Behavioral Economics

Table of Contents

Introduction

Rational Choice Theory

Homo Economicus

What is Behavioral Economics?

Homo Economicus as a Normative Theory?

Application of Behavioral Economics to Public Policy

Critiques

Introduction

Behavioral economics is best understood by first considering the theories of human decision-making it seeks to challenge. The “Homo Economicus” theory of human behavior makes predictions about both how individuals will make choices under risk and how individuals will behave in strategic interactions (defined below). This theory of human behavior is based on the axioms of a more general theory of human behavior called rational choice theory. However, experimental findings by behavioral economists have demonstrated that economic actors do not always make choices under risk in accordance with the predictions of the Homo Economicus model. Moreover, research suggests that economic actors do not always behave in the manner predicted by the Homo Economicus model when engaging strategic interactions. Further still, behavioral economists have found that individuals do not always conform to the axioms of rational choice theory, striking at the foundations of the Homo Economicus theory of human behavior.

Rational Choice Theory

Rational choice theory, in essence, imposes a set of restrictions on how an individual makes a choice among a set of alternatives. It says, “If your individual preferences, which lead you to make a particular choice, satisfy the following criteria, then that choice is a rational one.”

Rational choice theory begins with the assumption that people have fixed and stable preferences over alternatives (Levin & Milgrom, 2004). That is, if a person is faced with a choice between an apple and an orange, and prefers an apple today, then he must also prefer the apple to the orange tomorrow. 

Rational choice theory goes on to impose two conditions on individuals’ preferences amongst alternatives. The first condition is that individuals’ preferences must be complete (Levin & Milgrom, 2004). An individual’s preferences amongst a set of alternatives X are complete if for any pair of choices xi,xj ∈ X, he considers xi to be at least as good as xj, or he considers xj to be at least as good as xi, or both. In other words, an individual’s preferences amongst a set of alternatives are complete if he has a preference amongst any two alternatives with which he is faced. He is allowed to be indifferent between the two alternatives, but he must always be able to state a preference. So, for example, if faced with a choice among apples and bananas, the individual must be able to say that he considers apples to be at least as good bananas or he considers bananas to be at least as good apples or he considers both of them to be at least as good as each other. But he may not say that he does not know which he prefers. 

Rational choice theory also requires an individual’s preferences to be transitive (Levin & Milgrom, 2004). An individual’s preferences among a set of alternatives X are transitive if, whenever he considers xi to be at least as good as xj and he considers xj to be at least as good as xz, then he considers xi to be at least as good as xz. For example, if he considers apples to be at least as good bananas and bananas to be at least as good oranges, then rational choice theory requires that he must consider apples to be at least as good oranges. This condition implies that preferences cannot cycle. That is, an individual cannot prefer apples to bananas, and bananas to oranges, and oranges to apples. 

Building on the assumptions that individuals’ preferences are stable, complete, and transitive, rational choice theory states that given a set of alternatives, a decision-maker will choose the alternative that is at least as good as all the other alternatives available (Levin & Milgrom, 2004). That is, given that an individual is capable of stating his preference amongst alternatives, that his preferences do not cycle, and that his preferences do not change over time, it must be the case that, when faced with a set of alternatives, an individual will choose the one he most prefers. For example, if I am able to state my preference for apples over bananas, and bananas over oranges (i.e. completeness), and apples over oranges (i.e. transitivity), and these preferences never change (i.e. stable preferences), then it must be the case that when I’m faced with a choice between apples, bananas, and oranges, I will always choose apples.

Rational choice theory assigns a numerical ranking to each possible choice (Levin & Milgrom, 2004). This numerical ranking represents the “utility” an individual derives from making a particular choice (Levin & Milgrom, 2004). That is, the numerical ranking represents the benefit one receives from making a particular choice. Once utilities are assigned to the various choices, picking the preferred choice simply amounts to picking the choice with the highest utility (Levin & Milgrom, 2004).

Homo Economicus

A large intellectual edifice has been built upon the foundation of rational choice theory. Economics draws heavily on rational choice theory. Moreover, game theory is built on a rational choice theory foundation. And, game theory itself is used as a tool in economic analysis. This reliance of economics upon rational choice theory foundations has led to a theory of human behavior that is pejoratively called “Homo Economicus,” or “economic man.”

In going about his daily life, the Homo Economicus is faced with a variety of circumstances in which he has to make a decision under risk. According to the rational choice theory-based economic literature, the Homo Economicus uses a tool called expected utility maximization in order to decide how to make choices under risk. Another situation that Homo Economicus often faces which affects his economic well-being is what are called strategic interactions. According to the rational choice theory-based economic literature (drawing on ideas from game theory) the Homo Economicus will perform the same utility maximization procedure as when he is facing a decision under risk. Moreover, according to this literature, because every individual will behave in this way during strategic interaction, the outcomes of strategic interactions are predictable and result in what are called Nash equilibria. Let’s address each of choice under risk and decision-making during strategic interaction in turn.

In going about their daily lives, individuals often face a situation in which they have to choose amongst a set of alternate gambles (i.e. prospects) under risk. How do individuals choose among these prospects? According to the Homo Economicus model, individuals will compute the expected utility of each alternative by performing the following computation:

EU(A)=∑𝑜∈𝑂𝑃𝐴 (𝑜)𝑈(o)

where O is the set of outcomes

PA(o) is the probability of outcome o 

U(o) is the utility of outcome o

A is an alternative (a possible choice)

The Homo Economicus will then choose the alternative with the highest expected utility. That is, when faced with a choice under risk, the Homo Economicus will maximize his expected utility.

Another situation Homo Economicus often faces which affects his economic well-being is what are called strategic interactions. Strategic interactions are situations in which two economic actors are each faced with having to make a choice amongst alternative actions and the payoff to each actor depends not only on the action chosen by that actor, but also on the action chosen by the other economic actor. An example of a strategic interaction would be negotiations between two individuals who are trying to agree on the terms of a contract. In this case, the payoff to both individuals will depend both on how they bargain in the negotiation and on how the other individual bargains. How do individuals decide which action to take when they are engaged in a strategic interaction? According to the Homo Economicus model of human behavior, the results of such strategic interactions will be what are called Nash equilibria. A Nash equilibrium is the action profile at which no player can make himself better off by unilaterally choosing a different action (Osborne, 2004). In simple terms, each player chooses his action according the model of rational choice, given his belief about what action the other player will take (Osborne, 2004).

What is Behavioral Economics?

Behavioral economics is a response to this Homo Economicus model of human behavior. It is an attempt to increase the realism of the psychological assumptions made by economists (Camerer, Loewenstein, & Rabin, 2004; Ch 1). Note that behavioral economics does not challenge Homo Economicus as a normative theory (Camerer, Loewenstein, & Rabin, 2004; Ch 1). Rather, it only challenges Homo Economicus as a descriptive theory. That is, behavioral economics claims that economic actors do not act in accordance with the predictions of the Homo Economicus model, but it believes that they should. Behavioral economics’ claim that individuals do not behave in accordance with the Homo Economicus theory comes from numerous experimental studies. 

For example, behavioral economics research has shown that people do not always maximize expected utility when making decisions under risk. Consider Kahneman & Tversky’s (1979) demonstrations based on the responses of students and university faculty to hypothetical choice problems (i.e. subjects were not given the money). The outcomes refer to Israeli currency. To appreciate the significance of the amount involved, note that the median net monthly income for a family was about 3,000 Israeli pounds at the time the experiment was conducted. Consider the following pair of problems:

Problem 1: Choose between [N=72]

A: 2,500 w/ prob .33              B: 2,400 w/ certainty 

     2,400 w/ prob .66

      0 w/ prob .01

Of 72 respondents, 82% choose gamble B. But EUA = (.33)(2500)+(.66)(2400)+(.01)(0)=2409. On the other hand, EUB = 2400. So the expected utility of gamble A exceeds that of gamble B. Yet most participants preferred gamble B. So participants were not maximizing expected utility; they prefer 2400 with certainty to a risky lottery with EU 2409. The same 72 participants were next presented the following problem.

Problem 2: Choose between [N=72]:

C: 2500 w/ prob .33            D: 2400 w/ prob .34 

      0 w/ prob .67 

   0 w/ prob .66

Here, 82% of respondents choose gamble C. The EUC = (.33)(2500)+(.67)(0)= 825. The EUD = (.34)(2400)+(.66)(0)= 816. So here, participants are maximizing EU. This suggests that, when neither option is certain, risk-taking behavior increases.

These results suggest that people overweight outcomes that are certain relative to outcomes that are only probable. In the first choice, the majority prefers 2,400 with certainty to a risky lottery with an expected utility of 2,409. Kahneman & Tversky (1979) label this the certainty effect. In this case, the majority is risk-averse. In the second choice, when faced with two risky gambles, the majority is risk-seeking and so prefers the one with the higher expected utility. This does not conform to the Homo Economicus theory’s prediction that individuals always maximize expected utility when making a choice under risk.

Furthermore, there is evidence to suggest that economic agents do not conform to the game theory-based predictions that Homo Economicus makes about how people will behave in strategic interactions. For example, studies have shown that economic agents do not play to a Nash equilibrium in a particular strategic interaction situation called the ultimatum game. The ultimatum game is played as follows. The experimenter gives a Proposer a sum of money, $x, to divide between himself and a Responder. The Proposer states a proposed allocation of money, $p, to a Responder, from the sum of $x. If the Responder accepts the offer, the Proposer gets $x- $p and the Responder gets $p. If the Responder rejects the offer, no one gets anything. The game is only played once. The Homo Economicus model predicts that the agents play to maximize their payoffs. Thus, it predicts that the Responder will accept any offer greater than $0. Knowing this, the Proposer should take almost all the money. 

For example, suppose the game is played for $10 and the Proposer may propose in 1 cent increments. Then the Nash equilibrium is that he will propose to give 1 cent and keep $9.99. The Responder will accept the offer. If the Proposer chooses to give more than 1 cent, he is made worse off. If the Responder chooses to reject the offer he gets nothing and so is worse off. Thus, no player has a financial incentive to make a different choice.

However, Cameron’s (1999) study offers evidence that economic actors do not play to a Nash equilibrium in the ultimatum game. This study was conducted in Indonesia in 1994. At the time, per capita GDP in Indonesia was only $670. This made it possible to use very large stakes: the largest stakes used were approximately 3 times the average monthly expenditure of participants. Cameron (1999) found that Proposers did not propose to take as much as possible for themselves. Moreover, Responders would often reject offers that were small, despite that fact that rejecting an offer meant the Responder got nothing. Behavioral economists have conjectured that Responders reject small offers because they feel they are being treated unfairly by the Proposer. This is a phenomenon that the Homo Economicus model does not capture. Thus, while the Homo Economicus model predicts that individuals will play to a Nash equilibria in strategic interactions, experiments have provided evidence that this is not always the case.

Beyond finding evidence that economic actors are not always utility maximizing when making decisions under risk and do not always reach a Nash equilibrium in strategic interactions, behavioral economists have struck at the very foundation of the Homo Economicus model: they have found that individuals do not always conform to the axioms of rational choice theory. Consider the findings of Kahneman & Tversky (1981). They posed the following problem to 152 university students:

Imagine that the U.S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume the exact scientific estimate of the consequences of the program are as follows:

If program A is adopted, 200 people will be saved.

If program B is adopted there is a 1/3 probability that 600 people will be saved, and a 2/3 probability that no people will be saved

Which of the two programs would you prefer?

Note that the expected number of lives saved in both programs is 200. Of the 152 subjects, 72% of people preferred program A, while 28% preferred program B. Thus, the majority choice in this problem is risk-averse: the prospect of saving 200 lives with certainty is more attractive than the risky prospect with equal expected value of saving 600 lives with 1/3 probability. 

A second group of 155 students was given the same cover story as the problem above with a different formulation of the alternative programs, as follows:

If program C is adopted 400 will people die.

If program D is adopted there is a 1/3 probability that nobody will die and a 2/3 probability that 600 people will die.

Which of the two programs do you prefer?

Note that, here, the expected number of lives saved in both programs is 200, just as it was above. In this formulation of the problem, 22% of respondents preferred program C while 78% of respondents preferred program D. In this formulation, a majority of subjects are risk-taking: the certain death of 400 people is less acceptable than the 2/3 chance that 600 people will die.

So, for the majority of people, program A is preferred to program B and program D is preferred to program C. But programs A and C are identical! In both programs, 200 people will be saved with certainty. The only difference is that program A is framed in terms of the number of lives saved, while program C is framed in terms of the number of lives lost. And programs B and D are identical! In both programs, there is a 1/3 probability that no one will die and a 2/3 probability that 600 people will die. The only difference between programs B and D is that program B is framed in terms of the number of lives saved, while program D is framed in terms of the number of lives lost. So, since programs A and C are equivalent and programs B and D are equivalent, in the first formulation the majority prefers program A to B, while in the second formulation, the majority prefers program B to A. In other words, majority preferences are not stable, violating an axiom of rational choice theory. And if individuals’ preferences do not conform to the axioms of rational choice theory, then the “Homo Economicus” must be invalidated as a descriptive theory because it is based on a rational choice theory foundation.

In sum, behavioral economics has suggested that economic actors do not always make choices under risk in accordance with the predictions of the Homo Economicus model. Moreover, experimental findings by behavioral economists suggest that economic actors do not behave in the manner predicted by the Homo Economicus model when engaging strategic interactions. Further still, experimental findings by behavioral economists suggest that individuals do not always conform to the axioms of rational choice theory, impugning the very foundations of the Homo Economicus theory of human behavior. Thus, behavioral economics has presented a challenge to Homo Economicus as a descriptive theory of human behavior. 

Homo Economicus as a Normative Theory?

Behavioral economics has presented a challenge to Homo Economicus as a descriptive theory. However, it does not contest the validity of Homo Economicus as a normative theory. That is, while it argues that Homo Economicus does not accurately describe how individuals actually make decisions, a rational individual would adhere to its dictates: humans do not behave as the model describes, but they should.  

However, Berg & Gigerenzer (2010, pg.145-149) have argued that there is nothing to support Homo Economicus even as a normative theory. They argue that there is no evidence to suggest that deviations from this supposedly normative model result in people being worse off: there are no studies which show that deviators from rational choice theory consistently earn less money, live shorter lives, or are less happy (Berg & Gigerenzer, 2010, pg.149). And if there is no evidence to suggest that deviators from rational choice theory are worse off, then, according to Berg & Gigerenzer, there is no justification for the Homo Economicus theory as a normative theory of economic behavior.

Application of Behavioral Economics to Public Policy

Some academics have begun applying behavioral economics to public policy. Indeed Cass Sunstein and Richard Thaler have developed a new political philosophy based on insights from behavioral economics (Thaler & Sunstein, 2003; Sunstein & Thaler, 2003). Specifically, they develop a new normative account for how the state should interact with its citizenry, which they call libertarian paternalism. They claim that, in certain domains, a state enacting libertarian paternalist policies can “nudge” its citizens to make better decisions without restricting the autonomy of its citizenry to make free and independent choices. You can learn about libertarian paternalism here. 

Critiques

Written By: Aiden Singh March 27, 2020

Sources

Amos Tversky & Daniel Kahneman. The Framing of Decisions and the Psychology of Choice. Science. Vol 211, Issue 4481. January 1981.

Cass Sunstein & Richard Thaler. Libertarian Paternalism Is Not an Oxymoron. The University of Chicago Law Review. Vol. 70, No. 4 (Autumn, 2003), pp. 1159-1202.

Colin Camerer & George Loewenstein. Behavioral Economics: Past, Present, Future. in Advances in Behavioral Economics. Edited by Camerer, Loewenstein, and Rabin. Princeton University Press. 2004. 

Daniel Kahneman & Amos Tversky. Prospect Theory: An Analysis of Decision Under Risk. Econometrica. Vol. 47, No. 2 (Mar., 1979), pp. 263-292

Jonathan Levin & Paul Milgrom. Introduction to Choice Theory. September 2004. 

Lisa Cameron. Raising the Stakes in the Ultimatum Game: Experimental Evidence from Indonesia. Economic Inquiry. Vol. 37, Issue 1. January 1999. pp. 47-59.

Martin Osborne. An Introduction to Game Theory. Oxford University Press. 2004.

Nathan Berg & Gerd Gigerenzer. As-If Behavioral Economics: Neoclassical Economics in Disguise? History of Economic Ideas. XVIII/2010/1.

Richard Thaler & Cass Sunstein. Libertarian Paternalism. The American Economic Review. Vol 93. No. 2. May 2003.