This summary is written in 2013-2014.

Introduction

The structure of this summary is similar as the book; Explaining Social Behaviour by John Elster. It overlooks over five themes each being constructed by different subjects.

I. Explanation and Mechanisms

• Explanation

• Mechanisms

• Interpretation

II. The Mind

• Motivations

• Self-Interest and Altruism

• Myopia and Foresight

• Beliefs

• Emotions

III. Action

• Desires and Opportunities

• Persons and Situations

• Rational Choice

• Rationality and Behaviour

• Responding to Irrationality

• Some Implications for Textual Interpretation

IV. Lessons from the Natural Sciences

• Physiology and Neuroscience

• Explanation By Consequences and Natural Selection

• Selection and Human Behaviour

V. Interaction

• Unintended Consequences

• Strategic Consequences

• Games and Behaviour

• Trust

• Social Norms

• Collective Belief Formation

• Collective Action

• Collective Decision Making

• Organizations and Institutions

Social behaviour is subject to some very curious puzzles. To illustrate some:

• Why do some gamblers believe that when red has come up a number of times in a row, red is more likely to come up the next time?

• Why is it more likely for black to come up for other gamblers?

• Why is shame more important than guilt in some cultures?

• Why are people unwilling to break rules imposed on themselves, even when they are aware that those rules make no sense?

• Why did no one call the police when Kitty Genovese was beaten to death while there were some thirty bystanders?

• Why do people engage into projects with a negative expected value?

• Why do people take revenge when it has material costs for them without any material benefits?

• Why do people jump to conclusions?

• Why do individuals vote? There is no clear influence that the one vote can have…

• Why are successful people in the Western societies usually slimmer than non successful ones?

These puzzles are all connected to different categories. They can be puzzles on the mind, on action, lessons from the natural sciences or interaction.

Chapter 1. Explanation

Social science cannot rely on functional explanation. Functional explanation is through the fact that actions and behaviour can be explained by their consequences. Rational choice theory is overvalued in explanatory power as well. On the other hand, while people are not likely to go through a long list of mathematical equations before taking social action, game theory and the desire to act rational may give the rational choice theory credibility.

Social science is there to explain social phenomena. All explanation is causal. To explain a phenomenon (an explandum) is to cite an earlier phenomenon (the explanans) that caused it. An event can be explained by an earlier event which would be its cause. This is the basic event-event pattern. David Hume introduced this as the billiard-ball model. The first ball hitting the second is the cause and therefore the explanation of the second ball moving. Facts are assumed to be of more importance and result in a fact-event or event-fact explanation. For example:

Fact: Roads are slippery.

Event: Car went off the road.

Fact-event explanation can be used for the car accident.

Event: Attack on the WTC in New York.

Fact: Americans live in a state of fear.

Event-fact explanation for the level of fear many Americans have.

There can also be a fact-fact explanation. Here the existence of one fact fully explains the existence of another, there is no event necessary to prove it. A good example is the fact that the level of education of women explains the fact of the per capita income in the developed world.

Social scientists are not very interested in explaining the underlying influences that caused a statistic to exist. There is more emphasis on trying to find differences and changes in attitudes over time or across populations. The simple statement of a statistic has far less meaning than when it can be compared to something else. A 65% percentage in favour of the death penalty has far more meaning when it is compared to 80% which was the percentage 15 years ago.

Longitudinal studies are used to compare statistics over time. Cross sectional studies can be used when we want to compare populations at one point in time. The explanandum is not the phenomenon but its change over time or differences in other groups. To assess the success of an explanation we need to know the variance and how much of the variance is accounted for by the explanation. Perfect explanation of variance is absolute success. It does not say anything about the explanandum in and of itself, however.

In some cases the social scientist will try to explain non-events rather than events. If the non-event is in existence due to a decision making process, we will be able to regard this decision making an event. There will have to be a positive event somewhere down the line of reasoning to be used when we keep asking why certain non-events occur. When this positive event emerges we can explain that and the non-event will be explained as a by-product or residual.

Explanations should be only on individuals and their actions. In practise however, due to the second best-approach or as harmless shorthand, supra-individual entities such as families or firms are used as well. They are the result of lack of data.

Some events can be better explained by consequences than cause. This does not include intended consequences, just the ones that actually occurred. The actual consequences will be beneficial for someone or something. This seems to be incompatible with causal explanation; cause precedes its effect. One should look at this as a loop. The loop takes the consequences back to its causes. A child that cries could be in distress, but may also cry to get attention like it would when it would be in distress.

Before we try to explain a fact or event we have to make sure this fact or event actually exists. There should not be a blind belief in the data at hand. There may have been incentives to corrupt data. When trying to compare two groups we need to make sure that they are measured according to the same principals. Unemployment, for example, can be measured and defined in different ways. These different methods may paint a picture of large differences, while they may be effectively similar if looked upon the same way.

A puzzle is a well-established explanandum for which there is no explanation. They should be approached in a five-step method:

1. Choose the theory with the greatest promises of successful explanation.

2. Specify a hypothesis. The explanandum should logically follow from the hypothesis.

3. Come up with plausible alternative explanations with the explanandum being a logical fact to follow.

These steps can be in a different order. Different hypothesis can be looked at until we find one with the most promises.

4. Refute rival accounts by identifying additional testable implications which are not observed.

5. Test additional testable implications for the chosen hypothesis which are in fact observed in order to strengthen the proposed hypothesis.

This procedure is called the hypothetic-deductive method.

An explanation is supported from below if we can deduce and find verifiable facts from the hypothesis that go beyond what the hypothesis is intended to explain. This is called excess explanatory power. The support for a hypothesis is from above if we can deduce the hypothesis from a general theory. Leon Festinger has a theory that when there is inconsistency in believes and values, the mental adjustment to deal with this will be one of the least resistance. When someone buys a ticket for a show which costs him a great amount of money, it is not possible to deny this fact. If the show is not good enough to justify this expenditure it is easier to convince yourself of the opposite. This is support from above for the theory that standing ovations in Broadway shows are more frequent now due to the price increase of tickets.

Lateral support for an explanation is when we can think of alternative explanations and refute them. There needs to be a consistent process of thinking up methods to discredit yourself in order to maintain a high level of proof for your initial hypothesis. This process creates a higher level of credibility of any explanation you can come up with. They must of course be novel facts, which have a level of difficulty in refuting them to increase credibility. Theory is supported by the successful explanations rather than the other way around.

Explanation can be a tricky thing to do. To be sure the basics are clear, it is necessary to explicitly distinguish explanation from 7 others types of statements:

1. Causal explanation is not the same as true causal statements. In causal explanation the cause is accompanied by the causal mechanism. It does not need to be literally stated in each case, but there needs to be at least a suggestion on the causality. When we state that someone died of eating rotten food, we can assume the cause of death was food poisoning. There is no need for the explicit mentioning. Just remember that when we want to use causal explanation we need to look beyond the statement but also at the mechanism.

2. Causal explanation is not the same as statements on correlation. Correlation is not causality. Just because two events of facts seem to occur at the same time in many occasions does not imply they have a causal relationship. They can both occur due to the fact that a third factor causes both of them. There are many examples for explaining this distinction. One is the correlation between disturbed children and divorced parents. Did the fact that the children were disturbed cause the parents to split up or did the divorce disturb the children?

3. Causal explanation is not the same as statements on necessitation. Why did something happen as it happened? When trying to explain passed events it is nearly impossible to do so without being biased by hindsight. Just the fact that there was a factor that caused for an event to happen does not mean that if this factor was not there the event would have never occurred. There may have been another influence that would have caused the event. The first factor was therefore not necessary for the event to happen.

4. Causal explanation is not the same as storytelling. Storytelling is accounting for an event as is might have happened. Storytelling can be useful for new explanations, but it can be misleading and harmful when it is mistaken for explanation. There is a tendency to state things with as-if status which have no explanatory value.

5. Causal explanation is not the same as statistical explanation. Statistics cannot account for individual events. It is an intellectual fallacy to assume that a generalization is valid in individual cases. It is a moral fallacy to take these generalizations and treat individuals as a case rather than a person. The problem with the distinction between causal explanation and statistical explanation is the fact that in many cases we do not have another choice but to rely on statistical data.

6. Causal explanation is not the same as answering why-questions. To explain this consider the following example. When writing an article a scholar did not include a citation of a very relevant article. Asking yourself why may result in the answer that he did not include the article because he was not aware of its existence. This, however, is explaining a non event with a non event. Such an explanation is not causal explanation. If the article was not included because the scholar chose to do so, the answer to the why-question would have been a relevant causal explanation.

7. Causal explanation is not the same as stating predictions. We can explain without predictions and predict without explanation. It is much easier to explain something after the fact than to predict it beforehand. Also in some cases we can predict behaviour without needing to know the causal explanation behind it. When prices go down, consumers will purchase more of the same good. We do not need to explain the process to predict it to happen. When it comes to prediction we do not need to know the difference between causality, correlation and necessitation. As long as one happens when another does it is good enough for prediction, while it will never be usable for causal explanation.

Chapter 2. Mechanisms

Philosophers of science often argue that an explanation must rest on a general law. Citing a set of initial conditions with a statement to the effect that those conditions obtain an event will follow is explaining an event. But this is not sufficient and brings two objections to this idea. There can be explananda occurring when there is correlation or necessitation, there does not need to be causal explanation for that. If we replace the idea that the explanation must rely on general law with the one that an explanation must rely on a mechanism, we can get around this problem. The mechanism identifies the causal chain of events and facts that lead up to the explanandum. This is also referred to as opening the black box.

The credibility of an explanation increases with the extent to which general laws form a causal chain. The more links we have to build our chain with, the smaller the danger is that we have lurking variables disturbing the true causality. There may always be, however, the danger of lurking variables. Just because the chain will make the general laws less abstract, does not mean there is immunization possible against lurking variables. There is just a lower amount of variables that need to be controlled for.

General laws do not give satisfactory answers. Even when it is clear beyond any doubt that there is correlation or causality between two factors and all possible third influences are controlled for, we do not only want to know this. We are interested in how causality works, what causes what and why. The more general the law used, the less satisfaction a causal explanation will give.

The law of demand as it is known in economics is one of the very few laws that can be named in social sciences. Even so it is a weak law. The law only states that when prices go up, demand will fall and vice versa. The direction of the relationship may be evident; there is no indication of how much demand will react to price changes. This makes the law weak.

Furthermore, the law of demand cannot be used to explain. There is no way to explain why the individual requires less of a good when prices increase. To explain why consumers buy less of a good when it becomes more expensive we would have to adopt and test a specific assumption about an agent reactions to price changes. This principle reduces hypotheses to individual level is called ‘methodological individualism’.

Through mechanisms individual behaviour is explained. Mechanisms are frequently occurring and easily recognizable causal patterns that are triggered under generally unknown conditions or with indeterminate consequences. Mechanisms are used for explanations but not for predictions. By relating behaviour to a pattern we can explain this behaviour, but without explaining why the pattern occurs.

Mechanisms are frequently occurring and easily recognizable causal pattern, likewise as proverbs are passed down through many generations. Proverbs sum up a general principle or common situation. Proverbs state mechanisms, rather than general laws.

Many proverbs state opposites, such as the combination of ‘absence makes the heart grow fonder’ and ‘out of sight, out of mind’. But many opposite mechanisms are not represented by proverbs. And there are proverbs that do not have an opposite, while they still describe mechanisms. We believe easily what we hope and what we fear. It basically states that next to wishful thinking there is also a phenomenon called counter motivated thinking.

Sometimes a proverb indicates that two mechanisms with opposite effects on the outcome are triggered simultaneously. The interdeterminancy is created by the net effect of the two. Examples of such proverbs are ‘necessity is the mother of invention’ and ‘it is expensive to be poor’. The first describes a desire, the second opportunity. As we are driven by both, we cannot state what will be the net effect on behaviour.

We are not restricted to proverbs. When you look at the psychology of tyranny, there is a relationship between hatred and fear that cannot be ignored. If hatred exceeds fear, the tyranny will backfire. This is a mechanism that occurred in World War 2 as well, where the oppressed fought back by killing German soldiers in order to reduce the tyranny. In a smaller case, the same process occurs when someone faces a barrier to his or her goals. The barrier will call for action to re-establish freedom. This counteracts the function of the barrier and we cannot predetermine which will be stronger.

Even when the net effect of opposite mechanisms is known, this is not the same as explaining them. Maybe we can assess strength by looking at both effects in situations where the other does not occur. If we then have a look at the net effect they have we may be able to make an attempt at explaining the process that leads to particular individual behaviour.

The mechanisms discussed so far can be seen as atomic mechanisms. They cannot be reduced to other mechanisms at the same level. Combining atomic mechanisms, we can come up with molecular mechanisms. Molecular mechanisms are more complex and consist of atomic mechanisms (building blocks). To illustrate, consider the following proverbs. ‘Fear is often greater than the danger’ and ‘Fear only increases danger’. When you combine them, fear creates its own justification.

Opportunities for behaviour will not produce such behaviour by themselves. There must be a motive. Will a democratic regime restrain the desire of citizens to do what democracy allows them to? Tocqueville claimed that democratic citizens would turn to religion to satisfy the need for an authority democracy would not provide. There should be less focus on opportunities and more on desire according to him.

If the spillover effect is higher than the compensation effect, the lack of a strong authority will weaken religion. But we still cannot know the net effect. If there is a great deal of opportunity and only a small amount of desire restraint there will be very different outcomes. If the influence of democracy on religion is lowered through the compensation effect, democratic society will be religious. Schematically see the figure above.

If we know a mechanism, we may be able to find what triggers it and come up with a general law. Sometimes we can predict which mechanism will be triggered when there are opposite effect possible. This can be tested by placing the opposite possibilities to the test and using a third control group to make sure there are no lurking variables.

Sometimes different causes can interact. They can have additive characteristics. Such mechanisms can be the level of education and income of parents. They will both have an effect on the level of education of the child. There is an interaction between age and political attitudes when it comes to possible extremism. There is interaction between drug intake and the mental state beforehand which both determine the state of mind after drug use. Some say it is additive, other believe it to be reversal. A reversal interaction in drug use is when it is seen as a mood multiplier, a bad state gets worse, while a good one gets better.

If our intuition on how several mechanisms can produce the same outcome would be enough, we would not need to focus so much on the interaction. It is because we do not know because out intuitions are weak, that we need to take interaction into account when looking for causality. If we want to have law like explanations we tend to tinker with the models. As this comes with some dangers, we might be better of with lowering our expectations instead.

Chapter 3. Interpretation

Writings on the humanities focus on the interpretation rather than explanation. The question is not whether the social sciences rely on understanding or on explanation. This is because to interpret is to explain. Interpretation is nothing but a special case of the hypothetico-deductive method. Interpretation is directed to human actions and to the product of human actions.

Interpreting an action requires to explain it in terms of the antecedent beliefs and desires (motivations) of the agent. These mental states should be explained in a way that makes sense within the full desire-belief complex. An isolated desire or belief that does not have the normal kind of solidarity with other mental states is just a brute fact that may allow explaining behaviour but not understanding it. A paradigm mode of explaining action means that it was performed because it was rational; optimal from the agent’s point of view. If behaviour is rational, it is also intelligible. Here are three varieties of intelligible but irrational behaviour contrasted with some cases of unintelligible behaviour.

The first arises when the machinery of decision making is truncated in one way or another; a strong emotion may prevent the agent from gathering information before acting. The agent rushes into action rather than adopting a waiting strategy without considering the consequences.

A second variety arises in the short-circuiting of the machinery of decision that occurs when belief formation is biased by the agent’s desires. Motivated belief formation optimizing processes: they maximize the pleasure the agent derives from his beliefs about the world rather than the pleasure he can expect from encounters with the world.

The third variety is called a wire-crossing in the machinery of decision. The mind also pursues dissonance production. As a clarifying example; people believe easily what they fear.

Following are two examples to point out the ideas of intelligible beliefs and desires, both taken from civil wars in the past and present. First example, the belief in predestination which divided Calvinists and Catholics in the wars of religion. The origin of the war was the intense religious anxiety experienced by many believers in pre-Reformation times, due to uncertainty about one’s salvation. Since god had chosen from eternity who would be saved and who whould be damned, there was nothing one could do for one’s salvation and hence no reason to worry that one had not done enough. The key interpretive issue concerns the link between belief in predestination and the belief from anxiety. The Calvinists could not hold that rational and systematic effort would bring them salvation, but it would bring them subjective certainty of salvation. This emphasizes the twin mechanisms of wishful thinking and magical thinking lend intelligibility to the belief in predestination.

The next example covers the intelligibility of motivations. The main motivation for young Palestinians to give their life in suicide missions, defend their homeland, is not difficult to understand. Additional causal factors are needed to make the strength of the motivation intelligible. The gap between expectations and reality experienced by the many educated Palestinians who are without any prospects is a more plausible factor (relative deprivation). The most relevant features seem to be permanent feelings based on comparison between oneself and others (inferiority), the second interaction between oneself and others (resentment).

Behaviour is explained in terms of the antecedent mental states (desires, beliefs). Statements (verbal as nonverbal) about the agent’s motivation are used as evidence. All methods for consolidating individual preferences into a social decision create incentives for the participants in the process to misrepresent their preferences. There are many reasons why people might want to misrepresent their motivations and those of their opponents. Society has a normative hierarchy of motivations that induces a desire to present oneself as animated by a noble motivation rather than by a baser one. This impute a low-ranked motivation to the opponent. The problem of self-serving bias in statements about the intentions of social agents is serious, but not insurmountable. The solution is to consider the objective interests of the agent and assume that in the absence of strong contrary evidence they coincide with her subjective motivation. Alternatively, one might identify the actual consequences of the agent’s action and assume that in the absence of strong evidence to the contrary what the intendancy was.

Chapter 4. Motivations

To understand how people act and interact, first we have to understand how their minds work. This understanding is a mater of introspection and folk psychology. The outcome of the determination of beliefs and desires depend on many factors.

Some desires are irresistible, people feel like there is no alternative. Another extreme is a paradigm (caricature) of the rational agent who is unperturbed by visceral factors, including emotion. The agent acts only after having carefully weighted the consequences of each available option against one another. The distinction between visceral and rational motivation is provided by the difference between visceral and prudential fear. It is common to refer to fear as an emotion as a belief- , desire complex. Between the extremes of the visceral-rational continuum, we find behaviour that is partly motivated by visceral factors, yet is also somewhat sensitive to cost-benefit considerations. In more complex cases, one visceral factor might counteract another. As an example, the desire for an extramarital sexual affair might be neutralized by guilt feelings.

The seventeenth-century French moralists made a distinction among interest, reason and passion. Interest is the pursuit of personal advantage. Reason is used in relation to the desire to promote the public good rather than private ends.

Id, Ego and Superego are three basic forms linked to a separate subsystem of the mind. Respectively the three systems correspond to the pleasure principle, the reality principle and conscience. The id and superego represent respectively impulses and impulse control, while the ego defends itself vainly.

Motivations may be consequentiality, oriented either toward the outcome of action or non consequent, oriented toward the action itself. Economic behaviour is mostly purely consequentiality.

The distinction between wants and wishes is important if we look at the motivational component of emotion. Emotions can be accompanied either by a want or a wish that a situation is the case. Wishful thinking is based on wishes rather than on wants. The agent alternatively van refrain from the hard work of making the world conform to his desires and adopts instead the easy path of adopting an appropriate belief about the world.

The wish-want distinction is being complicated by a factor that in some cases; I can get X by doing A, but only if I do A in order to get Y. Indirect benefits are parasitic on the main goal of the activity. Example; A person who pursues a political career, may found at the end with it that he also obtained a character.

Why do people leave one city for another? Or a job? People move because they compare both situations, with the difference as a justification. Here we distinguish push motives from pull motives. Push motives are closer to the visceral end of the continuum and pull motives closer to the rational end.

The existence of competing motivations is commonplace. In general the strongest motivation wins. But this is to simplistic, since the idea of ‘strength of motivation’ is complicated. A motivation may owe its strength to its sheer psychic force; this is the sense in which, for instance, visceral motives are often stronger than ‘the mild voice of reason’. A strong motivation may also be one that the agent endorses strongly because of the high valued place in society. Each society or culture is characterized by a normative hierarchy of motivations. A person would rather perform a given action for a motive higher in the hierarchy, these we cal metamotivations.

Cognitive dissonance theory predicts that when one motivation is slightly stronger than another, it will try to recruit allies so that the reasons on one side become decisively stronger. In this situation strength of motivation cannot be taken as given, but should rather, to some extent at least, be seen as a product of the decision-making process itself.

Chapter 5. Self-interest and Altruism

Altruistic motivation is the desire to enhance the welfare of others even at a net welfare loss to oneself. An altruistic act is an action for which an altruistic motivation provides a sufficient reason. This makes the contrast between self-interested and altruistic motivation deceptively simple. But to understand the distinction between an altruistic act and a altruistic motivation is important. An example will clarify this distinction; giving money to a charity for the ‘’warm glow feeling’ can be seen as a altruistic act but the altruistic motivation is to make yourself feeling better. The motivation mimics altruism.

There exists a distinction between true and false motivations acting for the sake of what is right or what other people may think about you. Approbativeness is the desire to be well thought of by others and Shamefulness is the desire not to be thought badly of by others. This are both a bad motivation at first sight but can be seen as a substitute of altruism and morality. The main point of this section is to see that motivations are being substituted to induce all sorts of socially wasteful behaviour.

Original motivation of the agent can be transmuted from interest into reason. The mechanism behind this transformation is ‘amour-popre’ (or self-love), the desire for esteem and for self-esteem. Approbativeness and shamefulness are derived from the desire for esteem only affects outward behaviour. Desire for self-esteem can affect the inner motivations themselves. People try to put a non-self-interested glos on their actions wich fulfil their self-interest. People have to degrees of freedom in their efforts to justify their behaviour to themselves.

First one is plausible-sounding causal theories supporting that benefiting themselves will also benefit others. Example; lower taxes on the rich wil eventually benefit the poor also. Second is the plausible-sounding normative conception which says that a person would be pretty unfortunate if failing by coinciding with her self-interest.

Reciprocity is a simple dyadic relation when each party in an ongoing relationship faces the choice between cooperating and not cooperating. Direct reciprocity is A helps B if and only if B helps A. Indirect reciprocity is A helps B if B has helped C. The existence of indirect reciprocity suggests that people might behave altruistically in order to develop a reputation for having altruistic motivations.

Moral norms include the norm to help others in distress, the norm of equal sharing, and the norm of ‘’everyday Kantianism’’ (do what would be best if everyone did the same). Social norms include norms of etiquette, norms of revenge and norms regulating the use of money. Quasi-Moral Norms include the norm of reciprocity and the norm of conditional cooperation (cooperate if others do otherwise not). Social norms and quasi-moral norms are conditional. Social norms are triggered when other people can observe the actions of the agent, quasi-moral norms are triggered when the agent can observe what other people are doing. Moral norms are unconditional.

Next to the agent’s own motivation, the explanation of her behaviour must appeal to her beliefs about the motivations of others. The motivations of others cannot be detected at face value but with triangulations of the general kind and techniques that only apply to face-to-face interactions

Chapter 6. Myopia and Foresight

The tendency of the pleasure principle is to seek immediate gratification of desires. There are two manifestations for this tendency. First the adoption of the belief one would like to be true rather than the belief that is supported by the evidence second occurs in the choice between two actions that induce different temporal utility streams. Principle of pleasure dictates the choice of the stream with the highest utility in the first period regardless of the latter periods. Gradient climbing is when at any point in time an agent searches for whether one of the nearby options has yields a greater immediate benefit than the status quo. Gradient climbing leads to a place that is superior (in the short run) to all nearby positions, the agent reaches a local maximum.

Humans have the ability to re-present the absent which enables us to go beyond the pleasure principle and take account of temporally remote consequences of present choices. Planning ahead enables us to make better choices.

Here are four examples of acting on the basis of temporally remote consequences:

• Reculer pour mieaux sauter. The fundamental fact of economic life that to invest for greater consumption in the future one must consume less in the present. The agent accepts a state that is inferior to the status quo because it is a condition for realizing a superior alternative later on.

• Waiting. The agent is willing to reject an option that is superior tot the status quo because the rejection is a condition for realizing an even better outcome later. This comes with a risk namely the possibility that nothing better suited comes along.

• Shooting ahead of the target. To hit a moving target, one should not aim at where it is, but at where it will be at the time of encounter. Similarly to pursue a moving target, one should aim in a straight line at where the target will be rather than follow the curved path induced by always aiming at its current position.

• A straight line is not always the fastest way. When trying to reach a stationary target, a straight line is not always the most efficient path. An example to clarify this; a rescuer might impulsively run straight toward the drowning swimmer until he reaches the shoreline and then swims the remaining distance. Is the swimmer reflected he could realize that he runs faster than he swims. He could have reached the swimmer faster by taking an indirect path that, although longer on the whole, has a shorter stretch in the water.

The existence of the capacity for long-term planning does not imply that it will be used. For perceived long-term consequences to make a difference for present behaviour, agents must be motivated. For agents to subject to excessive time discounting two elements are needed; cognitive and motivational elements. Present behaviour cannot be motivated when future outcomes are shrouded in uncertainty or involve risk. The ability of future outcomes to shape present behaviour is affected both by the time at which and by the probability with which they will occur. The mechanisms by which they affect choice are, respectively, time discounting and risk attitudes.

Chapter 7. Beliefs

Believing in something generates action. The role of belief can be understood by understanding their nature, their causes, and their consequences. In everyday language ‘belief’ suggest less than full endorsement. Knowledge is defined as a justified true belief, example: I know I am married. Knowledge stands in a particular relation to the world (truthfulness) and to the body of evidence the agent possesses (it is justified).

We distinguish four cognitive attitudes to the world (with decreasing strength):

1. The mode of Certainty.

2. The mode of Risk; setting probabilities to exhaustive outcomes

3. The mode of Uncertainty; Knowing the set of outcomes but unable to attach probabilities

4. The mode of Ignorance; Outcomes and their probabilities are fully or partly unknown

Certainty and Risk are the most common. People feel irresistible compelled to form an opinion even when people have no grounds for having any belief on a topic.

Mechanisms generated on error rely on motivation. Yet error can also revise form ignorance (with confidence).

A judgment on probability comes from observation of objective frequencies or be purely subjective evaluations. Observation of objective frequencies means that by looking on a large number of observations of similar situations an agent can say something about a upcoming similar situation.

Expert likewise the average person often goes wrong because of violating the principles of statistical reasoning.

Selection bias: an error in choosing the individuals or groups to take part in a study. Ideally, the subjects in a study should be very similar to one another and to the larger population from which they are drawn (for example, all individuals with the same disease or condition). If there are important differences, the results of the study may not be valid. Selection bias comes in two types: (1) self-selection of individuals to participate in an activity or survey, or as a subject in an experimental study; (2) selection of samples or studies by researchers to support a particular hypothesis.

Regression to the mean: is a principle in statistics that states that if pair of independent measurements is made from the same distribution, samples far from the mean on the first measurement will tend to be closer to the mean on the second one. Moreover, the farther from the mean on the first measurement, the stronger the effect is. Random variance affects the measurement of any variable: this random variance will cause some samples to be extreme. For the second measurement, these samples will appear to regress because the random variance affecting the samples in the second measurement is independent of the random variance affecting the first. Thus, regression toward the mean is a mathematical inevitability: any measurement of any variable that is affected by random variance must show regression to the mean.

Availability heuristic: The availability heuristic is a phenomenon (which can result in a cognitive bias) in which people base their prediction of the frequency of an event or the proportion within a population based on how easily an example can be brought to mind. We make a judgment based on what we can remember, rather than complete data. In particular, we use this for judging frequency or likelihood of events.

Representativeness heuristic: The representativeness heuristic is a heuristic wherein people assume commonality between objects of similar appearance, or between an object and a group it appears to fit into. While often very useful in everyday life, it can also result in neglect of relevant base rates and other errors. People tend to judge the probability of an event by finding a ‘comparable known’ event and assuming that the probabilities will be similar. As a part of creating meaning from what we experience, we need to classify things. If something does not fit exactly into a known category, we will approximate with the nearest class available. Overall, the primary fallacy is in assuming that similarity in one aspect leads to similarity in other aspects. The gambler’s fallacy, the belief in runs of good and bad luck can be explained by the representativeness heuristic.

Magical thinking is the tendency to believe one can exercise a causal influence on outcomes that are actually outside one’s control. Causal and diagnostic relevance is when people act as though acts that are merely 'correlated' with a particular outcome actually 'cause' that outcome. This may explain why people bother to vote: They know that their decision to turn out is correlated with the decision by people like them, so they turn out thinking that this will cause people like them to turn out as well.

Content-generated beliefs are of two main varieties. Mainly the agent may be motivated to hold some belief or a specific belief. The mechanisms generating this variety are rationalization, wishful thinking and self-deception. With rationalization the behaviour occurs first and the belief follows. In wishful thinking and self-deception, the belief occurs first and the behaviour follows.

Wishful thinking is the formation of beliefs and making decisions according to what might be pleasing to imagine instead of by appealing to evidence, reality or rationality. This makes wishful thinking like self-deception. These two phenomena have in common that a desire for a situation causes the belief for the existence of that situation. In wishful-thinking this is a simple one-step process: the wish is the father of the thought. Self-deception involves four steps: first the evidence is considered, second the appropriate belief is formed, third this belief is rejected or suppressed because it is inconsistent with our desire, and last the desire causes another and more acceptable belief to be formed in its place.

Self-deception is the process or fact of misleading ourselves to accept as true or valid what is false or invalid. Self-deception, in short, is a way we justify false beliefs to ourselves. When philosophers and psychologists discuss self-deception, they usually focus on unconscious motivations and intentions. Examples of self-deception are things like their weight, their drinking habits and their faithfulness.

Chapter 8. Emotions

Emotions enter human life in three ways. First they are the most important sources of happiness and misery. The second way is on impact on behaviour. Third, emotions can matter because of their impact on other mental states.

There is no agreed-upon definition of what counts as an emotion. There are no common nor necessary features for emotions. For the purpose of social-scientific explanation we focus on occurrences of emotions in which a certain number of features are regularly observed and how these explain behaviour. These are the features:

• Cognitive antecedents. Emotions are triggered by beliefs, often by the agent’s acquiring of a new belief.

• Physiological arousal. Emotions go together with changes in the human body

• Physiological expressions. Emotions go together with characteristic observable signs

• Action tendencies. Emotions are accompanied by tendencies or urges to perform specific actions.

• Intentional objects. Emotions are about propositional objects or non-propositional objects

• Valence. Technical term for the pain-pleasure dimension of the emotions as experienced

The first important group of emotions are the evaluative emotions. Evaluative emotions involve a positive or a negative assessment of one’s own or someone else’s behaviour or character. If an emotion is triggered by the behaviour of another person, that behaviour may be directed either toward oneself or toward a third party. These distinctions yield eleven emotions altogether:

• Shame is triggered by a negative belief about one’s own character

• Contempt and hatred are triggered by negative beliefs about another’s character

• Guilt is triggered by a negative belief about one’s own action

• Anger is triggered by a negative belief about another’s action toward oneself

• Cartesian indignation is triggered by a negative belief about another’s action toward a third party

• Pridefulness is triggered by a positive belief about one’s own character

• Liking is triggered by a positive belief about another’s character

• Pride is triggered by a positive belief about one’s own action

• Gratitude is triggered by a positive belief about another’s action toward oneself

• Admiration is triggered by a positive belief about another’s action toward a third party

The second group are a set of six emotions generated by the position of someone else. The purpose of these emotions is neither individual action nor individual character, but a state of affairs.

• Envy is caused by the deserved good of someone else

• Aristotelian indignation is caused by the undeserved good of someone else. The closely related emotion of resentment is caused by the reversal of a prestige hierarchy, when a formerly inferior group or individual emerges as dominant

• Sympathy is caused by the deserved good of someone else

• Pity is caused by the undeserved bad of someone else

• Malice is caused by the undeserved bad of someone else

• Gloating is caused by the deserved bad of someone else

The third group exists of positive or negative emotions generated by the thought of good or bad things that have happened or will happen. Bad events in the past may also generate positive emotions in the present and good events negative emotions.

The three groups of emotions are induced by beliefs that are held in certainty. There are also emotions that essentially involve beliefs held in the modes of probability or possibility. These emotions are generated by the thought of good or bad things that may or may not happen. These emotions require that the event or state in question be seen as more than merely conceivable. Emotions are also generated by counterfactual thoughts about what might happened or what one might have done.

The role of emotions in generating happiness suggest a summation of happiness in a number. Emotions can’t be added because they can’t be given a objective number nor measure.

The mediating link between emotion and action is that of an action tendency (or action readiness). We may also think of an action tendency as a temporary preference. Each major emotion is associated with such tendencies.

Belief can be affected direct as well as indirectly by emotions. Direct effect produces biased beliefs, the indirect effect low-quality beliefs. The urgency of emotion acts on the gathering of information prior to belief formation rather than on the belief itself. The result is a low-quality belief, based on a less than optimal amount of information. The belief is not biased for or against any particular conclusion that the agent would like to be true.

Chapter 9. Desires and opportunities

Desires are that what is the most important for the agent. Opportunities are the options available to fill this desire. Opportunities can explain the behaviour. They provide a set with options the agent can choose from and therefore give an explanation for when an option is chosen. The choice can be related to the desires. In the end the agent will choose the opportunity that best fulfils his desires. Sometimes there needs to be input from the agents beliefs as well to provide a satisfactory explanation of behaviour. So far, however, the desire-opportunity framework is our focus.

When explaining behaviour we can also list all the possible actions an individual can take and filter them twice to come up with the best option. First we need to look at the constraints the individual is faced with. These are constraints that come from physical, economical, legal and other influences. When all constraints are knows, the options that satisfy all of them are left to choose from. The second filter then examines which option will be chosen. This is done by comparing them to the desires or preferences of the individual.

What if the first filter only produces one opportunity the individual can work with? In society it sometimes seems that there are many cases in which individuals do not have a choice at all. This is in line with those who believe in structuralism. The underlying explanation for this lack of choices is rather unclear. Most logical approaches end up with at east one individual having a choice. For example the fact that the poor do not have ay choice but to work for the rich implies that the rich have the choice whether or not they will let the poor work for them.

In some cases one has the opportunity to perform an action, but not the ability to do so. We all have the opportunity to score a perfect 10 on the exam, but not all of us have the ability to do so. Maybe this can be traced to an earlier stage where one did not have the opportunity to acquire the ability. Inability can come from psycho physiological constraints. We are not all able to perform numerous calculations to determine the optimal choice. Neither can we look into the future to see which option has the best outcome.

Variance in behaviour can be explained by opportunities. This is easily explained by the indifference curves used by micro economists to explain behaviour as well. When prices of one good rise, opportunities for buying it reduce and the budget line shifts. The optimal indifference curve changes and there is variance in behaviour.

This theory even holds for drug addicts. When prices go up, they are no longer able to purchase the same amount of drugs as with the lower prices. Regardless of desire, the opportunity set changes in a way that consumption must go down.

Some economic theory states that all individuals are subject to the same desires, they only differ in terms of opportunities. When desires differ this should be explained through abilities. As some desires require abilities, one can argue that those lacking such desires did not have the opportunity to acquire them. This all to lead to the conclusion that when it comes to explaining behaviour, the opportunity set an individual is faced with is of more explanatory value than is the desire.

Opportunities have the advantage that they are easier to observe. Any individual can observe opportunity sets for others when we know the economical, social and legal constraints. Opportunities are also more influence able. If the constraints are altered, opportunities can be completely different. When it comes to desires there is no quick fix. Human mindsets tend to be hard to change.

Such differences are important for governmental decision making as well. When society faces a problem with the root being the attitude of citizens, the government cannot do much about it. We will no spend our money when we think it will harm us, no matter how much the government tries to convince us that this will boost the economy. There should not be a focus on trying to change our mind set, the government should always try to alter our opportunities as they have much more influence on our actual behaviour.

Suicide. Just wanting to kill yourself will not do the trick. You need to have the opportunity to do so. This explains why there are more suicides among those in the medical profession. They have access to lethal drugs. Self help programs will not be able to beat this strong set of opportunities. Campaigns by the government will have no effect. The only way the suicide rate can be reduces is through making it harder to do it. Just changing bottles into blister packs made a huge difference. Of course, if the will is strong enough, the individual will find a way.

Both Madison and Tocqueville argue that there is a third variable which influences both the desires and the opportunities. Combined they shape the action undertaken. There are four possibilities:

A.

If a regime of direct democracy was introduced, or one of a small representative republic, fractions would be the cause and means of trouble. The common interest will be felt by all members of the majority. As society is small, the majority can oppress the minority because they can meet easily. If society is larger, such opportunities and abilities are smaller as well.

B.

In times of slavery, work is associated with slaves. This makes it nearly impossible for a slave owner to reach wealth, as there is no incentive for him to make an effort and work. The association of work with slavery even affects the desire of a man to make a fortune.

C.

The lack of resources has an effect on both the incentive of taking action as well as the possibilities to do so. One may wish to emigrate from the poverty that surrounds him in his home country. This poverty, however, prevents him from doing so because he cannot afford it. When nature is to lavish there is no desire to take action and where it is too scanty there is no opportunity to change it.

D.

Democracy creates the desire to behave disorderly where the opportunity was provided.

There can be a direct influence that desires and opportunities can have on each other.

E.

Opportunity shapes desire. People tend to want most of what they can get. We have a natural instinct to consume to the maximum level. We are also subject to adaptive preference formation so that we only want the options within our reach. It is a form of dissonance reduction.

F.

Desire shapes opportunities. This can be both intentional and unintentional. Tocqueville wrote: nothing makes for success more than not desiring it too ardently.

Chapter 10. Persons and situations

Human beings all have traits, virtues and vices which make up ones character. Character has an influence on action, whether it is your own character or that of someone else involved in your action. If character would be consistent in all cases and across situations, predictions on behaviour would be easy once we establish a pattern. But is it fair to say that if one cheats on an exam he would not be a trustworthy friend as he will cheat you to?

Believing that character will remain consistent is part of folk psychology. Folk psychology tends to be self fulfilling in some cases. If you feel that your actions will be used to predict future behaviour, you will keep this in mind and behave in a way you want them to predict you will. In other cases the folk psychology is simply not true. The only consistency is found when the effects of the folk psychology as explained above are not controlled for.

There are numerous examples in which behaviour in one case cannot be generalized as to say that the individual will act in the same manner all the time. Parents are often very surprised when their children seem to behave well at school or with others while at home they are disastrous. The people who inflicted electrical shocks in a well known experiment were not sadists in real life. People who are very clean and neat in their own house may not care for this at work.

There is a difference in our behaviour when the situation is no longer the same. One can be very talkative at lunch with a friend, but timid and quiet when at work. Even though they are both personality traits and therefore determinants of character, they are influenced by external factors. Character traits tend to be inconsistent when it comes to cross-situational analysis. There can be correlation, but very low. We just seem to get stuck in the folk psychology. Therefore we believe a doctor to have great medical skills by his good bedside manner. Excellence in one area is a good prediction for excellence in another, non related area, according to this way of thinking. To psychologists this is called the halo effect.

Behaviour can be better related to a situation rather than a person. The main reason for Germans to rescue Jews was not because they has heroic personalities, but that they were asked to do so in a particular situation where their action was to rescue them. It could be because the question revealed that the person in question needed saving. There could also have been the matter of shame which the German would feel if he’d ignore the request. So either altruism or moral reasons could have induced the behaviour. The main influence, however, is the situation the German found himself in.

Situational effects can be even better illustrated by the well known fact that people have the tendency to ignore someone who needs our help when there are other people around. There are various reasons. One may be that we expect the other to help. Another is that we may be afraid to act, as no one is doing it, it could be that interfering is dangerous. No matter the reasoning behind it, the more people surrounding an event, the less likely it is that one will take action and interfere or call for help. The responsibility to intervene is too diverted.

There is no common mechanism by which situations affect behaviour. The only thing you should keep in mind is that behaviour is only as stable as the situation that shapes it. There can be endless inconsistencies in behaviour in only one person, let alone the inconsistencies if we would conduct an excessive research on this subject alone. And the inconsistencies in behaviour of one person affects the inconsistencies of another. If you are in a loving relationship, there is a good chance that you will be much nicer to him or her than to anyone else. Your behaviour will encourage him or her to do the same, even when he or she is having a bad day and hasn’t been nice to anyone else.

We prefer character based explanations to situational based ones when it comes to trying to understand behaviour. It is easier to assume someone is a certain way and behaves accordingly than to look into the particular situation he was in. This is also the case when we accuse someone of being a bad person while in fact he did a bad thing. Telling someone that he or she is selfish is different from stating they are selfish in a particular situation. Attacking the person however, instead of the situation, will not leave much room for positive change.

Using explanations based on the characteristics of an individual instead of the situational factors is referred to as the fundamental attribution error. Sometimes people simply act a certain way because they were asked to do so. Just because you are capable of writing an essay that completely agrees with Hitler and all his viewpoints does not mean you agree with him. You can just fill out an assignment when you were asked to do so.

Western cultures have more difficulty with the fundamental attribution error than do the eastern societies. Asian individuals are more prone to explaining behaviour related to the situation. This is evident even in every day life. A dramatic shooting is reported in western media to be caused by a disturbed and mentally unstable person. In China the same drama was explained through access to guns and dramatizing events in the persons life. The reason for this may be that in Asian cultures the situational factors play a greater roll as it is. They have less bias to overcome than do western cultures.

We should attribute our success to our character traits and our failures to the situation while attributing others success to the situation and their failure to their character traits. This is the best way to be completely self-serving and maintain incentive to perform. We favour the person over the situation. This can be a result from the fact that we have a deeper focus on the moving foreground than that we care for the statistical background. In Asian cultures the focus is nearly similar, leaving a less severe bias to take into account.

Situational factors are not all powerful when we explain behaviour. The personal set of characteristics has to be taken into the mix. The character needs to be decomposed into a set of possible action the individual can undertake in a given situation. Variances of behaviour can be caused by differences in situations as well as differences in response contingencies. When behaviour shows similarities across situations, there may be a case where these two types of factors are cancelling each other out.

Chapter 11. Rational Choice

Many economic theories are based on the assumption that consumers act rational. There is no need for a distinction between choices based on desires or objectives. The individual will choose the option that best fits his desires. There are three optimality requirements:

• The action undertaken must be optimal given the beliefs of the individual.

• The beliefs must be well supported by the evidence.

• The evidence comes from thorough analysis from the information.

Actions should be caused by desire and belief. This makes the action rational. In some cases the right action is taken by mistake. Sometimes this is caused by two distinctively biased processes. Even though the outcome is the same as if it was through rational choice, there is no such thing in these cases. A good example is when one would walk down a forest path and see a stick. As there is a chance this stick is a snake, the first reaction is to run away. This would also be the rational choice if it were a snake, but, as there was no information gathering to make sure of it being a snake, there is no rational choice here.

Structure of rational choice explanation:

The optimal choice is that what seems best to the individual. There is no other option that would better fit the desires and beliefs this individual is subject to. There is no need for this optimum to be stable. The optimal choice depends on the situational factors that occur at the time the rational choice is to be made. In another time the preferences may have shifted completely.

To make sure the choice is the best, the preferences have to be transitive. If A is preferred to B and B preferred to C, A is preferred to C as well. When there is no transitivity, the best choice may not be available. Lack of transitivity makes it possible for another to exploit the individual as well, when money is offered to shift the choice to another option. Transitivity can arise when the individual ranks his preferences by counting aspects. Each option receives point for specific characteristics. It is possible for two options to receive an equal score.

Indifference can be transitive as well. One can be indifferent between A and B and indifferent between B and C. But there may be a difference between A and C, which makes the individual prefer A to C. This is not transitivity. The optimal choice may not be reached as this individual will easily be persuaded to choose an option less optimal than A. As she is indifferent between A and B, she will have no real objection when offered the choice B.

Preferences should be complete in order for the best choice to be the optimal one. There should be clarity why one option is preferred to another or why the subject is indifferent. If these statements cannot be made, the individual may not be able to rank his preferences. If this is the case it is not a failure. In some cases it is simply impossible to rank preferences and determine why they are ranked as such.

We say that preferences are continues if, when two goods are combined, no amount of the second good would make up for any loss of the first component. This continuity is not really something that applies to consumption, but it is one that is evident when it comes to politics. We vote for a candidate based on certain characteristics. If another would have these exact same characteristics we could look beyond them at the other points they make. But we would never consider another if he would not have this exact set of characteristics to start with.

When we have reached completeness, transitivity and continuity, we can establish a continuous utility function. This can be used to maximize utility. A is preferred to B if . This does not imply achieving as much as possible of physical things, there needs to be a maximum utility which is different for each and every individual.

Options available are often subject to risk as well. The most logical approach for solving such problems would be to calculate the expected utility that it attached to each option, according to the weight and utility of the possible outcomes, and choose the highest level of expected utility. The problem is that ordinal utility does not allow for this. If we consider a case with two options and assign utility levels to them, it could look something like below.

Preferences are set out equally well using either u or u=v^2, but they come up with contradicting results. This makes the approach useless. More effective measurements come from a cardinal point of view. There may not be much to say about the exact level of utility, but one can make inferences based on differences between levels of utility. We shift out focus to decreasing or increasing marginal levels of utility.

Each individual faces a set op options as well as probabilities assigned to obtaining those options. Preferences over these options are assumed to be transitive and complete. Preferences are also assumed to be independent when combined in the same way to an external factor. Continuity comes into play as well. P(C) is the probability at which one would be indifferent between receiving C for certain or a combination of P(C)*A + 1-P(C)*B where A is the better option and B the worst. U(C) is P(C), the cardinal utility.

If one option has a higher expected utility than another, this option will be preferred, no matter how the function is rewritten. Using this approach will allow an individual to maximize utility.

Cardinal utility is linear in probability. AE has a higher level of utility than does AC with 50% probability. Do not forget that with cardinal utility the risk aversion or preference is already represented when calculating the pay offs. Trying to control for them would be double counting and therefore not correct.

Intrinsic utility of some good or experience is very important as well. We can immensely enjoy something, but sometimes it is only satisfying. In other cases an experience or good can be mildly annoying or even dreadful. There is no reliable way to assign numbers to these experiences, but that does not make it less valuable when looking at rational choices. Many goods have decreasing marginal utility. The best-known example for this is income, where a poor person will derive much more marginal utility from a few extra dollars than would a millionaire.

In some cases the consumption of one good has two effects: Both the enjoyment of consuming as well as the reduction of enjoyment of future consumption. Food tastes the best when you eat it while you are quite hungry. Below an illustration of increasing marginal (intrinsic) utility.

Each extra unit of consumption is more valuable than was the one before. Cardinal utility functions are generated through risk attitudes and intrinsic utility. They cannot be measured separately, but it is possible to assign more weight to one factor due to intuitive reasoning.

So far we have focused on the first part of rational choice: choosing the best way to realize desire, given the individuals beliefs. Rational beliefs are shaped by processing all evidence available and then deciding on that what it most likely to be true. Bayesian learning identifies the process of belief revision as we find new evidence that has effects on our initial beliefs. Bayes gave us a frame in which we can identify probabilities, given the fact of some other factor being present:

Bayes’ theorem was introduced because Bayes believed that our beliefs are all based on what we already know. Over time, however, new evidence can change these beliefs completely. Each new piece of information has a smaller impact than the previous one.

Gathering information has an effect on both the beliefs and desires that make up rational choices. There is a price tag involved in this process, both the form of time and money. There needs to be an optimal level of information, given the costs of acquiring it and the benefits of having it. Determining this optimum depends on the individuals desire to obtain information. Another factor is the seriousness of the choice. In the case of buying a house information is much more valuable than when there is a choice between two kinds of bread.

The optimum is affected by the results that information gathering carries with it as well. The more valuable information is collected, the more one is attracted to looking for more. On the other hand, when there is enough information available, one can decide to have enough and not need to look for more. The same logic can be applied when the amount of information found is only little.

Rational choice theory is subject to limitations. When it comes to explanation, failure is caused by the fact that it cannot predict when people in a given situation will do. When predictions are stated, people may fail to follow them. This is indeterminacy. There are different reasons for this failure to occur. First of all, an individual may not identify the best option because he is indifferent between either of them. Preferences may be incomplete as well. See the figure below:

It is clear that this individual prefers A to C and C to E as well as B to C and B to D and D to E. In this case there is no clear preferential ordering between C and D. But that does not really matter, as they are both preferred to E and not as preferred to A or B. The problem lies with A and B, which are not comparable in this set. How to know if gathering information is worth it? Between the lower and upper bounds of effort and resources spent on information gathering, there is an interval of indeterminacy. Within this interval, the agent is rational.

Rational choice is subjective. Trying to explain it after the fact is very difficult, as we can only fully explain when we do not know the outcome and only the situations, preferences, beliefs and desires before the choice was made. The choices made can only be characterized as useful when we look at them from the point of view of the one making the decision.

Rationality of beliefs is not the same as judging them as being true or not. Truth relates to belief and the world. Rationality relates to belief and the evidence that is available to the agent. Investment in new information and therefore new evidence that could bring rationality and truth closer together is dependent on the expected costs and benefits the agent faces based on his beliefs. Some beliefs must be accepted to be true, as testing them would mean no one would have time to move on with their lives.

Chapter 12. Rationality and Behaviour

Rationality is something we want to have. Rationality is an explanation for why we behave the way we do. This desire for rationality can be strong enough to make us behave rational. Hyper rationality is when we want to achieve the optimal decision without taking costs into account that need to be incurred for the decision making process.

Costs are:

1. costs of means of deciding

2. costs of side effects of deciding

3. opportunity costs

Examples of hyper rationality are easier to come by than you think. Someone who travels across town to save the discount on a bargain while the saving is less than the price for gasoline. The effects of a custody battle on a child are clearly neglected in heavy trial cases. Opportunity costs occur even in the time you are thinking about your decision. Someone else would have been doing both options you are debating about in that time and saved a lot of trouble. Opportunity costs arise in the first example as well, bargain shopping can effectively be a waste of time.

Behaviour is subject to patterns that are not rational. To further explain rationality, I will now introduce some canons that are the fundamental principles of rationality. The following only applies to rationality when it comes to individual choices:

• A rational agent will only act if the expected benefits exceed expected costs.

• When the choice is between two evils, a rational agent will choose the least evil option.

• Opportunity costs and direct costs are weighed equally.

• There will never be preference for a subset instead of the full spectrum op options.

• X preferred to a glass half full means that X is preferred to a glass half empty as well.

• Previous results will not be consulted when it comes to decisions that involve gambling.

• When a project is started, choosing to continue will only depend on future cash flows and therefore not on sunk costs.

• If a choice is made in period 1, this choice will remain the same in period 2, unless desires and beliefs have changed.

• Rational agents will choose based on the expected value, not on best -or worst case scenario’s, when there is risk involved.

• Rate of return on all assets should be approximately the same.

• A choice made from the full spectrum of choices will not be altered when there is only a subset available to the agent.

• There is no action to an effect in order to suppress the cause.

• If X is preferred to pY, qX will be preferred to pqY. (Independence Axiom)

• If an agent will do X when C occurs and if he will do X if C does not occur, he will do C no matter the situational factors if he is acting rationally.

• An offer that could reveal a negative expected value upon acceptance will never be made by a rational agent.

• Vengeance will only be made at the point where there is maximum success or minimal personal risk, no matter how long the rational agent needs to wait.

• Rational agents take fencing lessons when challenged to a duel.

• Rational agents conduct research on behaviour and emotion before proposing an important question.

Even though everyone reading these principals will only agree with them, they are quite often violated by agents, who believe themselves to act perfectly rational. Your vote will not have a significant effect on the outcome, still a large number of people goes out to vote in elections. The costs exceed the benefits in this case. Another example is the fact that we are willing to pay someone for doing a job, but would not do the same job for even more payment. Gamblers fallacy incurs quite often as well, where a gambler expects a certain outcome based on previous outcomes, while each trial is independent.

In many businesses the sunk costs are still considered when one is thinking of terminating a project. Even though the expected cash flows have a negative present value, the fact that there is too much vested in the project as it is already will persuade many ‘rational thinkers’ to continue the project. In business the effect of irrelevant alternatives is evident as well. When there is difficulty choosing between two options, adding a third, inferior one will make it easier to see the best way to go. Rational thinking would not need this irrelevant option.

There are numerous examples in both business and individual decision-making, some in real life situations but others in experiments. Many agents making these contradicting decisions are ones who have a good degree and are expected to have the intellectual capacity to make rational decisions. There is no clear explanation for these repetitive violations. As we are eager to explain our own behaviour, however, we explain these puzzles through alternatives on rational choice theory. These alternatives take into account some factors human beings are subject to, which have an effect on rational thinking.

• Non-probabilistic weighing of outcomes. This factor is evident when there is a

different weighing of outcomes than would be if the expected utility theory were upheld. Agents are more sensitive to probabilities, which are close to 0 or 1. The probabilities are not linear as is assumed in expected utilities. One would be willing to pay much more when the chance of failure is reduces from 1/6 to 0 than they would pay for a reduction from 4/6 to 3/6.

• Hyperbolic discounting. This explains why people cannot handle the freedom of

holding their savings in a normal account and prefer to keep them in a special account where they cannot withdraw anything while they do not receive any compensation for this loss of freedom.

• Heuristics. These are rules of thumb. Heuristics lead to gamblers fallacies when they

believe that trials are not independent. The fact that red has come up in the previous four trials can make them to believe that red is more likely to come up again as well as black coming up since it is blacks turn. A peak-end heuristic makes people prefer a worse scenario because it has the same peak but a better end.

• Wishful thinking. We tend to play down the risks we face if the end result is something

we want really badly. This goes for people in a risky occupation as well as cancer patients facing chemotherapy. Even though the suffering is very hard and we may have had higher utility if we’d have chosen not to take the risk, the end result is so desirable that we tell ourselves its worth it by hoping the chances for success are higher than they actually are.

• Inability to project. Sometimes we are not capable of seeing what will happen in the

future based on decisions we make today. This makes us choose an option that in the end will prove to be less favourable than any other choice we could have made. The winners curse can be explained by this. The winners curse comes into play when there is notion that another party has information that you do not. Chirac’s choice to call for early elections made the French voters vote against him as they believed he called early elections to hide something unknown to them.

• Desire to act for a reason. This factor makes is harder to choose when there are more

options to choose from. As the amount of options increases, there is less certainty when trying to convince yourself that a particular option is the best one. If you need to be sure of your choice, you will decide not to choose anything if you cannot convince yourself of this. Stand with smaller variety of product will therefore sell more than will stands with an endless amount of options. On the other hand, adding one option when indecisive between two close possibilities may increase the likeliness to make a choice.

• Categorical imperative. When we make a choice we usually base it on the best option

for ourselves. The categorical imperative is something that happens when we do not consider only our own best interest, but we consider an option as one taken by everyone else as well. If we all choose this option it may be better for all of us. A good example for this is a national election. This, however, is irrational.

• Emotions. Emotions affect action directly, when there is weakness of will. Desires are

affected in two ways. The first way is by association with the action effect. The second way is through time discounting. The discount rate may be increased, which means that a less preferred option with long term consequences which are not so good could now be one that is preferred over any other. Beliefs are affected through wishful thinking. There is a close relation to pride. And as emotions tend to be very urgent, they have an effect on the optimal acquisition of information as well. This all is illustrated in the figure below.

• Social norms. Many emotions are related to sustaining social norms. Shame, pride

and anger can be induced by conforming to these norms. They may hinder the capability of rational thinking and therefore rational decision-making. Social norms may also prevent rational choices to even be considered as an option as it is simply not done. Take for example the fact that many people do not even consider the cheaper stores when looking for towels, while the prices are better for the same quality material. Avoiding the shame of being caught there while the social circle tells you not to be worth paying extra for an accepted retailer.

Chapter 13. Responding to Irrationailty

Agents want to be rational. Precautionary strategies should be distinguished from experience (learning). Precautionary strategies are second-best rationality and experience occurs when the propensity fades away as a result of improved insight. Agents can use intrapsychic strategies or extra psychic devices as a solution to cope with irrational behaviour.

A sophisticated agent is an agent who knows he is subjected to hyperbolic discounting. One form or irrationality (magical thinking) can offset another form of irrationality (hyperbolic discounting). For this offset a good framework of binary choices has to be made: always doing it or never doing it.

Another case can be when the agent confronts the choice between rewards at several future dates. The agent may then adopt the intrapsychic device of responding strategically to the know propensity of ‘future selves’ to discount the future hyperbolically. The danger lies here that if an agent predicts that he will deviate from your best plan, he may end up deviating even more that without his awareness.

Bunching strategies and precommitment devices involves affecting the external world for the goal to make it les likely to choose the earlier smaller reward. Five strategies come to place with this:

• Eliminating the choice of the earlier, smaller reward

• Imposing a penalty on the earlier choice

• Adding a premium for the choice of the delayed reward

• Imposing a delay between the choice and the actual delivery of the reward

• Avoiding cues that might trigger preference reversal

These precommitments can solve two kinds of temptation. First there is procrastination and second premature gratification. These two temptations stem directly from hyperbolic discounting. Merely resolving not to take a ction is less likely effective given ‘the power of the situation’.

Chapter 14. Some Implications for Textual Interpretation

The scientific enterprise has three distinct parts or branches; the humanities, the social sciences, and the natural sciences. Interpretion of a work of art and explanation are closely related enterprises. Understanding a work of art is to explain it in terms of the antecedent mental states of its creator. Dealing with the internal development of the work of art, intelligibility rather than rationality is the most useful idea for the task of interpretation.

Authors often make many drafts is irrefutable evidence that they are engaged in a process of choice and that they possess explicit or implicit criteria for betterness. These drafts involve small variations suggests that they are aiming at a local maximum. The difference between an author and someone who is merely climbing along a gradient is that the former’s creativity goes beyond mere choice. Work of literature cannot be reduced to rational choice is that the number of meaningful word sequences is too large to scan and select ‘the best’. A rational creator may try to make the problem more tractable by deliberately excluding some sequences. The author relies on his or her unconscious associative machinery. Rational creation is largely about getting precise or shift to the next metaphor. Here are some demands that rationality imposes on the author:

• The acts and utterances of the characters have to be intelligible; Intelligibility can be absolute or relative, and if relative, global or local. The question of absolute intelligibility is whether any human being could behave in this way. The question of relative global intelligibility is whether the behaviour of a frictional person is consistent with his or her overall character as displayed earlier in the work.

• The author has to meet the twin requirements of fullness and parsimony; The ideals of these requirements are best expressed by a passage from Aristotle; ‘the story … must represent one action, a complete whole, with its several incidents so closely connected that the transposition or withdrawal of any one of them will disjoint and dislocate the whole. For that which makes no perceptible difference by its presence or absence is no real part of the world.’ By reading this passage the thought comes up that the author has represented all the information necessary to understand the plot. But not all the information is necessary to understand the plot.

• The work has to flow downhill, in the sense of minimizing the appeal to accidents and coincidences; Good plots should not turn on unlikely events, accidents, and coincidences.

• It has to offer a psychologically gratifying pattern of the build up and resolution of tension; The overall plausibility of a scenario depends much more on the plausibility of its weakest links than on the number of total links. Authors should respect this particular quirk of the readers.

Interpretation of a work of literature is a claim important features of the work can be traced back to decisions that the author made for the purpose of enhancing the aesthetic value of the experience that some specific audience could be expected to derive from the work. Authorial failures are intelligible. Authors receive a double pressure. First one is to make the plot move on, and to do so through intelligible actions and statements by the characters.

Chapter 15. Physiology and neuroscience

Emotions are related to beliefs. If we believe someone will try to harm us intentionally we will respond with the emotion of anger of sadness. If we believe this process is unintentionally and not to be blamed on the person in question, there will be no emotional response towards him or her. Sometimes, however, emotions are triggered by mere perceptions, which are not the same as beliefs. Experiments that test for fear in the neurophysiologic science confirm this. When the emotion is caused by belief, it is combined with the thinking part of our brain. When the emotion is caused by perception, there is no thinking involved and the emotional reaction appears much faster.

Science also gives us the opportunity to distinguish between different motivations when it comes to punishment. The majority of people will punish because they feel the subject of punishment deserves it. Only sadists will punish for the sheer enjoyment of another’s suffering. Such analysis can be done for altruistic behaviour as well. Only a small amount of people will do so for the sheer joy of giving. Others will do so because they feel the subject deserved it. Some will do it to feel better about themselves.

Inflicting punishment yourself has a different effect than when you are an observer when someone else inflicts punishment on another. If you feel the punishment is not fair, you are likely to feel the pain of the punishment yourself. If you feel the punishment is fair, the areas in the brain that respond to this are the ones which induce pleasure. It is a difference between ‘I feel your pain’ and ‘you had it coming, good for you’. Women are more sensitive to the pain.

Our brain can fill in gaps in observations. Look at the figure below. In the first part we can see the circles underneath the rectangulars, because we assume the rectangulars to cover the circles. In the second part, only the parts of circles remain and we cannot ‘see’ them anymore. They now look like triangulars touching at the point.

Filling in the gaps is part of daily occupation of the brain. Just looking at a somewhat familiar word can lead to filling in unconsciously, while reading texts with more caution would reveal something completely different. This is why you need to take your time reading questions in exams, since our brain will spot the familiar parts and fill in the rest. Motivated misperception is when the brain is making the best guess based on our observations. This is better known as jumping to conclusions.

Chapter 16. Explanation by consequences and natural selection

Events are explained by its cause. Causes occur before the event. Therefore an event cannot be explained by its consequences. The only way consequences can have an explaining effect is if there is a pattern of recurrent behaviour. The way this is done is through reinforcement or selection. Reinforcement is pretty straightforward. When doing something comes with a reward we are more likely to do them again. When the consequence is punishment or disappointment, the likeliness of doing it again is smaller. This can happen without intentionally choosing it.

Knowing the effects or reward systems, one can use them to induce the preferred pattern of behaviour. Numerous experiments have been constructed to find out which pattern of rewards induces which pattern of behaviour. This knowledge is of great value to parents as well as casino owners or teachers. Con men also rely on this kind of reward system. This kind of manipulation only works if the reward system is unknown to the target. Therefore, clearly announced incentives that will be paid out when hitting certain targets do not induce the preferred behaviour in the same way.

Some believe agents to adapt to their environment. You could also think that agents are selected by the environment. Many selection mechanisms do not occur by a rational choice mechanism. This means there is some kind of natural selection. Natural selection optimises the agents available, makes sure the strongest survive and therefore that the entire species becomes stronger. This is, however, different from social selection. Social selection does not optimise.

The phenotype of an organism is its fitness, the number of its offspring for instance. The genotype of an organism is the set of instructions that induce the phenotype. This is due to DNA and environment. The DNA can be related to any common written language. The letters are in codons, the molecules that build any organism. A gene is a segment of the DNA. Early organisms appeared as the result of random mutation of the genome. Some theory thinks that these mutations occurred due to errors made in the mutation process. There are different types of mutations.

The figure below shows the resemblance between DNA mutations and the building of a sentence. Errors in print are made when there are mistakes in typing. There can be many reasons for these mistakes, but one of them is when a letter is replaced for another in an incorrect way. When there is supposed to be a d to make up dots, changing it to pots will change the whole meaning of the sentence.

When DNA mutations are replaced incorrectly, the entire specie can change and therefore evolve into something else. There can even be a complete lack or relation to the original. There can be different reasons for mistakes to be made in typing. A reason could be that there is no clear distinction between m and n. maybe the writer is under the influence of alcohol. Just like with sentences, the error made in DNA mutation can be caused by a mutagenic agent.

Just like with words, there are different steps to take when changing one DNA mixture to another. There are four steps to go from HAIR to DEAN ( HAIR – HEIR – HEAR – DEAR – DEAN ). In DNA there would need to be four mutations as well to create such a change. As you can see, the latter is nothing like the original. Some changes are, however, not possible. Going from HAIR to LYNX would not be possible with the words in the dictionary,

Natural selection has an effect as well. The environment an organism is in will reach a local maximum due to natural selection. Any other mutation in the DNA would lead to a less-fit organism. There may be higher peaks in the nearby landscape, but for this local area the maximum has been reached. This is not the process as described in rational choice. The figure below is an illustration of the natural selection process:

Human beings are capable of using indirect strategies, wait and be aiming ahead of a moving target to reach their goals. Natural selection fails already at the first part, the use of indirect strategies. There can be movement from A to B, from A to C and not from B to C. A change in the population from A to B will leave it stuck at that level. Even though it would be better to wait for an opportunity to change to C, natural selection is incapable of this and will jump to any opportunity to change.

Natural selection cannot adapt ahead of environmental changes. Even though adaptations occur fast, natural selection can only make sure that organisms can effectively keep up with the changes in their environment. Humans can anticipate changes and prepare themselves for it. This is where rational choice and natural selection differs completely. We can rationally choose to take an umbrella with us, because we know there is a good chance there will be rain by the time we return home. If we anticipate something we really do not want to happen we even have the possibility in some cases to prevent it.

Both mutation and the lack of mutation can lead to extinction. If the environment changes too quickly, the ones that evolve the fastest and therefore have the most optimal adaptation will be most likely to survive and therefore determine the new face of the population. On the other hand, such adaptations can induce the loss of valuable information that can limit the survival possibilities. There is an optimal rate in between. Natural selection here leads to getting stuck in the lower optimum as well. This is illustrated in the figure above.

To reach the global maximum depends on the relative speed op the processes illustrated in the figures above. These are the rate at which the inferior varieties extinct as well as the rate at which favourable mutations appear. There is no mechanism, however, that could mimic the capacity of intentional beings to anticipate, wait and use indirect strategies. There might be a lag when the change in the population is not fast enough.

Pleiotropy is when natural selection favours certain mutations that are suboptimal when they are in isolation, but which are optimal when they are a package solution. This is when a gene has more than one effect on a phenotype. When positive effects outweigh the negative ones, the gene will be favoured through natural selection. The harmful effects of male insects on their mates when copulating are outweighed by the advantage it has on the reproduction.

Natural selection favours the individual, not the species nor the population. When there is a mutation, which is beneficial to the individual, it will cause this individual to survive. In some cases, however, the mutation is not beneficial to the population at all. Natural selection does not really care for this. There are some exceptions to that rule. The first one is kin selection. This is when the gene is the unit of selection and not the individual organism. In some cases the organism will sacrifice himself to make sure the gene survives in those surrounding him.

Group selection is a form of supra-individualism. If organisms in a population have a gene that prevents them from the tragedy of the commons, they may be out-produced by less inhibited organisms that lack the gene. Group selection and kin selection provide the basis for cooperative behaviour. Reciprocal altruism is a form of this as well. When there is enough interaction there can even be the expectation of punishment, which will affect the behaviour of the individual towards another.

Chapter 17. Selection and Human Behaviour

Any selection mechanism needs raw material to work on. The source of variation can be intentional or non intentional. But how much of human behaviour is by choice?

Human behaviour is subject to the fact that we can create opportunities for action as well as restrain ourselves from it. Furthermore, we have evolved as emotional beings, emotions guide our ways and therefore our decisions. These emotions have an underlying natural selection as well. As men are less likely to be sure their children are actually theirs, they are more likely to be jealous then women. Children who are brought up by one or more parents other than their biological ones are more likely to suffer from abuse.

Emotional reactions also induce the tendency to punish another for any kind of wrongdoing at the expense of the punisher as well. Emotions create rules of conduct, where it is more accepted for a man to commit adultery than it is for a woman, while adultery as a whole is still frowned upon. Natural selection would encourage us not to commit to only one partner, as this way it would truly be the survival of the fittest. The same goes for incest, which is not accepted and less common when the siblings actually grow up together.

Humans are very well evolved in the act of self-deception. This may have evolutionary benefits. Self-deception makes it possible for us to deceive another as well. The problem here is that it causes false beliefs and this can be quite disastrous for behaviour. Therefore there is no clarity on the reliability of the claim that self-deception would actually benefit the evolvement of human behaviour. The same goes for the argument that depression evolved as a bargaining tool, somewhat similar to a strike.

Non-intentional variation and intentional selection is when a new organism arises by accident, but is preferred over others by choice. A good example here is the fact that in China and India parents prefer to give birth to a boy. If by accident the infant turns out to be a girl, the parent may intentionally choose not to carry the pregnancy to term. This causes the balance of male and female in society to be unequal and leads to all sorts of problems in the long run. Another illustration is when mentally ill or mentally retarded individuals are not allowed to reproduce. One day we may be able to determine the sex and mental or physical state of a foetus at the moment of conception and at that point even the variation will be intentional.

Intentional variance and non-intention selection are very common in economical markets. There are some similarities with natural selection. Efficiency is assumed as those who are not efficient will not be able to survive and will extinct in the long run. This is a strong basis for the assumption of efficiency, as managers are not capable of doing all the calculations necessary to actually determine the optimal level of efficiency. Their rationale is simply too restricted.

Firms all try to increase profits. We can assume that the majority of firms do this through innovation and imitation. When this leads to production at a lower cost level, the company is able to outperform competitors and maybe even drive them out of business. They can only fight back through increasing efficiency. As innovation and imitation are small steps to take, the process, in a constant environment, will lead to a local maximum. In this local maximum there will be equilibrium profits.

If we let go of the assumption of constant innovation and imitation, we can say that companies do their work in a routine fashion, sticking to processes and rules of thumb to maintain a satisfactory level. This does not need to be the optimal level. Statements as ‘never change a winning team’ reflect this line of thought. Until there is a threat of a hostile take over or a strong decline in profits, these companies will not change the way that they do business. In this setting a process of imitation and innovation will only come into the picture when the environment changes and income is below the satisfactory level.

What is the rate at which inefficient firms are eliminated compared to the rate of change in the environment? The answer to that question determines the usefulness of any kind of model used to explain behaviour like the ones above. Firms face unprecedented rates of change in the markets they operate in today. To maintain or achieve success, firms have to broaden their scope to a level that anticipates change and therefore creates the possibility to act in advance. When size is sufficient, some firms can even shape the climate they are in, just because they have the power to influence it.

Chapter 18. Unintended consequences

Thing do not always turn out the way we planned. Sometimes we face consequences that are trivial. This is when we unintentionally hit the delete button and loose all the work we did. Other times the consequences take the form of more schematic mechanisms. There are cases in which consequences are not only unintended, but also completely unexpected as they were not even foreseen.

Unintended consequences can be caused by the behaviour of one individual as well as the interaction between more individuals forming a social group. When the individual causes the unintentional consequences, this can be explained through the beliefs and desires combined with opportunities that make up the individuals behaviour. Actions can, in turn, shape desires, beliefs and opportunities as well. A simple example of this is that you can be determined only to drink two drinks at a party, but after doing so your state of mind is influenced by the alcohol in those drinks and you walk home singing.

Choosing an option at one point in time will eliminate some of the options that are available in the future. In some cases we can see the limitations that arise when choosing that option. Other times the irreversible consequences may not be anticipated just yet. The reason for this is cognitive deficit. This is when an agent cannot predict future consequences. When motivational deficit is at work the agent attaches low weight to the future consequences as they come with present gains.

When the unintended consequences are caused by the interaction of individual agents, we can often speak of externalities. Externalities of behaviour may aggregate into outcomes that have never been foreseen by any of the agents involved in the process. All will act in service of their own best interest, but carry unintended side effects with their actions as well. These side effects can be both positive and negative for the agents surrounding him. As all agents are subject to many of these externalities, they form much larger consequences.

Positive externalities are the basis of many economical theories. As all agents act in self-interest we choose the most optimal goods that are made by the most efficient companies, which increase profits through increasing efficiency, etc. This way the positive externalities we all force upon each other create theories like the invisible hand. Weak negative externalities are when the benefit is less due to the costs imposed on others. Examples of this are congestion on the road while it is beneficial to all to be in the car.

Strong negative externalities make all agents worse off as one is trying to improve his personal standards. This is a doom scenario the socialists proclaimed about capitalism. The idea here is that when costs are cut by laying off labour there can be a gain in profit. If, however, all firms would be letting go of large portions of their employees, income levels would be incredibly low. The market would not be able to absorb all the products offered for sale and profits would decline even further than before the layoffs.

Internalities are the benefit of harm that an individual places on himself by choosing a certain path of choice. They are externalities on the same persons later self. Time spent with children imposes a positive internality, while addiction comes with negative internalities. There sometimes is not even need for addiction. Some goods loose their level of satisfaction with every consumption. Listening to a great song on the radio can be very pleasurable at first, but after hearing it ten times a day you may not even be able to stand the song anymore.

The figure above illustrates the younger sibling mechanism. This happens when we all think we are the only one with the capability of acting rationally and making the most profitable decision. Everyone else does not have the same information and will not face the exact same optimum. That is why, when we expect p1, we produce q2. But when we flood the market with q2 products, we will have to sell them for only p2 if we want to sell them at all. For p2 we only want to produce q3. Producing q3 will lead to a shortage and therefore a price increase to p3. This leads to another increase in production and a cobweb of quantities and prices arises.

If we all assume only ourselves to be rational and capable to make the decisions we make, we are subject to the younger sibling mechanism. Even though the optimum is reachable, we fail to do so as we do not expect others to think like we do. This happens in daily life more than you think. Building new roads leads to adaptation in the form of even more traffic. Unemployment laws can backfire in making job security even worse. The only way we can come to the level at which it is better off for all of us is if we try to cooperate and find the optimal level together.

Chapter 19. Strategic Interaction

Interdependent interaction is what’s being explained by the game theory. It seemed rather puzzling, the behaviour that is present here. But when you see it through the game theory, it all makes sense. We have a number of players who all have a number of options. These options satisfy the conditions of representing preferences as cardinal utilities and can be ordered according to their preference. The players can communicate, but cannot enter into any binding agreement. The reward structure assigns a utility to a strategy. The psychological reward is the one that is important here.

In some cases there is clarity on what an agent will do. The outcome of this choice may be dependent on the choices of others, but the chosen strategy is clear and independent. Other cases have clear interdependence. There is equilibrium in this interdependence if none of the players can make a choice that would make the situation better for him, without the cooperation of the other players. So as all players choose their equilibrium strategy, none of them can do any better, given the choices of the others. It gets tricky when there is more than one equilibrium in one game.

For now, however, we will assume that all agents have access to the same information and there is only one equilibrium. The players are rational and will in the end choose their option so that equilibrium will be reached. They will rationally infer what the others will do according to their preferences. In many such cases there is a dominant strategy. This dominant strategy will make sure the players choose with the aim to reach the equilibrium. Most of the times both players will have dominant strategies. They do require more access to knowledge, however, so that one can decide on what to do as he knows the dominant strategy of the other.

There is a special class of games in which none of the players have an incentive to change their course of action, but they also prefer that no other player does so. These are called coordination equilibriums. A good example of this is traffic. We all agree on which side of the road we are supposed to drive. Changing this by yourself will cause an accident, but if another would decide to deviate from the agreement there would be an accident as well. There is no unique equilibrium, we just agree to do the same thing. Another good example is words. We all agree on their meaning. They are somewhat like conventions.

There are games that have no relevance to dominant strategies. Duopoly behaviour is such a game.

If firm 1 produces a low amount, firm 2 will have an incentive to produce more as the prices rise due to the shortage. Each firm has a production function that is dependent on the output of the other. Equilibrium is the best response of each firm, taking the output of the other as a given. They could do better if they form a cartel, but as this is illegal the next best thing is as it is illustrated above. These firms face a prisoner’s dilemma. Duopoly games also explain why ice cream vendors are most likely to be next to each other in the middle of the beach and why political parties have agenda’s that are close to the middle of societies preferences. It is the best possible response to what the other players are doing.

Frequently played games are ones that offer the choice between cooperation and defection. Social interaction is based on cooperation and coordination. If there would be no cooperation for mutual benefit possible, society would be a very nasty place where each has to fend for his own. This is not what we see around us most of the time. If coordination would not be possible, we would not be able to minimize the unintended consequences. Game theory is about the success and failure of the combination of coordination and cooperation. One of the most commonly used examples is the prisoner’s dilemma.

Two prisoners have committed the same crime and face the following options. While separated they have to choose whether to remain quiet and let the police guess their way with all the other evidence, or to blame the other and try to go free themselves. They know that if one tells and the other does not, the one who is blamed will go to prison for four years and the one who told will go free. If they both tell on each other they both face a prison time of 3 years. If they both remain quiet, they will have enough evidence against them to spend 1.5 years in prison. Both would be better off if neither would try and let the other be punished for the crime. However, the combination of fearing to get caught and spend 4 years by yourself as well as the temptation to go free will make them both speak up.

In some games, there is more dependence on the believed preference of the other than on actual fact. As they both assume the worst of the other, the simple fact that then they both act as if the other has the preference of the worst option will only confirm this belief. Such games happen a lot in many societies. If you believe no one pays their taxes, you will not pay them as well. As we all face these beliefs we then turn it into reality. If one nation thinks that the countries surrounding him are not taking the necessary precautions against some infectious disease, there is no incentive for it to try and eliminate it all by itself. As the surrounding nations see this country fail to take the precautions they will refuse to do so themselves, confirming the initial beliefs. This game is called a Stag Hunt Game.

Assurance games are of a different class. They exist when the payoff is something more than just material rewards. We cooperate if we are assured that others will as well. There is a strong desire to be fair and reluctance toward free riding. Such games can also exist if there is a penalty for not cooperating. Reversing this means that when a reward is attached to cooperation it will be more likely that we are in an assurance game than in a prisoners dilemma. Unionised workers may have higher wages, but can also benefit from pension plans and other offers that only members can access as the higher wages apply to non-unionised workers as well.

The game of the chicken is based on the story where two kids steal cars and drive them off a cliff. The one who stops first is chicken. The idea is to stop at the very least moment, it is not a suicide attempt. It can also be played when two cars drive towards each other and pull away at the very last second. Equilibrium is reached when they both do the same thing. Which of that is chosen is undetermined. In real life this situation has occurred in the arms race between the Soviets and Cuba on the one hand and the US on the other. In the matter of setting a standard it is much like the game of the chicken as well. Standards are reached in political sense as well as in commercial situations.

When it comes to setting a standard, there is the matter of coordination as well. We all prefer there to be a standard. In the recent case of Blue Ray vs. HD, there was no clarity for producers as well as consumers what would be the best option to invest in until there was a standard. No one wanted to buy the wrong option, or bring out a line of HD players while Blue Ray turned out to be the standard in the market. In politics we all prefer a society with structure, rules and regulations that set out the standard. The question here is which standard is the best. As we all prefer having a standard to not have one, we also prefer certain standards to others. This is called the battle of the sexes.

A variant to the battle of the sexes is the focal point game. Here there is a general consensus on some issue in which there is really only one unique option that would be the best. A good example of this is having a meeting point in a central station, all who have to meet up with someone know where to go. There are not two meeting points that would make it confusing. Deciding not to use nuclear weapons is also a unique focal point, as there is no good standard to come up with if there would be ‘limited use’. Democracy is a focal point game. The majority rule is the best way to come up with consensus on a standard. The focal point game allows for cooperation through coordination.

So far we have assumed the choices to be made simultaneously by all agents. They may be able to anticipate the others strategy, as well as logically infer it, but there is no possible reaction after the fact that the other has already chosen his option. In many cases, however, there is such a sequence. First one player acts and then the other reacts. Both face different payoffs, which lead to a different kind of game theory. The choices are made based on backward induction. To illustrate, consider the example of two generals facing each other at war. If the first general decides not to do anything, the payoff will be 3,3. When however, he decides to attack, the other faces the choice to either fight back or surrender. The situation is graphically set out below:

The second general can try to send out a threat that when attacked he will fight back. But such a threat needs to be credible. If the first general knows the payoffs facing the second one, his threat is not credible at all. The first general knows that surrendering is the dominant strategy for general 2, as this yield a payoff of 2 against a payoff of 1. Knowing this, general 1 will attack, yielding 4 against 3. Such games are played in many situations. A duopoly faces this structure often as well.

Credibility is a large factor in this kind of game. To be seen as a credible player, one can invest in its credibility. A good way to do this is by building a reputation. Threatening can be a good way to establish a reputation that you are tough. Pre-commitment is also a good way to build credibility. A good example of this is receiving warranty on products you buy. If the warranty is a full refund even after a full year of usage, the signal is quite clear that the producer of such a product does not think that the quality be any less than expected. In the case of the two generals above, the second general would be able to make his threat to fight back credible if he would make sure there was no possible way for him to retreat.

Chapter 20. Games and behaviour

Is the game theory a good explanation for behaviour? Burning your bridges to gain credibility about fighting back when under attack may not be the correct reasoning. Maybe the bridges have to be burned out of fear that otherwise it may be too attractive to retreat. It is important to not only look at the actions undertaken, but also at the intentions before the actions were done. What was the expected utility? This is more important than just looking at the actual utility. In some cases, the actors don’t even care about the actual consequences of their actions.

Some cases are subject to passion rather than reason. When a state is offered to join a union at a lower grade than the ones that are already part of that union, it may reject this for passionate reasons, while rationale would make it better off even at that lower grade. Such situations are looked upon as an Ultimatum Game or a Dictator Game. In the Ultimatum Game, the initiator proposes a distribution between him and the responder. When accepted, the distribution is as was proposed. If it does not get accepted both players are left with absolutely nothing. This is the case when a state is offered a membership according to certain rules. If it does not accept, the existing members do not receive any benefits either.

The dictator game is not really a game at all. The initiator offers a distribution, and this simply has to be accepted by the responder. When the Ultimatum Game and Dictator Game are put into experiments where the two actors are connected through an computer and therefore have no communication or even know who they are dealing with, the Dictator Game turned out to have much less generous distributions than did the Ultimatum game. The distribution of the Ultimatum Game is affected by the fact that really bad offers can be rejected. The responder will be more likely to reject a very low distribution, as he is driven not only by rationale, but also by feelings of fairness.

So, even when a distribution of 9.99, 0.01, it would be of advantage for the respondent to accept, the feeling of being treated unfairly would make him reject. We are willing to accept small damage to ourselves if this means unfair behaviour is punished.

The distribution chosen in the Ultimatum Game is come by through backward induction. Anticipating which distribution would be rejected is what makes the initiator determine which distribution he will propose. Calculations were quite simple in this case. Sometimes these calculations are very complex and involve long chains of subsequent reasoning. An example of this is when the respondent can react with a certain proposal at which the initiator needs to look at and determine whether or not he will accept:

Making the first proposal, the initiator needs to determine whether or not the responder will prefer a larger share of a smaller pie to the one he is offering. Every time an offer is rejected, there is 50% less to divide. A rejection makes the initiator and respondent worse off. The most logical distribution to propose in the first round is 3.75, 1.25. The respondent would accept this. In experiment, however, the mean distribution is 2.89, 2.11. There can be three reasons for this:

1. The initiator is too altruistic

2. The initiator doubt the capability of the respondent to conduct backward induction and fears he will reject the equilibrium distribution

3. The initiator may not be capable of backward induction.

It sometimes, however, just seems that human beings are simply not that rational. Looking back at the younger sibling syndrome, it seems that all the theory on rational behaviour is a large overestimation of human capacity.

When we repeat a finite number of prisoners dilemma’s we see that in some portion of the cases there is a choice to cooperate. The reason for this is because the actor choosing to cooperate hopes that his cooperation is one round will lead to the cooperation of the other in the next. In the last round, however, there is no incentive for either to cooperate, so they both defect. This means that the round before that has no incentive to cooperate as well. This reasoning can go back all the way to the very first round.

Backward induction should therefore result in a zero amount of cooperation when prisoner's dilemmas are repeated. In reality, however, we see that this is not the case. This does not necessarily mean that we are really not capable of thinking rationally because we hardly ever face a definite number of prisoners' dilemmas. So then we never know when will be the last one and therefore live for the here and now. And if this means we have to give some to get some we usually decide that’s the best way to go.

Furthermore, we do not really know who we are dealing with in these games. Some may always cooperate, others may always defect. There is a good probability for any kind and sometimes we just need to test the waters to be sure. Predatory pricing to scare possible entrants away is some strategy used quite often. As this is not rational, it causes us to raise our eyebrows. But if a business is known for being a fighter, the incentive for others to enter his market becomes less and less.

Chapter 21. Trust

Trust is an important thing in everyday life. To trust someone is to lower one’s guard, to refrain from taking precautions against an interaction partner. This includes when even the other could act in a way that might seem to justify precautions against an interaction partner because of opportunism or incompetence could act in a way that might seem to justify precautions. Opportunism is having a short-sighted self-interest unconstrained by neither ethical or prudential considerations.

Distrust exists in two forms. The first abstaining from interaction with a vulnerable party. Second form takes place when an agent engages in the interaction but takes precautions against these risks. Trust is the result of two successive decisions namely to engage in the interaction and to abstain from monitoring the interaction partner. The object of trust can be someone’s ability or motivation. People refrain themselves from taking precautions when the costs of taking the precautions exceed the benefits.

On four grounds people may be characterised as trustworthy; past behaviour, incentives, signs and signals. From observations of others we can conclude of they consistently keep their promises. Incentives can be seen as motivations for agents to act or postpone an action. Signs of features of individuals that are thought to indicate trustworthiness. Signals are behaviour that provides evidence of trustworthiness including the deliberate production or mimicking of sings. Trustworthiness is being induced by being trusted. The blind trust situation is manifested when precautions are excluded, as distinct from the case in which they are not chosen.

Chapter 22. Social Norms

Collective consciousness: a set of values and beliefs shared and known and believed by members of a community. Values are the social norms, moral norms, religion and political ideology. Beliefs are opinions, which can be about facts as well as causal relations. When stating that one should wear black in strong sunshine this is instrumental rationality. Wearing black in such weather gives the air between the body and clothes a faster circulation and therefore less heat. Stating that one should wear black to a funeral is a social norm with no instrumental significance.

Some social norms state not to do something, like not to eat another human being. Others tell you to do something, like starting with the fork and knife on the outside at dinner, working your way in. Some norms are conditional, doing one thing means doing another as well. Suggesting doing something sometimes makes you the one to carry it out. There is no outcome to be realized by such norms.

Causal efficacy is maintained through social sanctions that are directed at those who do not follow the rules as set out by the norms. These can affect the material state of the one who the sanctions are targeted at. This can be a direct effect as well as indirect through loss of opportunities. Gossip multiplies this. Ostracism, cutting off all contact with the violator, or avoidance is the most important reactions. This is because social norms are connected to emotions. The violator will feel ashamed, while the one he hurt with his violation will feel contempt. If the sanction comes at a cost to the sanctioner it even shows more how much the violation offended him or her.

Why is there an incentive to punish? Especially when it is, as in most cases, costly to punish, there is a need for further reason than just rationality why there must be punishment involved when someone violates a social norm. A good reason can be that if you do not punish, you can become a target of punishment yourself. This is very much evident in the behaviour of schoolchildren. There are, however, emotions of contempt present here. Sometimes this can even be combined with anger. It is these emotions that trigger the need for punishment, especially when the violation is flaunted with.

Social norms are not the same as moral or quasi-moral norms. Moral or quasi-moral norms shape the behaviour of one as he believes to be unobserved by anyone. When violating a social norm, however, one wants to hide away from anyone who is looking. Social norms are not legal norms either. Legal norms can be enforced by specialised agents who can direct punishment. There is a clear interaction with social norms, however, as social norms can turn into legal norms. Sometimes some things are legal, but social norms prohibit members of society to take action on that.

Social norms are not conventions. An agent purely based on self-interest can enforce conventions. Conventions are most likely to be respected, but that does not necessarily need to be. I can choose my seat in class and sit there every day, but when someone else decides to sit in that exact seat there is nothing I can do to make him leave. I do not own the seat. I might get angry, but there is nothing I can do. Political conventions act in the same way. Some things are agreed upon and even though they are not part of any constitution, we all respect them and violation is not an option.

Some norms reduce the small negative externalities that are imposed on others. Social norms tell us not to throw out garbage everywhere, not to spit in the street and not urinating in the swimming pool. Getting caught doing these things will not subject us to severe punishment, but we still refrain from acting in this way because we do not want others to think bad of us. Such norms make everybody better off. They follow from legal norms and turn into social norms because everybody follows them. It is, however, not clear if such norms would have come about if there hadn't been any public intervention.

There are social norms that prevent one group to impose a negative externality on another. Smokers are prohibited from imposing their smoking habits on non-smokers in any public place. When children are around, social norms make sure that smokers feel guilty for lighting a cigarette as they can severely harm the child. The same kind of norm goes for children who make too much noise. Their negative externality of noise can make other children avoid them, or the parents of other children can avoid their parents.

In some cases the violation of social norms feels like an attack on honour. Such a violation calls for strong retaliation. The form in which retaliation should take place can even be predetermined. Not retaliating is therefore another violation of social norms and can have effects on the person who was trespassed in the first place. He can be completely ignored by his own family and the rest of the community he is a part of. Not retaliating therefore leads to a feeling of shame and contempt. Anything that is seen as an insult to ones honour can have this effect. That is why the retaliation in such actions is quite severe.

Retaliation in the aforementioned class of social norms is in order to maintain the honour. In other cases the aim can be to gain honour. In societies where the strong reactions of violence towards those trespassing ones honour are accepted, the crime rates and homicide levels are higher as well. Some argue that the strong existence of violent behaviour is such societies serves to protect oneself against theft or any other kind of crime. It therefore creates an equivalent to organized law enforcement. The problem with such an explanation, however, is that cultures that have these points of view have much higher crime rates and deaths due to crime than do any other.

Norms of etiquette form another class of social norms. They are very detailed and need to be followed quite strictly if you do not want to be condemned into an outcast. In every culture they are different and depending on the social status they can be less detailed. One common problem is the difference between cultures when it comes to the amount of personal space. The Latin cultures tend to stand much closer to each other than do the North-Western Europeans. Other rules of etiquette are highly codified and seem quite pointless. Not following them can lead to very cruel consequences.

The rules of etiquette gain importance as they are seen to be important. Not following them shows that you are careless of their existence and lack some respect towards the system. Some may feel that if you cannot follow these rules of behaviour, you are more likely to violate other, more important, social norms as well. In the very least case it shows that you really do not care what other people think. For the elite, the etiquette makes it harder for outsiders to join. But this does not explain why there are such useless rules for behaviour in the first place.

Money cannot be used to buy grades, votes or organs. These are legal determinants of society. When it comes to the use of money, however, there is a set of social norms to follow as well. We do not ask each other how much money the other has. Nor do we ask what another earns. In some cases the relationship we have with others is as such that there is no money exchanged ever. This is the case for instance when parents take their children out for dinner and one invites a friend. In Western societies this friend should not ever offer or feel the need to pay. It is also a norm that when giving a present the receiver does not know how much you’ve spent on it.

There are social norms that determine the use of alcohol. It is in some cases not allowed to consume even a drop, like with religious groups. In other cases you need to prove your worth by consuming so much you cannot even spell your name anymore, like in a frat house. In other cultures it is seen as a lack of trust when you do not join in a round. The social norm that prevents one from drunk driving can give an easy out in such cases, you can use your driving as an excuse and not fear any more pressure from the host to have a drink. Using your religious beliefs as an excuse is also accepted. The social norm to drink is then offset by those that prohibit drinking.

A very puzzling social phenomenon is that of tipping. Doing it in order to make sure service will be good the next time you visit that restaurant or the next time you get your hair done is a good reason. But why tip when you will never go there again? Why tip when the service was not that good anyway? In some cases it can be explained by social conformity. We do not want our date or friends to think we are cheap by not leaving a tip. But we even do it when there is no one around to judge us. So it seems we only tip so that the person lending us his services will not feel any resentment towards us.

We can explain the effects of norms, but not their origin. We do not know what caused feelings of shame and contempt to form the norms in the first place. And why do they differ in different societies? What makes them develop and what causes them to disappear again? The main reason for their existence is our care for what others think about us. There is a strong fear for disapproval. The reason why they come about is to regulate externalities. In some cases the norms follow from actions taken by the formation of a law. But then they remain in place even when the law is no longer enforced. Economic scientists explain the existence of norms as equilibrium of behaviour we came about by accident and history.

Chapter 23. Collective Belief Formation

What one person believes should in fact not have much effect or mean anything to another. Still, we can find groups of people who believe the same thing. This can be because they have identical backgrounds and therefore develop the same beliefs. The conditions they face can be the same as well. This makes them look at things from the same angle. There can also be the pressure to conform at work here. Conforming outwardly is done through social pressure, we act the way that is socially expected and acceptable.

When this conformism is carried out long enough it will have an inward effect as well. This is through motivational and cognitive mechanisms. If the majority feels a certain way or has a certain opinion, it is very hard to think that they cannot possibly be right. Therefore the majority is accepted to have the correct point of view and opinion and theirs is the collective belief. On the other hand, there are intellectuals who are widely respected and their point of view and opinions are taken for the truth due to their superior wisdom.

In an experiment conducted on the inward and outward conformity, it turned out that when a question was asked in isolation the majority gave the correct answer. When the same question was asked when the target was exposed to a number op people who first all gave the same wrong answer, the target was more likely to answer the same as those he saw. This can be explained through the effect that the target could be learning or dissonance reduction. When the target was in the presence of those giving the wrong answers the chance of conformity was even higher. The reason for this increase is due to the fear of disapproval.

It is logical that the beliefs of another will affect you. Especially when it is a matter of reconstructing an experience, the testimony of another can definitely add to your own memory. How much the other person will affect you depends on how reliable you believe the other person to be. In some experiments the effect wore off after a series of repetitions with new additional targets, but in real life we see that cultural beliefs with absolutely no reality attached to it remain in practise a long time. The reason for this can be related to religious determinations. The discrepancy can also be so small that it is hard to detect.

People at a given time may hold similar beliefs due to correlation because they affect each other, i.e. causation. The first effect was clearly evident when Newton and Leibniz invented calculus around the same time. The conditions that determined the correlation were unknown, but there had to be something that made them do the same thing at the same time. Affecting each other happened daily as well. In religious cases it is very evident. In times where the communication with the rest of the world was not as easy, the individual was basically obliged to have the same religious beliefs and carry them out like the rest of the community. They prefer heresy to isolation.

These are beliefs that people do not hold sincerely, but only profess to hold. But it is not always clear what it means to believe something. Just the simple statement of a belief may give reason to want to endorse it. Pluralistic ignorance is when everybody believes that everybody believes something, while in fact there is not one member of the group who actually feels that this belief is true. This is an extreme of course, in more normal cases most people believe that most people believe something, while there are only a few who really think its true. And there are of course the pathological beliefs that everyone holds in public while we all know that no one holds them in private.

Communism is a form that showed an extreme degree of a culture of hypocrisy. This is also brought about by pluralistic ignorance. They are both brought into existence through a fear of disapproval or even punishment when you state that you do not agree with the common beliefs. In pluralistic the effect is horizontal, if you do not conform it are your neighbours, family and friends who will disapprove and shun you. In a culture of hypocrisy it is a vertical effect, where punishment or disapproval will come from people with some degree of power over you. This can again create a vertical effect.

Pluralistic ignorance is not the same as the bystander effect. In the bystander effect you feel that the fact that no one is doing anything justifies the fact that you do not do anything either. There is no social pressure not to do anything. It is not conformity, it is inference. As no one does anything, there cannot be something serious going on to worry about. Pluralistic ignorance shows a difference between the actions of an individual in order to satisfy the will and need to conform, while inside the beliefs are something completely different.

There is also the fact of the older sibling syndrome. We all know our own fears and insecurities, but we do not know that they inhibit others as well. So when a group of students was asked to read a completely incomprehensible article, they did not ask for any assistance while they were told this was an option to them. The reason for this is because they all feared the embarrassment of getting up and admitting they didn’t understand this. But on the other hand they all assumed they were the only one thinking this, and that the rest was not getting up because they understood the text. We see others as more mature and self-possessed.

Pluralistic ignorance can reach a stable state. This can happen when the false beliefs come true. Such a thing is usually called a self-fulfilling prophecy. It is most likely caused by dissonance reduction, as we are not so comfortable with saying one thing but actually believing another. People may also just stop doing it. This is when the situation unravels. When a small groups starts to stop conforming, this effect can snowball in a way that everyone comes out and tells the truth about their true beliefs. A single non-conformist is enough to start this, as others will feel strengthened in its claim as well. Unravelling can also be induced through the publishing of an opinion survey.

Rumours, fears and hopes have a great effect on the public belief. Rumour has an incredible magnifying effect. The more the same story is passed on from one to another, the more is added and there is no trace of the original occurrence anymore. Sometimes the speed and mechanisms of propagation of rumours can be identified. Actions elicited by rumours were by themselves inspiration of rumour. Letters and newspapers are an important source for rumours. Nowadays with the Internet and telephone possibilities, combined with television and other communication techniques the rumours travel even faster.

Official denials often seem to feed rumours. Furthermore, rumours are quite often set out by the government or the elite in an attempt to conspire against the people. Bad events can then be blamed on some other factor than the government or the elite. When a rumour started due to a bad harvest or another natural factor, the government or elite thought the people were conspiring against them and that there was a common agent spreading the rumour around, Conspiracy theories and the rumours about them were inseparable.

Rumours are believed to be spread by people who really want the rest to be convinced of the same truth. There may also be an incentive to exaggerate so there is no feeling of cowardice towards the storyteller. Some also seem to tell stories only based on what they believe their audience wants to hear. Some stories are only told to provide some sensation. The more reliable the source, the less the people listening to the story are inclined to question its accuracy. And then the snowball effect starts again, where telling the story means adding an extra layer of exaggeration.

We believe easily what we hope and what we fear. Rumours are in general more pessimistic than that they are positive. There is more fear than that there is hope. Only the rumours that are based on fear make people modify their behaviour. In the case of hope there is an effect, like exaggerating your chance of winning the lottery, but there are no real actions, like spending the prize money beforehand. When you fear that there will be something bad happening, you will take action on letting the effect be as small as possible. This is also how ethnic riots usually start when there is a rumour that the other side will attack. The asymmetry between the power that non-wishful thinking and wishful thinking have on behaviour is robust.

Rumours can also arise through informational cascades. Informational cascades are based on a mechanisms involving rational belief formation. There is private and public information combined, which each individual shares with the group. This is the same effect that happens when there is voting by roll call. You combine what you already know with the information shared by the rest of the group already. You know the outcome of what the others think, but you do not know the private information they accessed to form their opinion. Rational individuals may then have false beliefs.

Chapter 24. Collective Action

Sometimes collective action problems are solved by centralized action. There are rules to make sure we all do what is necessary for the greater good. There is also a sanction on free riding. This is a very common problem when it comes to public goods. There have been very elaborate experiments on how the behaviour of individuals comes about as far as public goods are concerned. The problem is that we do not see exactly what the benefits of our contributions are. We pay for goods, but we know that if we do not pay the good would still be available to us as well.

In experiments it became evident that the amount people contribute to public goods is dependent on what we feel others are paying as well. If 90% of the passengers in a bus travel without paying for it and there is absolutely no penalty or just not the chance of getting caught, the chance some additional passenger will pay is quite small. That is why the public transportation system needs to have people in employment whose only task is to check whether or not someone paid for their trip.

The figure below shows the process that contributions undertake as a function of the contributions that the rest of the people donate. If at first you believe the rest of society donated OD on average, the total amount is OH. If this is the actual amount received, the next time you face the choice on how much you need to pay the expected contribution of others is equal to OH and this is the same as OC. This in turn leads to a total of OG, which equals OB. Etc… As the level of contribution falls, the equilibrium is reached at OA = OE. Actual and expected contributions are the same. This is the process as is it unravelling.

There can be the opposite effect as well, we then call it snowballing effect. As actual donations exceed expected donations the total amount contributed grows. This is not the same figure as below, the process is observed in different settings. Think more in terms of how a small number of people can mobilize large crowds. This effect can take place when there is a public issue to protest against. Movements that result from this kind of action can bring down governments. Individual choices can have collective outcomes. The relation between these is still pretty difficult to explain, there is just great importance to understand that the relationship is there.

The dependence on how people react in matters of public interest is related to the externalities they face. These can be positive or negative. Paying your taxes has the positive externality that the schooling system works, the roads get cleaned and your trash gets picked up. It can also have the negative externality that you feel there is nothing done for you by the government and your payments are useless. The motives for each individual to take collective action are different. They are, however, supposed to be motivated only by the costs to them and the personal benefits they derive. But this does not explain the personal motivations.

Taking collective action means that each participant created positive externalities for the others at an individual cost. A worker on strike takes the risk of losing his job (cost) for the benefit of a higher wage possibility for all. Universal cooperation is more profitable than is universal defection. At the same time, with the same case and cost benefit analysis the individual defection would be a smarter decision than is individual cooperation. The whole is greater than the sum of its parts. The question is, however, how to determine when the group is large enough to have the benefits of universal cooperation. How do you even know that there will not be individuals who take advantage of the rest and just free ride?

The figure below shows the technology of collective action. The group is n+1. The behaviour of others is the horizontal line. If the individual cooperates the utility is line R (A-B). If he does not cooperate the utility is line L (O-C). The best outcome is unilateral non-cooperation, as then utility level C would be possible. The next best thing is universal cooperation with utility B. In all cases the choice not to cooperate will yield a higher utility than does the choice to cooperate. This is just like the two persons prisoners dilemma. If M people cooperate, however, their utility is better off than not cooperating, even when there are free riders in the game. OB shows the average benefits to the collective. This is a function of the number of participants.

The figure above is the situation when the costs for cooperation remain constant while the amount of participants increases. The figure below shows the case when the costs of cooperation increase. In some cases the last participants to join will actually reduce the marginal benefit of joining. Sometimes the costs start high and decrease as more people join. Furthermore, the figure below shows the average benefit as a linear function of the amount of participants, while increasing marginal benefits of the next participants are more frequent. In many cases the first and last participants add little to the average benefit, while the middle group has the most effect. This is perfectly illustrated in the figure below:

This is the fairest description of how the snowball effect works. The first contributors have to incur high costs to produce only few benefits to others. Then as more join in the benefits grow exponentially with the amount of additional co-operators. Eventually there will be a maximum reached and the marginal benefit goes down. Examples of this are when there are too many cooks volunteering for a class trip and they spoil the broth. If there are too many volunteers for the army, who will stay behind and make the bullets?

In most cases it is more profitable for the individual to do nothing and reap the benefits at the cost of others. There are two exceptions. When the individual faces the same collective problem repetitively, there is more incentive to cooperate as long as everyone else does. As soon as one defects, the best thing is to follow that persons lead and continue doing so forever. This is the behaviour in cartels. The second case is when there is an outside entity, which can enforce cooperation through punishment when defecting or rewards when cooperating. They rely on centralized enforcement and not by decentralized action by individuals themselves.

So if there is a clear incentive not to cooperate, why do we do it? There is a mixture of motivations, at the individual level as well as across individuals.

There are some motivations present in this mix:

• Some people are driven by more than just cost and benefit analysis. They are utilitarian, selfless or fully. Fully utilitarian means you will only join if your contribution has increased average benefit. Selfless utilitarian means you will join is benefits are raised, no matter your personal costs.

• Another category of people is the unconditional cooperator. These actors are not a function of the expected consequences of their actions, nor a function of the number of participants.

• A category that would never be a first mover is the group that is triggered by watching others cooperate. There is the need for fairness that triggers them. Then the social norms kick in. If they will be socially punished by not joining in, they will definitely participate.

• There are people who join because they think it is fun to do so.

This is not an explanation of how to successfully plan collective action; it is a framework on how to explain its existence. The outcome depends on the distribution of motivations among the population.

Chapter 25. Collective Decision Making

When there are collective issues at hand, we all have the option to make our individual choices. Guided by social norms and morality, we can have the possibility of achieving the best option some of the times just by interacting with each other. In other matters there needs to be centralized decision making. Collective decision-making is about making a policy choice. There are personal policy preferences, which are based on individual values and beliefs. These preferences are aggregated through three different mechanisms and then transformed to come out as a collective decision. Many cases do not involve all the people who have an interest with these decisions, but they are represented by those that are elected to do so. These representatives need to keep in mind that if they do not handle it according to the preferences of those he is representing, the change of re-election becomes smaller.

The mechanisms through which aggregation takes place are arguing, bargaining and voting. Pure argument happens when there is a jury that must come up with a unanimous decision. Some members may try to bargain in order to just get it over with. Time always matters when there are decisions to be made. Bargaining in its purest form is when you experiment with dividing a dollar games. Parties make offers and receive counteroffers. The bargaining power that parties have is of great effect to the outcome here. Voting means that the decisions are made in isolation and then the votes would be cast in isolation. There is no influence of one voter on another. This is not an easy setting to create in reality.

Arguing, voting and bargaining follow up eachother in the process of decision-making. At first there is a need for unanimity and the agents argue. As time becomes a pressing matter, the members decide to take a vote and therefore reach a decision faster. To some the matter is more important than it is to others. Therefore, the process of bargaining starts, during which logrolling agents try to exchange their votes for the promise of others votes. If there are no other votes to promise, maybe because there is only one case to vote on, the votes may be bought as well. This is of course not the legal course of action.

Arguing is the attempt to persuade another through reason. It is closely linked to democracy, as in parliaments the members of assembly argue about everything all the time. The idea is to make sure your standing point is viewed as to be the best for public interest. No one ever persuades someone by stating what the advantages are for himself only; you have to make sure the other thinks it is beneficial for him too. This is called misrepresentation of preferences, which is not the same as transmutation. Misrepresentation is more of a form of deception, while transmutation is self-deception. Misrepresentation can backfire if it is done to obviously. The need to disguise ones fundamental preference may even lead to a shift in someone’s policy preference.

Some arguments have been on the fundamental preferences. Due to the debate these fundamental preferences may even shift. There may be hidden similarities between cases that make it acceptable to change your opinion. A change in fundamental preferences can also be evident when a general principle contradicts intuitions about a particular matter. Initial unqualified utilitarianism can be revised to take account of non-consequentialist values.

Benefits of arguing may be undermined by the fear some have of speaking in front of an audience, On the other hand there are those who love the public eye and will say anything just to be in it. Nevertheless it remains quite difficult to stand up in front of a lot of other people and admit that you have been wrong or state that you support something controversial. Changing your mind in front of an audience is not easily done either. Arguments without an audience are therefore much more brutal and obviously catered to the individual preferences. When there is an audience there are more statements towards serving the public good.

Voting is needed when arguing is not doing the trick. It is not very simple to generate a consensus on policy through arguing. There are different voting systems. Some are closed, while others are open. Some need a 50% majority, while others require 66% before a policy is accepted. Public voting is possible as well, where the assembly is observed by an audience that does not participate in the voting process. The exclusion of an audience is not the same as secret voting, however. Some decisions require a minimal percentage of those who are eligible to vote to be present and actually cast a vote in order for the outcome to be valid.

There are different reasons why there needs to be a vote because arguing did not reach a consensus.

Consider the example below:

The three groups above are roughly of the same size. The reactionaries want absolute monarchy, the moderates want a monarchy with some parliament power as well and the radicals want absolutely no monarchy at all. From their differing standpoints, they will never reach a consensus through argument. Reactionaries and radicals voted in a way that there was only one system implemented.

Bargaining is the process of reaching an agreement by posing credible threats and giving credible promises. The outcome of such a procedure is completely dependent on the credibility of the person threatening or promising. The credibility of threats depends both on objective and subjective factors. Objective factors are the outside options available if the situation falls through. In case of a strike, the outside option is the wage the worker would get is he went to work somewhere else. Inside options are objective factors as well. They are what the parties have at their disposal during the negotiations. The inside options determine how long one can hold out. In the case of the strike the inside option is the size of the strike fund.

Subjective factors on the credibility are for instance the time horizon. The more time you have, the larger your bargaining power, as the one who is pressed for time will need to make larger concessions in return for an agreement. Risk neutrality gives an edge in bargaining as well. Subjective factors increasing credibility can also be placing a low value on that what is being offered compared to what the other party has to give up.

Credibility of promises is also affected in both the objective and subjective way. The credibility of a promise depends on the ability to deliver. The problem with most subjective factors determining this is that they are hard to observe and can therefore be misrepresented. The process in which collective decisions are made is illustrated below:

Chapter 26. Organizations and Institutions

Organizations are collective actors. They share the same goal and are guided through centralized decision making. Some are established, others emerged. Institutions are a way of doing things. They evolve gradually and will take a definite shape at some point in time. Once they have their definite shape, they can be implemented by organizations. Institutions which can be seen in most nations are those of a market economy as well as constitutional courts. As today it seems quite natural for all these institutions to be in place and functioning properly, their development took centuries. Many countries have experienced a lot in which their institutions evolved. Other nations or organizations can implement them as a result of these experiences.

Within an organization there are members and employees. Members interact horizontally, while employees have vertical relations. Vertical relations give rise to the principal-agent problem. The principal has different interest from the worker on how the time at work is spent. The problem in this case is the difficulty of monitoring the workers. When the work place is a public office, management may have in their interest to have their employees not accept any bribes and keep expenses to a minimum so that the public sector is as small as possible. For the employees, however, it is very lucrative to accept bribes and the size of their department is correlated with power, so there is an incentive to enhance this.

A solution to this problem could be the elimination of corruption opportunities. This can be established through competitive bidding for contracts although there can still be some influence carried out there. There is a bigger problem when the people who are supposed to be monitored are the ones with power. A corrupt president has very little people who can actually do something about it unless they play it out completely and the president will have to resign. The same goes for other authority figures, such as the president of the IMF or the dean of a university. But when they do get caught, they have to resign and carry the public shame with them.

You cannot completely restrict the employees freedom as an attempt to restrict their possibilities to act on their own interests. A good way to guide the agents’ behaviour is through incentives. Hard workers are therefore rewarded and those who do not work hard enough indirectly punished. Problems arise when it is difficult to monitor individual performance. Sometimes only the output of a group of employees is visible. Team bonuses would be the answer here, where the team members have to monitor each other. Here the problem of free riding shows its face, however.

Sometimes the costs of monitoring are higher than the benefits of catching any form of wrongdoing. Furthermore, who guards the guardian, who monitors the monitors? The best solution is a circular system. There is a chain in which there are material incentives to make sure all the work is done as it should be. Everyone controls everyone else. This is a very difficult thing to implement in practice. Trust is a strong issue when it comes to the problem of free riding.

A circular system of checks and balances is the best way in which each institution monitors the performance of the others. In most cases, however, there is one institution monitoring the others, while being an unchecked checker itself. This is the role of the US Supreme Court. In other countries this role is given to the parliament. Constitutions determine how the control systems ought to be. They are a form of organized distrust. As the people with power are not willing to give it up, even more they will try to expand it, the constitution protects us from this.

When constitutions are formed, there is much care for closing any loopholes through which corruptive behaviour could slip in. To explain in more depth, here follows a list of what should be in a constitution:

1. Prevent the government form engaging in political injustice through the use of juries or random assignment of judges to cases.

2. Prevent the government from manipulating the electoral system by placing the electoral law in the constitution.

3. Prevent the government from manipulating the flow of information such as newspapers and radio stations. The government should have no influence in the daily governing of these media and ensure that all information can reach the public freely.

4. Prevent the government from manipulating monetary policy the central bank should be an independent entity.

5. Prevent the bankers from enacting disastrous monetary policy by making it possible to dismiss them. Furthermore the length of time they can hold their position should be limited.

6. Prevent the government from manipulating statistical information the data-gathering task should be given to an independent bureau of statistics. The statistics represent reality, so the accuracy of statistics is just as important as the freedom of the media.

7. Prevent the government from starving the opposition; subsidies to political parties must be fixed in proportion to the amount of votes received.

8. Prevent the government from enacting self-serving legislation by stating in the constitution that certain types of law take a certain time before they are implemented after they have been accepted.

9. Prevent the government from bypassing the restrictions in the constitution by making amendments possible only if there is a supermajority in parliament.

10. Prevent the government from ignoring the constitutional restrictions by providing judicial review.

11. Prevent the government from manipulating judicial review by increasing the number of judges in the Supreme Court.

12. Prevent the Supreme Court from ignoring large portions of the population; they should be able to be dismissed if there is a supermajority for this in parliament.

There needs to be a balance in the strength with which the government in interlocked to prevent is from wrongdoing and the freedom it has to respond to the needs of the public.

If the twelve conditions are met, one can say the politics will be optimal. This is hard to establish in real life. The majority of governments will try, however, to approximate the optimal situation as closely as possible. Having said this, it is not true that the more conditions are met, the closer politics is to the optimum. Even so, it does not necessarily mean that if less conditions are met, the situation is less successful than a similar situation in which more or even all conditions are met. It depends on a number of conditions as well as the current situation of a given nation.