IKE’s Change and Expectations

Roman Frydman and Michael D. Goldberg Change and Expectations in Macroeconomic Models: Recognizing the Limits to Knowability March 23rd, 2012 Prepared for the Institute for New Economic Thinking (INET) Plenary Conference Berlin, April 12-15th, 2012

In modern economies, individuals and companies engage in innovative activities, discovering new ways to use existing physical and human capital, and new technologies in which to invest. The institutional and broader social context within which these activities take place also changes in novel ways. Moreover, market participants search for and occasionally adopt new ways to forecast returns from their activities. Thus, change in capitalist economies is to a significant extent non-routine, for it cannot be adequately represented in advance with mechanical rules and procedures.

Non-routine change alters how market outcomes unfold over time in ways that puts an overarching probabilistic representation of the process driving outcomes out of reach of economic analysis.

[Conventional theory] has obscured our understanding of markets’ role in modern economies, and of how, in many markets, their participants’ forecasting drives outcomes. It has also led to a false dichotomy: rational decision-making based on fundamental considerations versus irrational behavior driven by psychological factors.

IKE [Imperfect Knowledge Economics] holds out the possibility that individual behavior and market outcomes exhibit regularities, which makes economic theory possible, though we cannot expect these regularities to conform to mechanical rules. At best, these regularities can be portrayed with qualitative conditions that begin and cease to be relevant at moments that no one can fully foresee.
Moreover, such qualitative and contingent regularities are context-dependent. In this sense, IKE picks up where Keynes left off.

Although respecting the limits to knowability does entail abandoning the contemporary search for overarching probabilistic accounts of outcomes, … it does not necessarily mark the end of formal economic theory that can be confronted with empirical evidence.

Assuming Away Non-Routine Change

Economists typically regard the market price as the value that balances the total demand and supply … The [properties] of such a relationship … constitute the causal structure of the model.  That much is common to the way in which macroeconomists and finance theorists represent the relationship between individual behavior and aggregate outcomes, regardless of whether they rely on REH, behavioral considerations, or IKE.

[T]o account for time-series data on economic outcomes, an economist will need different causal structures at different points in time.

Rationality Turned Upside Down

 [Conventionally,] individuals … are presumed to adhere steadfastly to a single mechanical forecasting strategy at all times and in all circumstances. Thus, in the context of real-world markets, REH presumes that participants are obviously irrational. When institutional developments occur that alter the process driving market outcomes, they supposedly look the other way, and thus either abjure profit-seeking behavior altogether, or forgo profit opportunities that are in plain sight.

Between Animal Spirits and Fully Predetermined Models

As in their fully predetermined counterparts, the key assumptions that impute empirical content to IKE macroeconomics and finance models are those that characterize change, particularly in how market participants revise their forecasting strategies. Although IKE stops short of fully prespecifying change, it recognizes that economic behavior must display some regularity if formal economic theory is to generate implications that can be falsified by empirical evidence.

[T]he conditions that specify an IKE model’s representations of individual decision making – and thus of the aggregate outcomes that they imply – are not only qualitative and context-dependent, but also contingent.

Non-Standard Use of Probabilistic Formalism

 … Although it does not represent outcomes with one overarching probability distribution, IKE departs from the position of Knight and Keynes and makes non-standard use of probabilistic representations in formalizing the qualitative and contingent regularities that specify its models.

By contrast, because the conditions that they use to characterize change are qualitative, IKE models represent outcomes at every point in time with myriad probability distributions. Nevertheless, the conditions that specify IKE representations constrain all transitions across probability distributions to share one or more qualitative features. These common features, which are embodied in what we call partially predetermined probability distributions, enable economists to model mathematically some aspects of the causal mechanism that underpin individual decision-making and market outcomes. Such partially predetermined probabilistic representations constitute the empirical implications of IKE models.

By not fully predetermining diversity, IKE enables an economist to recognize its importance. Indeed, as Hayek emphasized, the diversity of expectations is crucial to the distinction between resource allocation by an individual or group of individuals – such as central planners – and that by markets in capitalist economies.

Although our model predicts that, under “certain conditions,” an asset price will undergo a sustained movement in one direction, it does not predict when such upswings or downswings will begin or end.


This paper provides a good critique of both rational expectations and behavioural approaches. It makes a minimal change to the mainstream approach in allow forecasting strategies to vary between people and to vary with time. Thus, for example, people with very different life-expectancies may have different strategies, and people’s strategy may change if they get some bad news. This introduce Knightian uncertainty into macro forecasts, and some degrees of freedom in fitting time-series data.

A criticism I have is that while IKE does -in a sense – follow on from Keynes, it is only following on from Keynes as seen by subsequent mainstream economists, overlooking his Treatise on Probability. IKE’s model is of people who think very conventionally about uncertainty, occasionally changing their mind to another conventional view. Keynes’ view, in which he takes a broader view of risk and supposes that others might, seems more general. For example,as presented, IKE depends on the notion that supply and demand balance, but as Keynes points out this assumes that the market is near equilibrium. It is not clear that this was the case through 2007/8.  Further, the predictions of IKE seem unnecessarily vague compared with Keynes’ approach. By Easter 2008 it was clear that Keynes’ conditions for a crash were in place, that one was unlikely as long as the Chinese were spending in the run-up to their Olympics, and that it was unlikely to be delayed long after that while the key players were in denial. This seems usefully more precise than IKE.

See Also

My notes on economics and uncertainty. In particular, von Neumann and Morgenstern for criticisms of the conventional approach that also apply to IKE.


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.

%d bloggers like this: