Rational Decision-Making and the Information Paradox

Oftentimes people do not act rationally in the sense that they maximize ends and means or adapt aspirations in view of past experiences.

Rationality refers to purpose-directed action, guiding all actions towards the attainment of a goal or a set of goals (or ends).

Rational decision making requires knowledge and willful choice between known alternatives. However, resources and time to obtain information about alternatives are scarce and costly; endless information gathering before we can make a decision would lead to paralysis by analysis.

It is just not feasible. Economics Nobel Prize winner George Stigler Opens in new window once remarked: “Information costs are the costs of transportation from ignorance to omniscience, and seldom can a trader afford to take the entire trip” (Stigler, 1967; p. 297).

Rational ignorance is the behavior of people not to acquire certain types of knowledge in the face of the costs and uncertainties of information search.

We often choose to remain ignorant because it is simply too costly to be fully informed.

The question therefore arises to what point people should carry their information search: to the point where the expected marginal cost is equal to the expected marginal benefit (Stigler, 1971), or to a point where experience suggests to decision makers that they probably know enough to decide?

The answer is that individuals cannot know the expected costs and benefits of obtaining certain types of information, before they have acquired it.

Consequently, they are unable to maximize net returns from knowledge not yet acquired. Paradoxically, they would need the very information before they have acquired it. This logical insight has been called the information paradox (Arrow, 1971/1962).

Unlike the production of goods and services, where one has the knowledge about costs and benefits Opens in new window beforehand so that one can optimize resource use, the production of new knowledge cannot be subjected to such a rational calculus.

We can illustrate the point by thinking of a student who wonders whether it is worth $10 of his money to watch a specific movie. The only way to really find out is to spend the money and watch it!

Although the risk can be reduced by reading film reviews, it is always possible to conclude with hindsight Opens in new window that the money would have been better spent on something else, whatever a film review conveyed by way of partial knowledge of the movie.

In short, when searching for new knowledge, one never knows what one will find and whether the information is of as much value as expected. Often, one is even completely ignorant of what one misses until one makes a discovery.

There is another peculiarity in the production of knowledge: the costs for knowledge search have to be considered as sunk costs Opens in new window.

This means that the costs for knowledge production—once incurred—have no relevance to the extent to which the knowledge is used, whereas the costs of producing goods have a direct bearing on the amount which can be profitably produced (Streit and Wegner, 1992).

We speak of information search, when the general context is known and details have to be found out. This is, for example, done by software developers who want to improve a computer program.

By contrast, we speak of discovery when totally new knowledge is found and the discoverer is genuinely surprised. It takes good luck and a certain quality of enterprise to make such discoveries.

In practice, people engage in the search for information until they perceive that they have incurred sufficient expense and then make a decision within the bounds of what they have been able to find out.

We make decisions all the time that we may regret ex post. A classic example of this is going out to eat at a new restaurant — ex ante we expect to enjoy the evening of good food and drink, but if our meal does not meet our standard, we often wish we had gone instead to the older, more reliable restaurant.

The critical issue for economic activity is not our regret per se, but how that disappointment plays into our decision calculus in the future. It is the very discrepancy between our ex ante expectations and our ex post realizations that spur our learning.

This is, at least in part, what Shackle (1972, p. 156) means when he writes that “being consists in continual and endless fresh knowing. “

Information economics Opens in new window with its core idea of optimal search runs into the information paradox that we referred to above.

Because obtaining and analyzing new knowledge is costly in terms of time, effort and resources, no one on his own will acquire all the knowledge needed for complex operations.

People will rather seek to exploit the knowledge of others by interacting with them. Indeed, it is rational for people to acquire only certain bits of information and to remain ignorant of other information, given the high costs and uncertain outcomes of knowledge search, a phenomenon known as rational ignorance.

Different Kinds of Rationality

What has been said so far has a further important consequence:

people who incur the costs of searching for useful knowledge under conditions of uncertainty cannot be all-knowing and will often not be able to make absolutely rational choices.

They will normally even find it hard to stick rigidly to pre-conceived, set of objectives and aspirations. In the process of interacting with others, their aspirations will evolve.

If people are frequently disappointed, they tend to peg back what they aspire to. If objectives are reached easily, they may become more ambitious and hope to attain new wants.

In other words, people will often display adaptive behavior, recognizing the bounds of what they can attain. This is, however, not invariably the case. At other times, enterprising people will act to break through constraints which they normally accept as given, or they may discover new opportunities by chance.

We therefore have to distinguish between three kinds of rationality:
  1. End-means rationality, where the ends and the means to attain them are known and actions are taken to attain them, for example from earlier experience;
  2. Bounded, adaptive rationality, which relates to situations when people do not all-knowingly maximize their given objectives, but adapt their aspirations (goals) in the light of experience to what is feasible for them; and
  3. Entrepreneurial-creative rationality that, on occasions, breaks out of recognized bounds or exploits chance discoveries.

Instances of end-means rationality can be observed where agents have set those goals for them, and where people use available resources and techniques to maximize the attainment of these goals.

However, frequently the other kinds of rationality are prevalent. What if you are not a top-rated long-distance runner, but nevertheless want to compete in marathons? You will rationally adjust your objective to your own capacity in the light of past performance and get great satisfaction from running at a speed commensurate to your capability.

Instead of following absolute goals, people gradually discover what they can achieve and handle their goal-attainment rationally in an adaptive way. This is so because people are frequently limited in their capacities to absorb and evaluate more information.

American economist-sociologist Herbert Simon Opens in new window called this bounded rationality or procedural rationality (Simon, 1983). Instead of optimizing given objectives, people and businesses ‘satisfice’, he wrote: they adjust their aspiration levels in the light of past experience.

At times, people do not act rationally in the sense that they maximize ends and means or adapt aspirations in view of past experience.

People, who normally accept given constraints and act within them, may at certain junctures decide to tackle these constraints head-on in order to widen or bypass acknowledged obstacles to their goal achievement (Schumpeter, 1961/1908).

In these cases, individuals act out of a creative and entrepreneurial motivation and overcome technical or other constraints.

Acting in creative-entrepreneurial ways sometimes also involves the breaking of institutional constraints Opens in new window, for example by violating established conventions and customs, or by political lobbying to gain support for the change of a certain rule.

We therefore have to recognize creative rationality, which propels people to take risks and break new paths and keeps them on the alert for new discoveries, as a third type of rational behavior, and one that is apt to enhance human knowledge.

It would be plain wrong to describe entrepreneurs Opens in new window who try to market a new product as irrational, although they do not optimize known ends and given means.

One may go further and acknowledge that a considerable part of our daily behavior is not rationally and logically targeted at any specific, identifiable objective. Much actions is coordinated by habits which cannot be explained in any one specific instance by rational calculus:

Often people simply follow habitual patterns or imitate others.