Another way in which irrational behaviour dominates practicality is when consumers measure costs as proportions instead the absolute. This is best described by Robert Frank in “Microeconomics and Behaviour”, in which he describes two situations. The first being that a consumer could buy a clock radio for $20 at the shop near their home, or drive to the supermarket and get it for $10. In this situation, the rational consumer would choose to drive to the supermarket as they make $10 savings. However, if this consumer now wants a big television set, priced $1010 at the nearby shop and $1000 at the supermarket. The rational consumer should again choose to go to the supermarket as they again make a saving of $10. However, because in the second situation, the consumer only makes a saving of 1% rather than the 50% saving in the first situation, they may decide to get the television from the nearby shop. It is by doing that, that they are not behaving irrationally.
Even some consumers presented with, in principle, the exact same situation, behave very differently. Amos Tsversky and Daniel Kahneman experimented with this idea in their paper “The Framing of Decision and the Psychology of Choice”. They did this by telling one group of people to imagine that they were going to see a play with tickets priced at $10. Upon arrival, they discover that they have lost a $10 note. A second group were then told that they had previously bought a ticket for $10, and upon arrival discover that they have lost the ticket. Both these groups were then asked whether or not they would proceed to see the show. In principle, these problems are the same, whereby the consumer is effectively losing $10, or $10 worth of goods. However, their results show that in the first case, consumer loses $10 note, 88% said that they would continue to buy a new ticket. Whereas in the second case, where they lose the original ticket, only 46% said that would buy another ticket and proceed to see the play. In this paper they then go on to propose that consumers have mental accounts which affect their behaviour in scenarios like the one depicted above.
Therefore a new model should incorporate consumer’s psychological accounting, and measuring costs as absolutes rather than a proportionate, and the utility of a certain option, with respect to sunk costs.
To get to this new model, an understanding of the utility function is required. Below is a diagram of the utility function, by where utility is drawn with respect to wealth:
As shown above, if a consumer is originally at equilibrium M0 and U0, and then if their wealth increases to M1, their utility is increased to U1. Notice how, because of the angle of the slope, that it takes a more significant increase in wealth to increase utility a little bit. Whereas if the original equilibrium was more to the left of M0 and U0, then there would be more of an increase in utility.
The utility function helps explain the new model, which was proposed by Kahneman and Tsversky. It is called the Value Function, and is shown below:
(Robert H. Frank, 2008, p240)
The Value function is a descriptive model of the “regularities in the ways people actually seem to make choices” (Robert H. Frank, 2008, p.240) the diagram above demonstrates the following example. The scenario is that you have just returned home from holiday, and have received an unexpected gift certificate for £100. At the same time, you also receive an invoice from the hotel you were staying at, for a water pipe that burst whilst staying at the hotel totalling £80. The rational consumer would still be happy, due to the fact that they are still effectively £20 better off. However, in practice, many consumers would be angrier by the fact that they are losing £80. The basic principle of this is that people value a loss more than a gain, this is demonstrated in the diagram above with a steeper curve when the value is a loss. A resulting factor of this is that the area demonstrating the loss is clearly larger than the area demonstrating the gain. It does not seem to be irrational to feel a loss more heavily than a gain, as Herbert Simons describes in his writings that we are “saticficers” not maximisers. However what does seem irrational is treating each scenario independently. As for the example above, it is irrational to consider both the £100 increase and the £80 loss separately, instead of realising that the end result is actually a £20 increase in total welfare.
Another theory on irrationality was put forward by Von Neumann Morgenstern. It demonstrates how consumers are to act rationally when under a degree of uncertainty, or when there is an element of risk involved, a very good example would be gambling. His theory uses statistics to describe what some consumers would choose given situations where there are elements of risk. The diagram that represents his theory is shown below:
The chord that runs from A to B is the expected utility of the scenario. For instance if a consumer was presented with a 50:50 outcome of winning £100 or nothing, the expected utility would be half way along the chord AB. Therefore any points lying right of B, and left of A, are scenarios where the expected utility will be less than the outcome, and as a result, rational consumers would choose against the taking part in that particular scenario.
However, work that was undertaken by Kahneman and Tversky after this work was published described how this model does not describe how rational consumers actually behave, only how they behave when there are elements of risk.
Another way to consider consumer preferences is to consider cardinal utility and the analysis of consumer surplus. In order to construct the diagram, a couple of assumptions have to be made. The first of which is that the utility of consuming one good, call it good X, is independent of consuming another good, call it good Y. In other words, good X and Y are not complements or substitutes. The second assumption is that the utility of consuming good X is constant which is represented by a vertical line on the indifference map. The final assumption is that the consumption of good Y has a diminishing marginally rate of utility, represented by a downward sloping utility curve with respect to good Y. Using these assumption the “Marshallian” Indifference Map can be constructed. As shown below.
Since the utility of consuming good X is constant, we can measure the increase in welfare by an increase in good Y. For instance, if a consumer is originally on I0, and moves to I1, then the increase in welfare is the I1 - I0. This is also true for I1 and I2. However, one of the underlying characteristics of cardinal utility and consumer surplus, is that it a descriptive measure, and as a result cannot be used for numerical analysis. As well as this, every consumer’s consumer surplus for good Y is going to be very different from consumer to consumer. As a result, it would be impossible to derive a model to sum up an aggregate population using this theory.
The main theory that works parallel with the rational choice model is the theory of expected utility, which helps derive the Kahneman and Tverskys Value function, whereas the Choice under Uncertainty describes consumer behaviour in certain scenarios. However, since this model is used to describe consumer behaviour in its own little niche, I have to agree with Kahnman and Tverskys argument that it cannot be fully representative of an aggregate population since every consumer is not put up against a risk factor in everyday purchasing habits. The value function does explain the irrationality of valuing losses more than gains which was one of the main drawbacks of the rational choice model. However, I do still believe that the rational choice model, despite its failures, is still the best way of describing consumer behaviour in normal circumstances.
Bibliography
R.H. Frank, Microeconomics and Behavior (7th edition), McGraw-Hill (2008)
Simon, Herbert, Models of Bounded Rationality, Vol. 3. MIT Press , (1997)
Amos Tversky; Daniel Kahneman, Science, New series, Vol 211, AAAS, (1981)
Herbert Simon, Administrative Behavior (3rd ed.), New York: The Free Press (1976)
Milton Friedman, Essays in Positive Economics, Univ. Of Chicago Press, (1966)
Thomas Kelly, Sunk Costs, Rationalilty and Acting for the Sake of the Past, Univ. Of Princeston Press, (2004)
Harvey A. Levine, Sunk Costs Don’t Matter, PM World Today Vol. IX Issue 1, (2007)
Amos Tversky; Daniel Kahneman, Prospect theory: An analysis of decision under risk, Cambridge University Press, (1989) (found in “Decision, Probability and Utility”)
Amos Tversky; Daniel Kahneman, Rational Choice and the Framing of Decision, Cambridge University Press,(1991) (found in “Decision Making: Descriptive, normative and prescriptive interactions)