INET Conference in Berlin: Day 1 Videos
Now that the Internet connection problems are resolved, at least for the moment, here are the videos from the 1st day of the INET Conference here in Berlin:
Intro: Challenging the Foundations of Economic Thinking
Keynote: Paradigm Lost
Session 1: Which Way Forward? Reflections on Global Turmoil and the Role of Markets, Governments, and Civil Society
Joschka Fischer
William Janeway
Dennis Snower
Joseph Vogl
William White
Discussion and Q&A
Session 2: What Can Economists Know? Rethinking the Foundations of Economic Understanding
Nancy Cartwright
Gerd Gigerenzer
Roger Guesnerie
Roman Frydman
Discussion and Q&A
Roman Frydman adds two essays explaining and elaborating his comments:
1. Non-routine Change, Imperfect Knowledge and Economic Science
2. From Post-War Economic Theory to Imperfect Knowledge Economics At the conference that he convened in 1969 in Philadelphia, Edmund Phelps unveiled a path-breaking approach to formal macroeconomic modeling that based macro-relationships on explicit micro-foundations. These foundations’ distinctive feature was that they accorded market participants’ expectations an autonomous role in economists’ models of aggregate outcomes. The conference contributions, published in what came to be known as “the Phelps micro-foundations volume,” provided radically new accounts of the co-movements of macroeconomic aggregates, notably inflation and unemployment. In discussing his vision for the micro-foundations approach, Phelps underscored the fundamental difficulty in portraying individuals’ expectations in a mathematical model: “isolated and apprehensive, these Pinteresque figures construct expectations of the state of the economy...and maximize relative to that imagined world.”
The papers presented at the Phelps conference did not attempt to formalize market participants’ “imagined world.” Instead, relying on the so-called “adaptive expectations” rule, they modeled the forecasting process as an automatic response to forecast errors: participants were assumed to revise up or down their one-period-ahead forecast of, say, inflation by a fixed proportion of the error between the realization of inflation in the current period and the previous period’s inflation forecast.
Nearly a decade prior to the Phelps conference, John Muth criticized such error-correcting rules. He argued that they assume away an important consideration: in forming expectations, market participants take into account their understanding of the causal process driving the outcomes that they are attempting to forecast. He proposed the rational expectations hypothesis (REH), according to which market participants’ expectations are “essentially the same as the predictions of the relevant economic theory.” In order to implement REH, every economist assumed that the predictions of his own fully predetermined model adequately represented market participants’ forecasting.
Although IKE jettisons the idea that fully predetermined accounts of participants’ expectations and market outcomes are within reach of economic analysis, like REH, it also focuses on participants’ interpretations of the processes driving outcomes and accords fundamental considerations the key role in its representations of forecasting behavior.
Frydman and Phelps and other early critics pointed out REH’s fundamental epistemological flaws. Frydman showed that profit-seeking market participants would not adhere endlessly to a mechanical forecasting strategy, let alone the one implied by an economist’s fully predetermined model. Building on this critique, Frydman and Goldberg argue that, even if viewed as a bold abstraction or approximation, REH is grossly inadequate as a representation of how even minimally rational participants forecast the future. Indeed, strict adherence to REH would be a symptom of obvious irrationality.
The Phelps volume is often credited with pioneering the currently dominant micro-founded approach to macroeconomic analysis. But what has been largely overlooked is that, in contrast to the models in the Phelps volume, their REH-based counterparts presume that market participants’ expectations do not play an autonomous in driving outcomes. As Thomas Sargent, one of the pioneers of the REH approach, succinctly put it: “[I]n rational expectations models, people’s beliefs are among the outcomes of [economists’] theorizing. They are not inputs.” Because they rule out an autonomous role for expectations, Frydman and Phelps argue that, rather than developing the micro-foundations research program, REH-based models derailed it.
As the epistemological flaws of REH became recognized, and as evidence of REH models’ inconsistency with empirical evidence accumulated, some economists returned to modeling expectations as autonomous – as an “input” rather than an “output” of their models. Some of the most influential alternatives that emerged to challenge the hitherto near-universal use of REH were the behavioral-finance models. In contrast to REH models’ focus on fundamental factors, behavioral-finance models singled out psychological factors as the key to understanding market participants’ forecasting. But, although they emphasize “realism” as the hallmark of their approach, behavioral-finance theorists nonetheless believe that REH models represent how rational participants should forecast the future. Consequently, they have interpreted the departures from REH that they have observed in real-world markets as a symptom of participants’ “irrationality” or “bounded rationality.”
Over the last four decades, George Soros has compellingly argued that comprehending financial markets requires recognizing that their participants cope with ever-imperfect knowledge about the processes driving outcomes, and that their interpretations of that process and subsequent outcomes unfold in an inextricably interdependent way that no one can fully foresee. Soros refers to these features of the market, and more broadly, of social change, as fallibility and reflexivity.
The importance of fallibility and reflexivity in markets, such as those for financial assets, raises a key question concerning the role of mathematical models in accounting for outcomes that are driven largely by participants’ expectations. Seeking to answer this question has led Frydman and Goldberg to develop Imperfect Knowledge Economics as a formal approach to macroeconomics and finance whose core premise is that there are inherent limits to what we can know about the processes driving aggregate outcomes.
Recognizing the inherent limits to our knowledge about the processes driving outcomes has enabled Frydman and Goldberg to incorporate into representations of market participants’ forecasting the roles played by both the fundamental considerations on which REH models focus and the psychological and social factors that behavioral finance theorists emphasize, without presuming that market participants are irrational.
Posted by Mark Thoma on Friday, April 13, 2012 at 12:24 AM in Economics |
Permalink
Comments (3)
You can follow this conversation by subscribing to the comment feed for this post.