Sunday, December 06, 2009

On Animal Spirits and Knee-Jerk Reactions

A couple of weeks ago Robert Shiller published a piece in the New York Times in which he explored the role of mass psychology in generating business cycles, and argued that "the economic recovery that appears under way may be based on little more than self-fulfilling prophecy." This led Mark Thoma to respond as follows:
I find that I have a knee-jerk, negative reaction to explanations based upon mass psychology, sentiment, story-telling, and the like. I have to consciously force myself not to dismiss them. I'm not sure why that is, though it probably has something to do with a feeling that such explanations aren't scientific, and hence have no place in serious academic investigations. That is, prior to the crisis I thought that the real economy drove sentiment, and not the other way around. Sentiment could definitely provide a feedback loop that strengthens negative or positive economic shocks, but psychology was not the prime mover. Thus, sentiment changes that did not have evidence to support them would quickly die out before having much, if any effect.
But this crisis has caused me to reevaluate. I still find the Shiller-type animal spirits, psychology based explanations hard to swallow, but when the foundation supporting your beliefs is called into question (in this case modern macroeconomic models), it's important to open your mind and at least give alternative explanations a chance. That's particularly true when the person pushing the stories has a pretty darn good record of using them to warn of bubbles, as Shiller does. So I'm trying.
This captures my sentiments almost exactly. I'm trying. I too have the greatest respect for Shiller and consider his 1981 paper on stock price (relative to dividend) volatility to be an absolute classic. But I can't help thinking that too much is being asked of behavioral economics at this time, much more than it has the capacity to deliver. In an earlier post on Elinor Ostrom I expressed this view as follows:
Behavioral economics... has been very successful in identifying the value of commitment devices in household savings decisions, and accounting for certain anomalies in asset price behavior. But regularities identified in controlled laboratory experiments with standard subject pools have limited application to environments in which the distribution of behavioral propensities is both endogenous and psychologically rare. This is the case in financial markets, which are subject to selection at a number of levels. Those who enter the profession are unlikely to be psychologically typical, and market conditions determine which behavioral propensities survive and thrive at any point in historical time.
If one is to look beyond economics for metaphors and models, why stop at psychology? For financial market behavior, a more appropriate discipline might be evolutionary ecology. This is not a new idea. Consider, for instance, this recent article in Nature. Or take a look at the chapter on "The Ecology of Markets" in Victor Niederhoffer's extraordinary memoir. Or study Hyman Minsky's financial instability hypothesis (discussed at some length in an earlier post), which depends explicitly on the assumption that aggressive financial practices are rapidly replicated during periods of stable growth, eventually becoming so widespread that systemic stability is put at risk. To my mind this reflects an ecological rather than psychological understanding of financial market behavior.


  1. Yes! Absolutely right. I missed Mark's comment at the time but I felt exactly this way when I read Animal Spirits.

    I do firmly believe in the potential for behavioural economics (perhaps cognitive economics would be a better term) to better explain the macroeconomy at some point. But Shiller and Akerlof are nowhere near providing a model that can do it.

    As an aside, my view is that behavioural phenomena in finance are much more relevant to the investors than the professionals; but even there, there is much more work to be done.

    The ideas you're expressing in this blog seem to be pointing in the right direction, and I hope my own research will make a contribution too. A short outline I wrote earlier this year is on VoxEU.

  2. I am not an economist, but my reading of the behavioralists is that that they explain limits to modeling, and that is their value.

    To say, ok fine, now that you've explained why our old models are bad, please give us new ones ...

    To what degree is that a typical human request for answers? Does that ever imply convenient answers? Cue Taleb.

  3. What Shiller and Akerlof have shown in Animal Spirits is that the structural models of the individualistic "rational" processing that entrepreneurs, workers and consumers are assumed to use in RE and DSGE modeling have little correspondence to how they actually think or how they behave.

    When the assumption that actors have perfect information and computational brilliance generate nonsensical predictions, a less than complete set of information is reverse engineered into the model to get the model to predict behavior consistent with stylized facts. To make the model work the modeler makes arbitrary assumptions such as the Calvo updating is once every X months not 2*X months or 0.33*X months. Without exactly the “right” delay, the model blows up, so the model maker chooses an updating periodicity that generates aggregate behavior similar to the recent past in simulations. Ask yourself this question. Why might it be rational for market participants to update their judgments about market conditions on a fixed timetable? As soon as you ask the question it is clear that reasonable people would increase the frequency of updating when they start suspecting a speculative bubble is forming or a large investment bank may implode.

    But then an event such as the Russian or Lehman Default or the bursting of a DotCom or housing bubble forces the analyst to throw the old model out and build an entirely new one. What good is a theoretical framework that fails to warn us that there is an elephant in the TV room. What good is a theory that must be completely rebuilt when the elephant knocks down a wall holding up the roof.

    Akerlof and Shiller are saying economists will not be able to model and predict the economy we actually have (rather than the idealized RE and DSGE models we simulate in our computers) until we recognize that social norms, beliefs about justice and fairness, regulatory capture, herd behavior, looting and animal spirits influence what it is rational for rational actors to do.

    Their insights are not that economic actors are irrational, it is that the characterization of rationality and of markets found in RE and DSGE models ignore important parts of the human psyche and how we relate to and influence each other. They are calling for a root and branch reconstruction of macroeconomics.

    I am with you Rajiv. Behavioral economics is not the answer. Let’s see how far non linear ecological models can take us.

  4. John, thanks for this thoughtful and informative comment.