Next Challenge for Rational Expectations Hypothesis

KOBAYASHI Keiichiro Faculty Fellow, RIETI

This article focuses on academic theory, away from the political debate on deflation and expectations. An analogy using physics is used to examine how economics has considered the formation of expectations in discussing future possibilities. It is a well-known fact that economics has imitated physics in advancing its theories.

♦  ♦  ♦

Both physics and economics have experienced revolutionary transitions in the 20th century. Physics saw the emergence of quantum mechanics in the early 20th century, while modern macroeconomics, based on the rational expectations hypothesis, became established in the late 20th century. These two developments share characteristic as a change of academic framework, i.e., starting to observe a system from the inside, rather than the outside.

In the world of physics, analytical mechanics (Newtonian mechanics), developed in the 19th century, and Einstein's theory of relativity, which emerged in the early 20th century, are called classical physics. Quantum mechanics is a completely different academic discipline from classical physics. Classical physics is about observing a physical system from the outside, while quantum mechanics observes a physical system from the inside.

In classical physics, observers do not affect a physical system as they remain on the outside. However, in reality, observers are on the inside of a physical system which they are observing. Quantum mechanics accurately describes this fact. In quantum mechanics, observers are on the inside of a physical system and therefore affect the observed system by engaging in the act of observation. This causes unexplained quantum-mechanical phenomena (This is one interpretation that emphasizes the issue of observation).

In the world of economics, classical economics and subsequent variations up to Keynesian economics can be grouped together, as opposed to modern macroeconomics, which was established in the 1970s onwards.

Up until Keynesian economics, observers (economists) were on the outside of an economic system, and assumed to be separate from those on the inside (market participants). Market participants do not observe and recognize the economic system they belong to as a whole. Instead, they merely respond to government policies.

However, in modern macroeconomics, market participants on the inside of an economic system are considered to be observers themselves. University of Chicago Professor Emeritus Robert Lucas pointed out that market participants must be observing and recognizing the entire economic system when making a decision on their actions.

In his 1976 paper, Lucas raised the question about the effectiveness of economic policies, arguing that when a government changes its policy, individuals in an economic system (observers) change their expectation about the overall economy, and therefore change their response to the policy. This changes the response of the entire economic system, which is an aggregate of these individuals, thereby diminishing the effect of the policy as initially intended by the government. This Lucas critique takes the same logical framework as that of quantum mechanics, which argues that observers affect the overall system as they are inside of a physical system.

♦  ♦  ♦

Since modern economics perceive individuals in an economic system as observers, the formation of people's expectations takes the following recursive structure:

The government implements a policy with expectation about how the market responds (Expectation G). Market participants respond to the policy with expectation about what outcome the government policy produces (Expectation M). Since the market response depends on market participants' expectation (Expectation M), the government's expectation about how the market responds (Expectation G) depends on Expectation M. At the same time, since the government's policy decision depends on the government's expectation (Expectation G), the market's expectation about what outcome the government policy produces (Expectation M) depends on Expectation G.

In other words, Expectation M depends on Expectation G, and Expectation G depends on Expectation M. People's Expectation M is therefore determined by and dependent on Expectation M itself (via the government's Expectation G). This characteristic of expectation being determined by itself is called "recursiveness."

Recursiveness is at the heart of modern macroeconomics. Researchers are keenly aware of this concept. This is why the most prominent textbook about modern macroeconomics, written by New York University Professors Lars Ljungqvist and Thomas J. Sargent, is titled Recursive Macroeconomic Theory.

The recursive nature of people's expectation being dependent on the expectation itself is the essence of an economic system. Recursiveness normally prevents any prediction about what is going to happen, and denies any meaningful policy analysis. This is why economists have decided to introduce a very powerful hypothesis (rational expectations hypothesis) that since people make predictions completely rationally, the basic form of expectation is completed from the beginning and would not change with time.

Assuming rational expectations define single expectations and facilitate policy analysis. However, in a real-life economic system, expectations might be recursive but not necessarily rational. Rational expectations are a very special example of recursive expectations, and do not necessarily reflect reality.

Numerous adjustments have been made to bring rational expectations hypothesis closer to reality. These include sunspot theory, high-order expectations and global game, as well as robustness theory of model selection. Yet, none of these adjustments produce results that are very different from those of rational expectations hypothesis. For this reason, the use of rational expectations hypothesis continues to be the mainstream approach in policy analysis.

♦  ♦  ♦

Compared to quantum mechanics, today's economics seems to lack a key characteristic, i.e., direct interactions between observers (people) and observation subject (market).

Werner Heisenberg's uncertainty principle is the essence of quantum mechanics. Observation in quantum mechanics refers to observers firing a photon (light particle) at an observation target, and catching the bounced photon with an observation device. Naturally, the photon's target changes its position and momentum. The act of observation cannot confirm the target's position and momentum at the same time. This is the uncertainty principle.

Meanwhile, unlike quantum mechanics, the economic theory of competitive markets does not involve direct interactions between observers (people) and observation targets (market). The theory of competitive markets hypothesizes that both companies and people determine their demand and supply by taking market price as given. This excludes the possibility that single person's action could directly affect market price. The grand economic premise that quantity is determined by taking price as given is functioning as the framework of completely excluding direct interactions between observers and observation targets.

The notion that both companies and people determine quantity by taking market price as given is a theoretical hypothesis, and does not reflect the real-life economy. In order to truly understand the real-life mechanism of expectations formation, and incite expectation to come out of deflation, it might be necessary to find the economics version of the uncertainty principle.

To this end, we might need a new branch of economics where people determine the probabilistic distribution of price and quantity, away from the premise that people determine quantity by taking the price as given. This is a challenging task for the current generation of economists alone, and it is indispensable to collaborate with explorers of the next generation.

>> Original text in Japanese

* Translated by RIETI.

February 20, 2017 Nihon Keizai Shimbun

April 4, 2017

Article(s) by this author