May 30, 2011
Unemployment and Output Gaps in a New Keynesian DSGE
The output gap is big and negative; the unemployment gap is big and positive. From "Unemployment in an Estimated New Keynesian Model" (ungated version), by Jordi Galí, Frank Smets, and Rafael Wouters:
In this paper we have developed a reformulated version of the Smets-Wouters (2007) framework that embeds the theory of unemployment proposed in Galí (2011a,b). We estimate the resulting model using postwar U.S. data, while treating the unemployment rate as an additional observable variable. This helps overcome the lack of identification of wage markup and labor supply shocks highlighted by Chari, Kehoe and McGrattan (2008) in their criticism of New Keynesian models. In turn, our approach allows us to estimate a "correct" measure of the output gap. In addition, the estimated model can be used to analyze the sources of unemployment fluctuations.
A number of key results emerge from our analysis. First, we show that wage markup shocks play a smaller role in driving output and employment uctuations than previously thought. Secondly, uctuations in our estimated output gap are shown to be the near mirror image of those experienced by the unemployment rate, and to be well approximated by conventional measures of the cyclical component of GDP. Thirdly, demand shocks are the main driver of unemployment fluctuations at business cycle frequencies, but wage markup shocks are shown to be more important at lower frequencies. Finally, our estimates point to an adverse risk-premium shock as the key force behind the initial rise in unemployment during the Great Recession. The important role uncovered for monetary policy and wage markup shocks at a later stage may be interpreted as capturing the likely effects of the zero lower bound on the nominal rate and of downward wage rigidities (as opposed to those of truly exogenous shocks).
First, and regarding the data on which the estimation is based, we use employment rather than hours worked, and rede ne the wage as the wage per worker rather than the wage per hour. We do so since the model focuses on variations in labor at the extensive margin, in a way consistent with the conventional de nition of unemployment. Given that most of the variation in hours worked over the business cycle is due to changes in employment rather than hours per employee, this change does not have major consequences in itself. We also combine two alternative wage measures in the estimation, compensation and earnings, and model their discrepancy explicitly. Second, we generalise the utility function in a way that allows us to parameterize the strength of the wealth effect on labour supply, as shown in Jaimovich and Rebelo (2009). This generalisation yields a better t of the joint behavior of employment and the labor force, as we discuss in detail. Third, for simplicity, we revert to a Dixit-Stiglitz aggregator rather than the Kimball aggregator used in SW (2007).
The paper is extremely interesting (as well as relevant) and definitely well-worth an in-depth reading. I will focus on three points: (1) the estimate of the output gap resulting from this approach is remarkably similar to the CBO's (and dissimilar to statistical methods such as Hodrick-Prescott and Baxter-King band pass filters); (2) while the natural rate of unemployment rises noticeably, by 2010Q4, the cyclical component accounts for the greatest proportion of the increase in unemployment; and (3) the wage markup -- which the authors attribute to wage stickiness -- account for a large proportion of the increase in unemployment.
Source: "Unemployment in an Estimated New Keynesian Model"
Source: "Unemployment in an Estimated New Keynesian Model"
Source: "Unemployment in an Estimated New Keynesian Model"
Risk premium shocks are shocks in the consumption Euler equation, while investment shocks are productivity shocks specific to the investment equation. Monetary policy shocks are deviations from a Taylor rule; the authors interpret the increasing share of unemployment attributable to monetary policy shoks as arising from the zero interest rate bound.
Some Concluding Thoughts
I thought about this set of results when I read Professor Stephen Williamson's critical remarks about the relevance of Keynesian-inspired wage and price stickiness:
... what I do not like in any of the blog/mainstream news posts I have linked to above is the notion that we know exactly what is going on, and exactly what we should be doing about it. There are good reasons to question Keynesian orthodoxy in the current context. Even if we thought that sticky wages and prices were important factors in the recent recession (which I do not), it is hard to believe that the effects of this stickiness would matter 14 quarters after the onset of the recession. Those pushing the Keynesian narrative need to provide us with some more explicit evidence about wage and price stickiness as it relates to recent events.
What this paper demonstrates is that not all New Keynesian DSGE's yield the same results, and that wage stickiness could potentially explain high unemployment.
Posted by Menzie Chinn at May 30, 2011 09:50 PMdigg this | reddit
Bare with the econ undergrad. Can someone expand/reword why are sticky wages and prices important factors here?
Posted by: Jim at May 30, 2011 10:32 PM
Stickiness is important in New Keynesian models because that is how the market does not rapidly clear even if there are no other distortions.
Wages and prices are higher than the market clearing rates for labor and in this instance, houses, so we see lots of houses and workers sitting on the sideline as "excess" inventory (I hate this phrasing as I was out of work for way too long in this recession). Most people are unwilling to drop their reservation wage too quickly or by too much at any one time because most people live in nominal debt worlds --- what they owe is nominal dollars, not a proportion of their income. So workers make the judgement that it is better to wait a while longer in the hope that something better comes along then grabbing the first low wage job possible because that low wage job which may be available still means a massive drop in the worker's standard of living and possible bankruptcy.
Posted by: Dave Anderson at May 31, 2011 04:59 AM
@Jim: What Dave said. I'd only add that it is an empirical 'fact' that many prices and just about all wages really are sticky. And it makes sense when you think about it a little. Many prices are set in contracts and only revised periodically, and workers strongly resists reductions in nominal wages - so much so that firms are reluctant to reduce nominal wages because they fear that workers will leave, shirk, retaliate, etc ...
Posted by: Patrick at May 31, 2011 07:31 AM
Drawing from Econbrowser (Learning about Long Term Unemployment (I)).
The prevalent conclusion was, that most of the unemployment was cyclical in nature as opposed to structural. The structural unemployment would account for a very little percentage of the unemployment ~2%.
Chart 10 of UR gap is consistent with the above concluding comments.Are they data available proving the unemployment trend resolution, that is elasticity of employment to wages in the private sector.
In Europe the figures do not match with the assertion.Indeed,not all new Keynesian DGSE s yield the same results.
Unit labour costs
Unemployment rate (as a % of labour force)
Posted by: ppcm at May 31, 2011 08:12 AM
@Dave and @Jim:
All what you say is well and nice, but still it is not an economic explanation for wage and price stickiness. Why do they exist, IF they do in fact exist? Just because anecdotally some people "...are unwilling to drop their reservation wage too quickly or by too much at any one time because most people live in nominal debt worlds..?" is not an economic explanation. There are no microfoundations for this claim. And yes, this is a rehash of the intellectual battles of the 80s and 90s, which apparently, is not overcome.
Posted by: Manfred at May 31, 2011 08:21 AM
Not to worry proved geniuses are at work.
“Before we restructure any country, we’re going to have to restructure the banking system in Europe.”
Laurence D. Fink
BlackRock CEO, on the European debt crisis
BlackRock’s Fink No. 1 in Wall Street Pay as Stock Sinks 16%
Posted by: ppcm at May 31, 2011 08:37 AM
If wages are sticky, then why include them in the analysis at all?! If the data doesn't depend on them and they are not useful for predicting employment trends then why not make them a dependent variable and see what happens?
Posted by: ed at May 31, 2011 11:09 AM
Keynes would no doubt find hilarious the notion this last episode was primarily caused by wage stickiness.
why dont you explain to your curious questioner-- its the amusing notion you are mainly out of a job because you didnt reduce your wage fast enough?
am guessing the "market clearing wage" for a subprime lender or a securitizer might be nagative?
Keynes would also be either amused or incensed, i am guessing the former, that it would be called "New Keynesianism"
Posted by: AWH at May 31, 2011 11:14 AM
Thank you. Loved the typo of "actuation's", which really should be a word.
I think there's a general misunderstanding about models. They are models in the more general sense, not like the Standard Model which describes subatomic particles. Want to challenge the Standard Model? You have some heavy lifting. Economic models take ideas about aspects of the economy and to oversimplify some come from the supply side and others from the demand side. The oddity, if I may, is I don't hear those who work with demand side models dismissing supply side issues as nonsense but I hear that kind of rhetoric from those who develop supply side models and people who don't have a clue but who buy the political content.
In this context, sticky wages, sticky prices is a model idea that represents a general conception that these provide friction. The idea isn't a perfect fit with reality; it's an aggregate idea whose general effects are evaluated. In this regard, the paper is cool.
I wish people would grasp that their dismissal of some models means they are just plain believing in others, that they are accepting as true a whole bunch of ideas and assumptions which are just as illogical as those they criticize. I think maybe things are hindered by the limitations of language; the word "model" implies something more and tends to deliver less than it really means. These aren't modeling the orbits of the planets, something we can do exactly.
Posted by: Jonathan at May 31, 2011 11:49 AM
The paper uses employment rather than hours worked. The authors claim that this is in line with what actually happens in a recession. Assuming that's right, aren't they also assuming that the same relationship is true during an expansion? I'm not sure that the relationship is symmetric. During an expansion isn't it more likely that firms will lag hiring and rely on increased overtime for existing workers? If so, then doesn't this asymmetry corrupt some of their empirical estimates? Also, employment status in their model is primarily binary...a worker is either employed or unemployed. But all of the utility functions and implied Euler conditions holding down intertemporal budget constraints really (implicitly) assume something like a representative agent and they are continuous. Not sure that's true if you use a state variable like employment status. So are these legitimate assumptions if they looked at unemployment rather than hours worked?
Posted by: 2slugbaits at May 31, 2011 04:40 PM
Today is the two-year anniversary marking the end of the great recession.
# jobs created in the last 2 years: 29,000!
Housing prices still falling!
Consumer confidence plummets!
Gas at $4/gallon!
Govt spending out of control!
Obama 2012 budget = full speed ahead with more spending, more debt...
The stimulus will unleash “a new wave of innovation, activity and construction” and “ignite spending by businesses and consumers.”
Obama, Feb 2009.
ROFCMAO! The Hopie and changie guy, like Keynesian economics, a total fraud, clueless on what ails the economy and how to fix it....
Posted by: Toto at June 1, 2011 09:40 AM
Manfred said: "There are no microfoundations for [wage and price stickiness]."
See, this is what gives economists a bad name. Wage stickiness does not need theoretical foundation; it is an empirically observed fact.
Posted by: Greg at June 1, 2011 03:16 PM
@ Manfred ---
Empirically wages and prices are discrete functions and not continuous functions in a Keynesian model instead of a simple neo-classical model. The basic reason is search costs are high, uncertainty costs are present and changes are costly to do continuously.
That is the basic economic reasoning for stickiness. Throw in a bit of loss-aversion even if the loss is purely nominal instead of real, and inefficient discounting mechanisms, and that produces a whole lot of stickiness.
Posted by: Dave Anderson at June 3, 2011 10:30 AM