Refuted/obsolete economic doctrines #7: New Keynesian macroeconomics

I’m writing a review article about Akerlof and Shiller’s new book, Animal Spirits. In doing so, it struck me that I had most of a new entry for my list of refuted economic doctrines, except that the target this time has not been refuted so much as rendered obsolete by events. I’m talking about New Keynesianism an approach to macroeconomics, to which Akerlof and Shiller have made some of the biggest contributions, but which they have now, on my interpretation, repudiated.

Here’s Akerlof and Shiller

The economics of the textbooks seeks to minimise as much as possible departures from pure economic motivation and from rationality. There is a good reason for doing so – and each of us has spent a good portion of his life writing in this tradition. The economics of Adam Smith is well understood. Explanations in terms of small deviations from Smith’s ideal system are thus clear, because they are posed within a framework that is already very well understood. But that does not mean that these small deviations from Smith’s system describe how the economy actually works
Our book marks a break with this tradition. In our view, economic theory should be derived not from the minimal deviations from the system of Adam Smith [needed to provide a plausible account of observed outcomes – JQ] but rather from the deviations that actually do occur and can be observed.

The central theme of new Keynesianism was the need to respond to the demand, from monetarist and new classical critics, for the provision of a microeconomic foundation for Keynesian macroeconomics.

As Akerlof and Shiller note above, the research task was seen as one of identifying minimal deviations from the standard microeconomic assumptions which yield Keynesian macroeconomic conclusions, such as the possibility of significant welfare benefits from macroeconomic stabilization. Akerlof’s ‘menu costs’ arguments, showing that, under imperfect competition, small deviations from rationality generate significant (in welfare terms) price stickiness, are an ideal example of this kind of work.

New Keynesian macroeconomics has been tested by the current global financial and macroeconomic crisis and has, broadly speaking, been found wanting. The analysis of those Keynesians who warned of impending crisis combined an ‘old Keynesian’ analysis of mounting economic imbalances with a Minskyan focus on financial instability.

Similarly, the policy response to the crisis, which now seems to be having some positive effects, has been informed mainly by old-fashioned ‘hydraulic’ Keynesianism, relying on massive economic stimulus to boost demand, combined with large-scale intervention in the financial system. The opponents of Keynesianism have retreated even further into the past, reviving the anti-Keynesian arguments of the 1930s and arguing at length over policy responses to the Great Depression.

There is of course, still a need to explain why wages do not adjust rapidly to clear labour markets in the face of an external financial shock. But, in an environment where the workings of sophisticated financial markets display collective irrationality on a massive scale, there is much less reason to be concerned about the fact that such an explanation must involve deviations from rationality, and seeking to minimise those deviations. Rather, as Akerlof and Shiller suggest, a sensible theory of wage determination should be derived from behavior that actually occurs and can be observed.

To put things more simply (an oversimplification, but this is a blog post, after all), New Keynesianism (as with most other attachments of the word “New” to left/progressive terms in the 1980s and 1990s) was a defensive adjustment to the dominance of free market ideas such as new classical macroeconomics, and to the apparent success of a policy regime in which active fiscal policy played a minor role at most. The New Keynesians sought a theoretical framework that would justify medium-term macroeconomic management based on manipulation of interest rates central banks, and a fiscal policy that allowed automatic stabilisers to work, against advocates of fixed monetary rules and annual balanced budgets.

But now that both the intellectual foundations of post-1970s economic liberalism (most notably the efficient markets hypothesis) and the policy framework that brought us the Great Moderation have collapsed, there is no need for such a defensive stance. The big question for the crisis and after is how to develop and sustain a Keynesian system of macroeconomic management that can deliver outcomes comparable to those of the Bretton Woods era, while avoiding the excesses and imbalances that brought that system to an end in the 1970s.

Note:

I was partly stimulated by this piece from Gregory Clark (hat-tip Brad DeLong). It’s mostly a rant (a very entertaining one) about the state of academic economics, but includes the (only slightly overstated) observation that

The debate about the bank bailout, and the stimulus package, has all revolved around issues that are entirely at the level of Econ 1. What is the multiplier from government spending? Does government spending crowd out private spending? How quickly can you increase government spending? If you got a A in college in Econ 1 you are an expert in this debate: fully an equal of Summers and Geithner.

The bailout debate has also been conducted in terms that would be quite familiar to economists in the 1920s and 1930s. There has essentially been no advance in our knowledge in 80 years.

40 thoughts on “Refuted/obsolete economic doctrines #7: New Keynesian macroeconomics

  1. Thanks John. I’m still trying to figure out what macroeconomics is so this is very helpful. I’ll get Vernon Smith’s book — if it’s anything like his Nobel lecture i’m going to like it a lot. I already avoid Schiller after reading his preachy doom-mongering in Irrational Exuberance.

    I’m still very skeptical about how much experimental evidence can be incorporated into theory. A lot of the experiments aren’t (and can’t be) specific enough to provide any useful information for theory. Experimental results are often used to give a veneer of scientific respectability to political arguments. Shackle was violently opposed to economics becoming an experimental science for this reason.

  2. John Foster @ 19

    If history repeats itself, perhaps in 80 years we’ll have Post-AkerlofShillereans whom remember long forgotton Post-Keynesians in the same way our Post-Keynesians remember the elsewhere overlooked Kalecki.

  3. Joseph: isn’t the whole point of behavioural economics to use laboratory experiments to determine which theories are correct?

    John: what in particular about New Keynesian models do you find to be empirically shaky? Certainly, models with lags do better than models with leads _within the same policy regime_. But if there is a structural break due to a regime shift, the situation is reversed.

    Also Keynes didn’t employ general equilibrium analysis but Hicks famously did in his exposition of Keynesian theory. Friedman aluded to Walrasian general equilibrium concepts in his 1968 Presidential address, but for the most part employed Marshallian partial equilibrium analysis in his work.

    I have some more points to add to my original post. Firstly, I think John Quiggin goes a little too far in equating the efficient market hypothesis with microfoundational macroeconomics and rational expectations. The asset price model that is actually based on microeconomic foundations, Breedens consumption CAPM, has a long standing chequered track record empirically, gives odd values for embedded microeconomic parameters, and as a result is most decidedly _not_ the flagship of the efficient market movement. That title went to Markowitz’s CAPM until about 20 years ago. This model has optimizing elements but lacks rigerous microfoundations. Now it would be some kind of multi-factor arbitrage free asset pricing model (again not really micro grounded).

    Also, one of the big empirical findings of the last 30 years is that, as the Lucas critique suggested, the parameters in econometric regression models typically change when the policy regime changes. Even if rational expectations is wrong this result has to be addressed; it certainly looks like expectation revision of some description based on known or infered characteristics of the new regime is at work here. There is no going back to adaptive or static expectations, at least where evaluating alternatve policies are concerned.

    – Tim

  4. “There is no going back to adaptive or static expectations, at least where evaluating alternatve policies are concerned.”

    I agree with this, and with quite a few points raised above. The Keynesianism that emerges from this crisis must take into account the lessons of the 1960s and 1970s. But the microfoundations will nonetheless involve significant departures from rational optimisation. As Joseph says, that means sacrificing rigour for realism, since there an awful lot of departures to model.

  5. Thanks JohnQ!

    I am speculating here, but I wonder if some of the theoretical basis for the new economics will come from the very recent field of neuroeconomics?

    Big progress is being made in finding the neural substrates of basic consumer theory; relating things like indifference between food bundles to equal firing rates between groups of neurons. This is not such a big deal, because consumer choice under certainty ain’t broke and doesn’t need fixing.

    Choice under uncertainty presents more challenges to the neuro researchers, but there is much overlap between neurofinance and gambling addiction research. For example, making profits on the stockmarket releases pleasure neurotransmitters and in the case of men testosterone; this can induce a state of stupour that impairs decision making – an identical result to the findings for gambling winnings.

    One doesn’t need to go beyond the experimental laboratory to test behavioural theories, but better understanding of the underlying neurological mechanisms may well inspire and inform theorizing.

    Speculating even more here, well established theoretical links between the brains pleasure mechanisms and the dynamic programming techniques so beloved of New Classical/New Keynesian Macroeconomists may yet provide a path back to rigour once realistic theories are established.

    – Tim

  6. […] Refuted/obsolete economic doctrines #7: New Keynesian macroeconomics at John Quiggin: [Here is] a new entry for my list of refuted economic doctrines… the target… has… [been] rendered obsolete by events… New Keynesianism an approach to macroeconomics, to which Akerlof and Shiller have made some of the biggest contributions, but which they have now… repudiated…. [T]he research task was seen as one of identifying minimal deviations from the standard [rational foresight, self-interest, and competiative markets] microeconomic assumptions which yield Keynesian macroeconomic conclusions…. Akerlof’s ‘menu costs’ arguments… are an ideal example of this kind of work. New Keynesian macroeconomics has been tested by the current global financial and macroeconomic crisis and has, broadly speaking, been found wanting. The analysis of those Keynesians who warned of impending crisis combined an ‘old Keynesian’ analysis of mounting economic imbalances with a Minskyan focus on financial instability…. [T]he policy response… has been informed mainly by old-fashioned ‘hydraulic’ Keynesianism… massive economic stimulus… large-scale intervention in the financial system. The opponents of Keynesianism have retreated even further into the past, reviving the anti-Keynesian arguments of the 1930s and arguing at length over policy responses to the Great Depression. […]

  7. JQ notes “There is of course, still a need to explain why wages do not adjust rapidly to clear labour markets in the face of an external financial shock”.

    I think wage reductions go against human incentives right through organisations. Hierarchical structures of management means some are always happier to contribute to mass head choppings off and maintain their own wage from the top to the bottom of organisations – protect thine own wage or salary first!.

    We see this behaviour both within organisations and from unions who are growing ever more irrelevant by focusing on protection of existing benefits for the employed while the ranks of casuals and the unemployed grow around them.

    It is a priority of most people to stand ready to defend their own remuneration if they have any power to do so (and does this willingness to defend increase when the wage is under downward pressure)? I would suggest that human nature makes it difficult for wages to fall.

    We dont behave quite like a pile of surplus wool…

    Tobin also noted as far back as 1960 an increase in the prevalence of administered wages and prices associated with increases in “monopoly power” in industries and the decline in agriculture industries and self employed sectors over time. Tobin claims the decline in agriculture as a sector of flexible prices and wages and as an elastic source of industrial labour contributed to the sluggishness.

    We could suggest the decline of competitive industries has proceeded even further rendering the foundations even shakier.

    If we look at the extraordinary salaries able to paid in the large financial firms it would seem that increasing concentration of industries may play a part. Any asssumption of perfect competition is a poor fit when considering markets for executive salaries and salaries paid at firms like the millionaires factory.

    People band together in groups to protect their incomes whether inside the firm (boards and CEOs or levels of management, or outside the firm eg unions and associations). So wages dont fall enough in recession and rise more (much more in some firms over the recent decade) than they should in times of excess demand.. Its been noted that whilst the underlying assumption of perfect competition is useful for some markets it may not be so for the labour market and this would seem to me to have been recognised by Keynes as it was a theory based on imperfect markets. So I am not quite ready to chuck it in the bin yet.

  8. I have been doing some more web reading on behavioural economics and am a little less inclined to radically go outside the bounds of economic theory as a result.

    Specifically, one finding that crops up is that some deviations from rationality are not universal but depend upon psychological traits like self confidence and anxiety proneness. They may thus be important for economic actors in the general population, but I think it is reasonable to assume that decision makers in big business, funds managers etc are selected to avoid these traits. There are still a lot of deviations from rationality that seem universal and need to be accounted for (eg the Alias paradox). I would be interested in seeing John Quiggin’s hit list in this respect.

    Having said that, I think rational expectations is wrong. Rational expectations does not mean rationality as applied to forming forecasts, as the name suggests. It means that expectations are formed with full knowledge of the structural economic model for the variables concerned, as well as its parameter values. It is a theoretical useful theoretical quick fix that cuts through the gordians knot of expectational issues aluded to in the Samuelson/Solow paper reprinted by Alice. But it also plain wrong.

    Mainstream economics widely ackowledges that precise knowledge of economic parameters is too strong an assumption, particularly with respect to the policy reaction functions used to model policy regimes. The standard solution is to use least regression models to learn these parameters ‘on-line’, yielding an estimate that grows more precise over time.

    But this just doesn’t go far enough. The differering forecasts in the Wall Street Journal poll of forecasters reflect different models – different assumptions about the way the economy works – not different estimates of the parameters of the known to be correct one true model.

    My thinking on this subject was stimulated when I looked at long bond data going back to the seventies on FRED for a post on Brad DeLong’s Blog (I only needed the latest point). I noted the twin peaks in the early and mid 80s. I think the first peak represented the effects of monetary tigtening/fiscal lossening, together with regime uncertainty over and (at the time) absymally low credibility of the Volker monetary targetting regime.

    The second peak might have something to with what the post monetary targeting regime would mean in terms of interest rate policy (monetary targetting was implemented from 1979-82), but I think the big factor at work here was monetarist based inflation forecasting. The latter (including the Newsweek articles of Friedman) forecasted a big resurgance in inflation from 1985 onwards as a result of the surge in monetary growth that occured after monetary targetting ended. When these forecasts proved to be false, bond markets increasingly ignored monetary aggregates, and bond rates fell back again.

    Some economists take this sort of thing on board. Mankiw, Riess and Wolfers did a paper on disagreement amongst inflationary expectations in surveys and published forecasts. Sargeant analysed a central banker who doesn’t know which model to use and employs Bayes’ theorem to combine the results of several models. And Kurz has a model where differences in asset pricing models between investors drive stockmarket dynamics and propagate waves of volatility, with data generated by outcomes occasionally making people change models.

    All of these models (but not the Mankiw, Riess and Wolfers paper, which is data based) are probably too optimization based for John Quiggin’s liking. But I think that they represent a step in the right direction in replacing rational expectations.

    – Tim

  9. Oops:

    For “The second peak might have something to with what the post monetary targeting regime would mean in terms of interest rate policy ” please Read “The second peak might have something to with UNCERTAINTY ABOUT what the post monetary targeting regime would mean in terms of interest rate policy”

  10. “There is of course, still a need to explain why wages do not adjust rapidly to clear labour markets in the face of an external financial shock”.

    How about overlapping annual wage contracts (or in some sectors in the US, multi year indexed wage contracts)? I guess it depends on what you mean by rapidly.

    – Tim

  11. […] May 12, 2009 by isummary * Refuted economic doctrines #1: The efficient markets hypothesis * Refuted economic doctrines #2: The case for privatisation * Refuted economic doctrines #3: The Great Moderation * Refuted economic doctrines #4: individual retirement accounts * Refuted economic doctrines #5: Trickle down * Refuted economic doctrines #6: Central bank independence * Refuted/obsolete economic doctrines #7: New Keynesian macroeconomics […]

  12. lark (#10) wrote: “This reminds me of ‘regulatory capture’. What conditions in the [economic] profession allowed this?”

    Take the captured economists at their word. The only reason anyone does anything is because they are rational actors seeking to maximize their expected future income.

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 )

w

Connecting to %s