Melberg, Hans O. (1993), Three arguments about rational choice in sociology

[Note for bibliographic reference: Melberg, Hans O. (1993), Three arguments about rational choice theory in sociology, http://www.oocities.org/hmelberg/papers/930520.htm]
[Comments: This essay does not present any new and revolutionary ideas. It is probably best suited for undergradutes who simply want to read a short paper about rational choice in sociology. The paper was written in 1993 and today I do not agree with all the arguments, nor the way they are expressed. Nevertheless, I believe that the main arguments are still sound.]

Three arguments about rational choice theory in sociology

Three arguments
This essay will present three arguments:

1. That rational choice models are valuable in sociology because they provide a "rule of thumb" about how an action is chosen (a mechanism). Though, rational choice models are limited because they take many factors, such as preferences, as given. So, to explain preferences one must use other models.

2. That rational choice models are useful in explaining social exchange in the sense that the choice of action in an interactive situation is often governed by attempted maximization according to aims. However, once again it only models the mechanism taking certain factors as external (such as aims and the meaning of an action). So, once again, other theories are needed to explain the aims and meanings that govern the situation (such as Symbolic Interactionism).

3. That rational choice models very good at explaining how the problem of collective action may exist and be solved.

The value of rational choice models
Aim of sociology: Liberate or Explain
The value of a model depends on what one wants the model to do. This creates a problem because people with different conceptions about the aim of sociology would judge the value of models differently. For example, Jon Elster believes that the ultimate aim of sociology is to increase freedom in society. With this aim rational choice models are very valuable because they focus on intentional choice between alternatives. This implies that being conscious of the alternatives increases freedom. However, the people who argue that the aim of sociology is to describe how society works (i.e. maximise the explanatory power) would put larger emphasis on how choice is governed by internalised norms, thus leaving individuals few or no alternatives. On this view a rational choice model would be of less value since it does not account for the formation of preferences (which could be governed by a dominant ideology). So, the first point is that the value of any model, including rational choice, depends on the aim of the model.

Rational choice as explaining behaviour: Maximizing
It is plausible that human behaviour is governed by the principle of intended maximization of aims under given constraints (Jon Elster's two filters). If this is true, as empirical evidence in general shows, then rational choice models do have a role in sociology in terms of explaining which action is picked from the feasible set ie. that action which is expected to be optimal given the aims. So by explaining individual motivation, one then provides a possible explanation for macro phenomena since they are aggregate of individual actions. For example, one could use this model to explain revolutions (and revolutionary failures) in terms of individuals who maximize their expected utility. So, by providing a mechanism for individual motivation rational choice models may serve an explanatory role in sociology.

A false counter-argument: Altruism or emotional behaviour
It is important to note that rational choice in sociology does not mean the same as rational choice in consumer economics. This is so because in consumer economics one also assumes that being rational means to be selfish. However, there are no theoretical objections to use rational choice models to explain altruistic or emotional behaviour. Even an altruist would have to chose the best possible action ; Even a person motivated by emotional hatred would chose the best action to satisfy his emotional demands. Rational choice is just a model describing how to best chose actions taking the aims as given. So, altruistic or emotional behaviour does not refute the model.

Problem: Ignore the formation of aims and utilities
At the same time that the neutrality with respect to aims is a strength of rational choice models, it is also a weakness. This is so because in sociology one would also want to go beyond taking aims as exogenously given. One would also want to explain why people have the aims they have and why people have the utilities they have. Why do people attach high utilities to Rolex watches? Why do people sometimes behave altruistically? Why does industrialisation lead to alienation? To provide the answers to these questions one cannot use models of rational choice. Instead one is forced to other models such as symbolic interactionism. This model provides a theory of why certain goods are linked to certain signals. For example, deviance such as green hair is a signal of rebellion. Rational choice can explain why a person who aims at rebellion will colour his hair green (maximization of aim), but it cannot explain why green hair is a symbol - a meaning - of deviance since this is taken as given.

This is not to say that rational expectations are useless. It only emphasises that it describes only one process (i.e. the choice from the feasible set) and it needs to be complemented by other models to provide a more satisfactory explanation of society.

Social exchange and Rational choice
As humans we engage in activities with other humans i.e. there is social exchange. The question is then why we engage in such behaviour. Rational choice would argue that it is because the benefits are larger than the costs, otherwise the activity will not take place. For example the act of giving a gift is motivated by the reciprocity of later receiving a gift. On this view to receive a favour is like incurring a debt i.e. it has to be repaid. One could go on to explain many interactions in such cost-benefit terms. For example, the act of giving advise could be costly in terms of time and effort, but rewarding in terms of higher status. The real question is then whether this kind of net benefit motivation really is how people are motivated in an exchange situation.

Anthony Heat thinks that there are exceptions, such as a mother - child relationship. Such a relationship he argues, is not governed by net benefit considerations. I believe A. Heath is at least partially wrong. The key is that rational choice does not, as argued, imply selfish behaviour. It is thus possible, according to rational choice, that the mother can chose to die herself rather than letting the child die. Seeing rational choice in terms of any aim it is clear that a mother - child relationship may well be governed by the application of rational choice to social exchange. For example mother - child relationships are often governed by altruistic aims in which behaviour can be explained in terms of rational choice of the most altruistic behaviour. Furthermore a mother - child relationship is governed by a general aim of doing what one think is best for the child. One may then use rational choice models to explain why certain decisions were taken, such as stopping the child from watching television all night. Lastly, there seems to be some element of calculation in the relationship since some women sometimes choose abortion because it is the "best" option considering all the costs and benefits.

Problem: Exchange with no prospect of reciprocity
There are also types of exchange with no prospects of reciprocity, such as when a person stops you on the street and asks you what time it is. Since you are never going to meet this person again there is little incentive in terms of reciprocity to do anything at all for this person. So it would seem that rational choice would be indifferent between lying, not answering and telling the truth.

One possible solution to this problem is that norms to tell the truth or to be helpful, could be beneficial to me. So in order to maintain this norm there is a small incentive in favour of telling the truth (since although this is only one instance out of a million it does have a little impact on the maintenance of the norm however small). Though, this is not a very robust solution. One could improve it by saying that people have an internalised norm that says "tell the truth" or "help" and that to break this norms would cause internal displeasure. This is not inconsistent with rational choice since it takes human nature as given so internal emotional reactions should count when deciding upon the course of action.

An even more robust answer would be to say that helpful behaviour is intrinsically pleasant. It is by helping one becomes human, one realizes oneself. This solves the practical problem by changing the payoffs facing the individual. Since telling the truth becomes intrinsically pleasurable, then this alternative is preferred in a rational choice model over not telling the truth when someone asks what time it is.

Alternative theory of exchange: Norms
Not all interaction can be described in terms of optimising behaviour. The reason is that sometime people do not even think before they act, they simply follow an internalised rule. It is, however, possible to explain much of this behaviour in terms of rational choice. First of all it is time efficient to follow norms, so that rational choice could well end up by recommending the use of norms. Secondly, it is possible that norms are beneficial to you (as with the norm of telling the truth). Thirdly, it is possible that norms are better for motivating humans since emotions may be stronger than reason (so that unconditional rules prevent a slippery slope development of conditional statements about net benefits). Thus, while conceding that there may be norms which are not explained by rational choice models, it is at least plausible that many norms can be explained by rational choice.

Rational choice and collective action
The problem of collective action is described by the famous Prisoner's Dilemma case. The essence of the problem is that for a group it would be best if all engaged in activity X, but for each individual it is better to not do X. Thus, some kind of mechanism must be created to exploit the potential welfare gain by cooperation by excluding the free rider problem. The question is whether rational choice can be applied to collective choice to provide such mechanism.

At first sight it would seems like rational choice models could not explain much co-operation. For example, there is very little incentive for me to vote according to rational choice models because the effect of my vote is neglectable. So, the argument goes, one needs other models such as conceptions of fairness to explain voting and similar cooperative behaviour.

I will argue that this view is wrong. Once again it ignores the distinction between selfishness and rational choice. Voting would be entirely consistent with altruistic aims for example. Furthermore, voting could be explained in terms of rational choice by the effect of generating greater self respect, despite the small effect it has on the outcome. So, the aim of voting is not mainly to support a party, but a personal aim of "participating."

Repetition, Uncertainty and Weak mathematical abilities generates cooperation
There are further conditions which makes rational choice models useful when applied to the theory of collective choice. For example, it is a fact that many relationships are repeated, there is often some uncertainty and people have limited mathematical abilities. These conditions enable rational choice to generate cooperation because in a world of uncertainty one might believe that the other individual will deviate from their selfish rational strategy to free ride (also uncertainty about when he repetitions will stop makes the game approximately similar to infinite games in which cooperation is generated). Because of repetition cooperation becomes more profitable and hence more attractive according to rational choice. Finally limited mathematical abilities improve the possibilities for rational choice cooperation because it stops the backward induction defection argument (The backward induction argument is that in a finitely repeated prisoner's dilemma case it is rational for selfish individuals to defect in the first round. This is based on the argument that it is rational to defect in the second last round, so then it becomes rational for the other player to defect one round before the second last round. But knowing that it becomes rational for you to defect in the round before the third last round etc. But, if people are not able to think in many layers (complex), then one will get cooperation for a long time and only defection on the last few rounds.

Failure/Outcome of collective action: Bargaining theory
Rational choice may also explain why collective action may not occur. One reason why this is so, according to rational choice models, is that people may not agree on how to distribute the costs and gains from cooperation. This problem is especially large when there are many participants needed (such as to create an army and pay for it), when the costs to cooperation is high and when the benefits of free riding are high. So, according to rational choice models voluntary cooperation would be difficult to find in the above mentioned circumstances and this corresponds to the real world. To solve such problems one then creates coercive solutions such as creating a state with the power to tax everyone.

The outcome of any bargaining about collective action, according to rational choice, is given by the bargaining power of the participants. This power is determined mainly, according to Elster, by time preferences and inside options (i.e. the available alternatives). The paradox for rational choice, however, is to explain why the outcome of a bargaining process is shaped by irrelevant alternatives i.e. alternatives which would never be chosen. The only solution I can think of is to say that the independent alternatives shape the emotional willingness of people to accept an outcome. This then effects the outcome because the more unwilling they are to accept the outcome, the bargaining power is restored even when the feasibility set is reduced.

Conclusion
This essay has defended the theory of rational choice without arguing that it explains all phenomena in the social sciences. The main distinction was that rational choice is good at explaining behaviour under given conditions, but that it was poor at describing how these conditions came about. It was also argued that many of the criticisms directed against rational choice derived from the mistaken view that rational choice necessarily involved selfish behaviour. The value of rational choice is seen as a tool in social science which is complementary to other kinds of explanatory devices.

[Note for bibliographic reference: Melberg, Hans O. (1993), Three arguments about rational choice theory in sociology, http://www.oocities.org/hmelberg/papers/930520.htm]