Everyday decisions, everyday decision-makers
Everyday decision-making is messy. Much messier than the well-structured risky prospects from which both the expected utility decision-maker and prospect theory decision-maker traditionally chooses. And much, much messier than the highly structured games of strategy that are constructed by classical game theorists. Risky prospects with outcomes x, occurring with probability p; are a part of everyday life and the expected utility decision-maker deals with them in one way while a prospect theory decision-maker deals with them in another way. Part of what makes ever)' day risky prospects messy is other people. This is not other people acting strategically but simply acting like humans, getting in each other’s way, generally being understandable but not completely so, generally being predictable to each other but not completely so, choosing too quickly, choosing too slowly, being too emotional or not emotional enough, not being analytical or being too analytical and, in general, proving to be spanners in the works.The decision-maker confronts the humanity of others and others confront the humanity of the decision-maker. Everyday decisions are made in the maelstrom of it all.
Messy risky prospects. Normally, the economist who wants to make things a little bit messy will introduce noise or uncertainty or incomplete information or some such thing into a formal, structured model. By contrast, we just introduce the idea that people are people and that includes people who work at law enforcement agencies and people who decide to join a terrorist group or who choose to do violence. People make risky prospects messy. From the perspective of law enforcement, a set of counter-terrorism measures produces outcomes x, with some probability pr The terrorists against whom these measures are directed are people and their actions and reactions change the expected outcomes and the probabilities. From the perspective of the terrorist group, attack methods produce outcomes x, (fatalities or media attention etc.) with probability ty.The law enforcement agents trying to stop them are people and their actions affect the outcomes and the probabilities. In thinking about the possible outcomes and their chances, each decision-maker is susceptible to cognitive biases and emotions. Furthermore, each side influences the other both by making rational, smart decisions and by making emotional decisions. This is what we mean by everyday messiness.
One thing, often overlooked, that makes human decision-making messy is that humans second guess themselves, rate themselves and assess themselves against some internal reference point. For example, the person who thinks that they are sometimes too vocal at meetings might resolve to be quieter next time or the person who thinks that they are too negative might resolve not to be pulled into negative discussions about their workplace around the water cooler.These perceptions of self are not necessarily shared by others.The person’s opinions may be valued and might not be seen as domineering and the gossip around the workplace water cooler might be a useful source of information for others or at least not seem as negative to them as it does to the person who is reflecting on their own contribution. Nevertheless, following a team meeting or a water cooler discussion, the person may feel as though they did or did not meet the reference point for behaviour that they had set themselves. This sort of self-referencing behaviour may be associated with self-signalling (see Bodner & Prelec 2002).
There are aspects of the self that the individual does not know. For example, when the time comes to attack, will the terrorist lose his nerve? Some terrorists do, some don’t. A prospective terrorist may be unsure as to which category he may fall into. As such, a prospective terrorist may view his decision to continue meeting with associates despite a detectable police presence as a signal (to himself) of his own courage or dedication.There are things about themselves that law enforcement agents do not know for sure either. It is only in the taking of action that a particular trait might be revealed (Bodner & Prelec 2002). Before an action is taken, however, there is only the likelihood that we are who we would like to be (or not). We know, of course, that people are very bad at estimating likelihoods. Is there a high or a low probability that the terrorist will follow through with a plan? He might not know for sure himself and his assessment of the likelihood is probably biased. And even if the probability assessment was correct, he might overweight or underweight it if his decision-making process is best described along the lines of prospect theory.
Everyday decision-making does not usually involve a ‘fork in the road’ choice. That is, a choice of one action at one point in time. It is messier than that. Hastie (2001, p.665) provides an alternative image:
The image of a decision maker standing at a choice point like a fork in a road and choosing one direction or the other is probably much less appropriate for major everyday decisions than the image of a boat navigating a rough sea with a sequence of many embedded choices and decisions to maintain a meandering course toward the ultimate goal.
While this type of image has given rise to constructions such as the ‘problem space’, the really crucial point was recognised by Shackle (1961).This is the fact that human decision-makers create their choices. They create the alternatives and then make a choice. They are not always hemmed in by available, pre-existing alternatives. Terrorists sometimes create something completely new. Law enforcement agencies continuously innovate. They are not subject to simply choosing the best counterterrorism strategy from a set of strategies.They can create new ones. In doing so, the decision-maker may surprise others and they may also surprise themselves.
Only a decision theorist or computer scientist could come up with the idea of the ‘problem space’. It would never occur to an ordinary person in a million years. Newell and Simon (1972) described what they mean by the problem space using the ‘nine dot problem’. In this problem, there are nine dots and the task is to draw four straight lines that pass through all nine dots without raising the pencil from the paper:
If the individual approaches the problem by delineating the problem space as being‘within the boundary of the dots’ (i.e. none of the drawn lines can go beyond the dots on the boundary), she will never solve the problem. If, however, the individual delineates the problem space such that lines can go beyond the boundary, then the problem can be easily solved (Newell & Simon 1972, p.90). As such, the problem space might not actually include the solution. In complex real-world problems, the problem space can be vast. The decision-maker tries to reduce this vastness by using new information or existing knowledge to limit the size of the problem space and delineate a suitable sub-space (Newell & Simon 1972, p.94). Of course, that sub-space too might not include the solution to the problem. Simple though it may be, this illustrates perfectly the idea that the individual in many ways creates a situation rather than just responds to a situation.
The playing field or social stage on which both the terrorist and the law enforcement agencies act is less structured than what it appears in analytical treatments. When a terrorist comes ‘face-to-face’ with law enforcement in this less structured everyday context, the progress from where both parties are now towards some ultimate conclusion is not a series of forks in the road where one side or the other is knocked ‘off course’. At many points along the way, there are not even ‘points’ until the decision-makers create them. What is interesting about economic analysis, and why it has had any success at all, is that the structures it creates are not dissimilar to the structures that people try to create in order to better understand and cope with the world around them. People do not face complete nihilistic uncertainty because they seek and create order. Much of behavioural economics is about responding to given situations. Our discussion is about responding to situations that are not just given but where the situation itself is partly the product of one’s own doing or one’s own attempt to impose order on a situation and in which other people also try to do the same.