathered by Bewley and others that employers recognize that cutting nominal wages damages morale.  Workers considering what is "fair" implies application of worker norms, which reinforce both that downward stickiness and also the effiiciency wage argument.  Again, while these arguments can be seen feeding into NK DSGE modeling, they also implied further effects and policies.

The latter came about with the later paper by Akerlof, Dickens, and Perry at Brookings in 1996 when Yellen was on the Board of Governors, "The Macroeconomics of Low Inflation."  This paper must be seen as the origin of the nearly global adoption of a 2% target rate of inflation.  The argument derives from that strong tendency to downward stickiness of nominal wages with the microeconomic need for there to be real relative wage adjustments over time. So, these must come about from arguably "excessive" wage increases for some.  In the end it was a matter of judgment that said that a 2% inflation rate was the "Goldilocks" amount, not too much, not too little, that could bring about these efficiency labor market wage adjustments while keeping inflation low.  It is now well known that it was Yellen that year who convinced Alan Greenspan to switch to supporting the 2% inflation target from his previously desirved zero % target, with this later being taken up by Bernanke and many others around the world.  In this regard, behavioral macro is already profoundly influencing not just Fed policy, but macro policy around the world.

I shall conclude by simply listing some themes mentioned by Akerlof in his Nobel address and his AEA presidential address, which we should assume will be at least somewhat supported by the new Fed Chair, even if not necessarily all the time and fully vigorously when too strongly opposed.  So, in his his Nobel address he mostly recounted the points made above about efficiency wages and second order effects.  He exteneded these to question the solidity of the natural rate of unemployment argument, noting that deflation stopped after awhile in the Great Depression.  He also supported Shiller style arguments about asset markets being subject to irrational bubbles, something that may explain her being the first among top Fed policymakers to openly worry about the bubbly nature of the US housing market starting in 2005. Finally, he brought in the matter of undersaving, which reflected another behavioral argument of his from his 1991 Ely lecture on "Procrastination and Obedience.."  This made respectable the long ignored idea from Robert Strotz in 1956 about hyperbolic discounting, which has now become fully entrenched in respectability with such people as David Laibson at Harvard publishing on it in the AER, QJE, and other proper outlets.

Finally, in his presidential address he arguably became more openly behavioral with his emphasis on the role of norms, something that Yellen clearly agreed with him on in some of their joint work on labor markets.  He argued that this "missing motivation" underlay why five supposed "neutralities" of macroeconomics do not hold in reality.  These are the independence of consumption from current income, the independence of investment from fiancial decisions, that inflation stability can only hold at the natural rate of unemployment, that macro policy is ineffective due to rational expectations, and Ricardian equivalence.  Now, several of these were pushed to the side already by the NK DSGE modelers, but some others have continued to be used in such models, with the now-under-attack Euler equation that underlies the first point being an example.  My guess is that by now most of those participating in FOMC meetings have already absorbed that these five neutralities do not hold, even if they do not do so by reasoning from people following norms as Akerlof argues and Yellen may argue.  But, even as I have already said that current Fed policy to some extent already reflects the behavioral macro of Akerlof and Yellen, I suspect that we are likely to do so even more so, with this perhaps even becoming more openly so in the near future.

Barkley Rosser 


Carola Binder said...

Very interesting post. I wonder what the operational implications could be, aside from the 2 percent target.

rosserjb@jmu.edu said...

Actually, Carola, I think things have gone beyond this 2% target. It is out of date. Maybe effectively at ECB and others, 1 or maybe 1 and a half percent are the real de facto targets. It is curious that this nearly 20 year old target is so entrenched, with the person who put it in place now in charge at the top. Which means that she is the one who can modify it; she now has the power, and you, at UCB, are more on the inside of the views there than I am, although, well, they are no longer there... :-).

Noah Smith said...

Great post. I think you should do a series highlighting behavioral macro research papers! If you want to guest-post those on Noahpinion I'd be happy to.

rosserjb@jmu.edu said...

Thanks, Noah, and I appreciate the offer. Probably not going to do it on a regular basis. There is a reason that I only have one foot in the econoblogosphere, and that continues to be there, indeed, at the moment with several commitments I have made, I am even less able to spend as much time as I might, and if I were to do so in a serious way, well, my wife and coauthor would kill me, or at least disable me in a way that I could only work on approved (in some cases, joint) projects, but not this.

However, maybe I shall try to follow up in a limited way, and you can share when I do, Noah.

Myrtle Blackwood said...

"...the widespread empirical phenomenon of the downward stickiness of nominal wages, with survey evidence gathered by Bewley and others that employers recognize that cutting nominal wages damages morale. Workers considering what is "fair" implies application of worker norms..."

Any research on the 'downward stickiness' of household bills for food, water, residential taxes also ceo pay, medical bills...?

Brenda Rosser