Home > Uncategorized > DSGE models in the ‘New Keynesian’ repair shop

DSGE models in the ‘New Keynesian’ repair shop

from Lars Syll

The problem of the DSGE-models (and more generally of rational expectations macroeconomic models) is that they assume extraordinary cognitive capabilities of individual agents. Recent developments in other disciplines including psychology and brain science overwhelmingly document that individual agents struggle with limited cognitive abilities, restricting their capacity to understand the world. As a result, individual agents use small bits of information and simple rules to guide their behavior.

aa-model-train-repair-584x300The fact that the assumption of rational expectations is implausible does not necessarily mean that models using such an assumption cannot be powerful tools in making empirical predictions. The problem, however, is that rational expectations macroeconomic model make systematically wrong predictions, in particular about the speed with which prices adjust. This empirical failure could have led the profession of macroeconomists to drop the model and to look for another one. Instead, macroeconomists decided to stick to the rational expectations model but to load it with a series of ad-hoc repairs that were motivated by a desire to improve its fit. These repair operations most often involved adding lags to the models so as to create sufficient inertia in variables. These repair operations most often involved adding lags to the models so as to create sufficient inertia in variables. These operations were successful in the sense that the fit was significantly improved. In another sense, however, they were failures because the inertia building tricks are really departures from rationality. As a result, the present DSGE-models create a dynamics the largest part of which is the result of the ad-hoc repair operations. These have nothing to do with optimizing behavior and rationality of expectations. In a way it can be said that these ad-hoc repairs introduced heuristics in the model through the back door.

The success of the DSGE-model has much to do with the story it tells about how the macroeconomy functions. This is a story in which rationality of superbly informed and identical agents reigns … We have questioned this story by presenting an alternative one. This is a story in which agents do not understand the model well, and use a trial and error learning strategy to discover its underlying logic …

No individual can ever hope to understand and to process the full complexity of the world in which he lives. That’s why markets are so important. They are institutions that efficiently aggregate the diverse bits of information stored in individual brains …

Paradoxically, the rational expectations revolution that was so much influenced by the Chicago School created a model that, like in the socialist models of the past, assumes an all-knowing individual, who can compute the optimal plans and set the optimal prices. In such a world, markets are indeed not necessary to coordinate the actions of heterogeneous individuals. The representative agent does it all in his mind. In the behavioral model presented here, we go back to the old Hayekian idea that we need markets to aggregate the information that is spread out in tiny little bits in individuals’ brains. It is this aggregation process that creates macroeconomic fluctuations.

Paul De Grauwe

De Grauwe’s critique of the repair shop treatment of DSGE modelling convincingly shows that ‘New Keynesian’ tweaking of DSGE models will not do the job. Why? Basically because they do not acknowledge​ real-world genuine uncertainty, and without this acknowledgement,​ the job will not be properly done.

It also underscores the necessity of being sceptic of the pretences and aspirations of ‘New Keynesian’ macroeconomics. So far it has been impossible to see that it has yielded very much in terms of realist and relevant economic knowledge. And — as if that wasn’t enough — there’s nothing new or Keynesian about it!

counterfeit‘New Keynesianism’ doesn’t have its roots in Keynes. It has its intellectual roots in Paul Samuelson’s ill-founded ‘neoclassical synthesis’ project, whereby he thought he could save the ‘classical’ view of the market economy as a (long run) self-regulating market clearing equilibrium mechanism, by adding some (short run) frictions and rigidities in the form of sticky wages and prices.

But — putting a sticky-price lipstick on the ‘classical’ pig sure will not do. The ‘New Keynesian’ pig is still neither Keynesian nor new.

The rather one-sided emphasis of usefulness and its concomitant instrumentalist justification cannot hide that ‘New Keynesians’ cannot give supportive evidence for their considering it fruitful to analyze macroeconomic structures and events as the aggregated result of optimizing representative actors. After having analyzed some of its ontological and epistemological foundations, yours truly cannot but conclude that ‘New Keynesian’ macroeconomics — on the whole, and not only when it comes to its repaired DSGE models — has not delivered anything else than ‘as if’ unreal and irrelevant models.

The purported strength of New Classical and ‘New Keynesian’ macroeconomics is that they have firm anchorage in preference-based microeconomics, and especially the decisions taken by inter-temporal utility maximizing ‘forward-looking’ individuals.

To some of us, however, this has come at too high a price. The almost quasi-religious insistence that macroeconomics has to have microfoundations – without ever presenting neither ontological nor epistemological justifications for this claim — has put a blind eye to the weakness of the whole enterprise of trying to depict a complex economy based on an all-embracing representative actor equipped with superhuman knowledge, forecasting abilities and forward-looking rational expectations. It is as if these economists want to resurrect the omniscient Walrasian auctioneer in the form of all-knowing representative actors equipped with rational expectations and assumed to somehow know the true structure of our model of the world. A model of the world — as De Grauwe puts it — “that like, in the socialist models of the past, assumes an all-knowing individual, who can compute the optimal plans and set the optimal prices.”​

  1. Prof James Beckman, Germany
    May 8, 2018 at 9:11 am

    Just speak to virtually any economist about what is NOT their specialty to find people struggling with a lack of data & little time spent analyzing that sub-field,

  2. May 8, 2018 at 4:37 pm

    The first and main problem with these models is not “rational expectations”, whatever it means, but that it depicts the choice of ONLY ONE agent – the so called “representative agent” (maximisation of his intertemporal utility function).
    If you agree that this is nonsense, there is a lost of time to “discuss” about the other “unrealistic” asumptions of the DSGE models – and you play the game of their authors, who love this kind of “discussion” (you admit implicitly that their models have a sense,even if they are highlly “unrealistic”).

    • Prof James Beckman, Germany
      May 9, 2018 at 11:58 am

      Right on, Bernard, as economics assumes markets have at least two sides, with buyers & sellers on each. Then there are the supervisors of each of these agents–all with different perceptions, agendas & estimates of the future. So an n-dimensional chess game, with n larger than any person can handle or any programmer can structure for lack of data.

    • Edward K Ross
      May 10, 2018 at 10:36 pm

      As usual Bernard Guerrien separates the chaff from the wheat and focuses on ‘the first and main problem with these models’. Ted

  3. David Harold Chester
    May 9, 2018 at 5:09 pm

    At least 6 agents are needed to represent the complete social system. SSRN 2865571 explains why.

  4. Muhammad Ali Nasir
    May 14, 2018 at 9:29 am

    I found comparison with idea of central planning interesting, I did some work in DSGE modelling in the past and I think you are absolutely right.

  5. Roy Langston
    May 14, 2018 at 3:17 pm

    No one who works with DSGEs can be taken seriously as an empirical scientist, any more than astrologers with their use of detailed astronomical calculations to identify the trines and alignments of the planets. The fact that students who don’t obediently chant this anti-scientific garbage are not even considered for acceptance into graduate study programs is disgraceful.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.