Is it Irrational Behavior or Risk Aversion? (Round Two)

On July 15, 2010, I posted a blog with the same title except that it was round one. Today, I write a follow up hence the label round two. Why the designation? In the first round, I have contrasted the outcome of choice involving two medical options (pp.2, 3). The options involve a risky choice. One of the options had a certain outcome—the individual knows with certainty the outcome of the choice whereas the other was labeled “risky” choice as it involves uncertainty. I discussed how the individual calculates the outcomes of the two options by assigning what economists call utiles, a scale of measurement for the utility of the choice. My example was intended to show that the risk averter is likely to choose the certain outcome over the risky choice. I have also shown, that to make the uncertain option equally valued would require (given my numerical assignment) a reduction of the risk valuation ascribed to the risky option. In this round I would like to put forth another scenario. I would like to suggest another way of making option Y the winning option. For this to take place option X is rendered an “uncertain” option. Given that my examples are medical intervention options, the agent who can alter this and hence the choice is the “physician”.
Now assume that the physician who monitors the patient were to inform the patient that the certainty attached to the outcome of option X is no longer valid. That is, the probability of success now is equals to zero. Moreover, assume that the physician were to point out that there exists the probability that a side effect that did not exist earlier will materialize if option X is chosen. This means that not only the intervention with 100 percent probability has become ineffective, but also it carries with it a “negative” outcome. The comparison between the two options will hinge on the value a risk averter will assign to this negative effect. Note that in comparison to option Y there is no benefit attached to continuing the treatment. This scenario then pushes the individual to choose option Y without a change in its probabilities or the negative valuation attached to the side effect, a component of the option.
In short, the purpose of the exercise was to show how difficult it is for the individual to exercise choice when faced with uncertainty, not only because information may be incomplete but also because of his/her dependence on the market (third party) to evaluate the risky options. When an option is chosen in situations involving risk or uncertainty it is not an easy task to label the choice as rational or irrational. The saving grace is that in some situations a choice can be amended, in others the loss arising from the “wrong” choice cannot be recouped. That brings me to the literature that brought the risky choice models to the theory of consumer choice.
I have mentioned in the previous blog the seminal article by Milton Friedman and L.J. Savage (1948). A year earlier a most influential contribution by Von Neumann, J and O. Morgenstern’s Theory of Games and Economic Behavior (1947) offered utility functions that permit the complete ranking of options in situations involving uncertainty; the comparison of utility differences and the calculation of expected utilities thus making it possible to analyze choice in situation involving uncertainty (see chapter 1 of their book). Since then we have gained insight into this issue through contributions by several economists about consumer choice in situations characterized by risk and uncertainty. It is worth noting that risk has an objective probability while uncertainty involves assigning “subjective” probability. Hence the importance of ascertaining the individual type: whether he/ she is a risk averter, a risk neutral or a risk lover as this is critical to understanding choice. It is of note that an individual may be a risk averter in one situation and a risk lover in another. Now, I turn to the contribution of Behavioral Economists to the study of consumer choice.

Behavioral economists reject the economist models’ assumptions of rationality and maximization of utility. Rationality is defined as the “cognitive abilities” for solving economic problems .Behavioral economists dispute full rationality on the basis of research findings by psychologists and some economists “people exhibit preference reversals; have problems with self control and make different choices depending on how the issue is framed.” For a number of reasons, people make errors and behave in a manner contrary to their self interest. Given this premise, placing constraints on the exercise of free choice may be called for. That is, “paternalistic” intervention by the state or the community may be called for.
There are many variants of state paternalism: Paternalism (Mead, editor 1997), Patronizing Paternalism (Burrows, 1993), Libertarian Paternalism (Sustains and Thaler 2003), Permissible Paternalism (Goodin 1991), Benign Paternalism (Choi et al.2003), and Asymmetric Paternalism (Cramer et al. 2003). The different labels notwithstanding, the underlying premise of paternalism is simple: intervention by the state or the community will generate significant welfare gains. Sources that give rise to “bad” individual choice are: bounded rationality, slow learning, framing and lack of self control.

A question that needs to be posed: when people choices are ‘bad’, should the state and /or the community (a) Override their choices? (b) Steer them towards ‘welfare’ improving choices? (c) Encourage “good” choices without being coercive? or (d) Do nothing?
I have explored this issue in a conference presentation: “State Paternalism and the Rules of Reason” at the International Atlantic Economic Society meeting, which took place in Savanna, GA, on October 2007. In the paper presented, I have summarized the arguments put forth in a number of papers pointing out differences in the policies advocated to deal with “irrational” behavior. In a nut shell, paternalistic policies are advocated to help those individuals whose rationality is bounded (i.e. less than perfect) from costly errors. In the medical intervention example, if the individual was to reject option Y then he/she will bear a costly outcome.
It needs to be emphasized that not all policies advocated by behavioral economists call for coercion. Libertarian paternalism for example allows for differences between individuals and ‘covert’ coercions are not contemplated. At this juncture, it is befitting to call upon one of the architect of liberalism in economic thinking, John Stuart Mill. In his essay on liberty (1859), Mill wrote: “the only purpose, for which power can rightfully be exercised over any member of a civilized community against his will, is to prevent harm to others” (1984 edition, p.92). Following the writing of Nobel Laureate James Buchanan in his Logic of Limits (Buchanan and Musgrave 2000, p.111), one may ask: Why should constraints be placed on the individual when his actions do not infringe on others?
An understanding of the logic of limits may be gained by examining individuals choice in situations where they voluntarily impose restrictions on own actions. Behavioral economists advocating paternalistic intervention do so because they doubt the validity of the logic of limit concept altogether or at least as it applies to those individuals exhibiting bounded rationality. Buchanan’s observation that persons do adopt rules that they intend to abide by is valid for many but not for all. Not all smokers purchase stops smoking aid; alcoholics join Alcoholics Anonymous (the examples given by Buchanan, p.112).
Ruling out the logic of limits as it applies in the example of medical intervention given above in favor of paternalistic intervention; it is imperative to recall that “state or community” intervention most often entails coercion for they command the tools to implement said intervention (power to tax, impose fines, outlaws certain actions and so forth). Paternalism exercised by a parent and/ or a care giver does not command the same coercive power. In the medical intervention example cited above, the physician may attempt to move the patient towards his preferred option but he cannot coerce the patient to do so. Unlike the state, or the community, his power over the individual is not absolute; the patient can opt out of his care.
To conclude:
Behavioral economists have made a significant contribution towards our understanding of human behavior. It is undoubtedly true that some “bad choices” at least from the point of view of society are likely to be made, others are not .But placing constraints on free or voluntary choice of the individual should not be taken up lightly. To err on one side or the other demand more empirical proof that we currently do not have.
Each and every one of us can relate to situations where choices made were far from utility maximizing and/ or ‘fully’ rational. Most of us believe that the choices we voluntarily make are optimal in the sense that their expected costs are below those associated with alternative-dictated choices.