Sociology 319

March 3-10, 2003

Rational Choice Theory (RCT)


Readings for this section are Peter Abell, “Sociological Theory and Rational Choice Theory,” Chapter 8 of Turner and James Coleman, Foundations of Social Theory, p. 292-299 and 220-228.


1. Introduction


A different approach to the theory of social action from that discussed so far is rational choice theory.  The theories discussed earlier consider individual and group actions to be social in that they are concerned with meaning, interpretation, interaction, symbols, self-definition, feelings, emotions, and how people do things together.  Weber and Parsons differ from the interactionists in terms of how they define meaning and the relative importance that each attributes to subjective consciousness as opposed to praxis.  But all these perspectives emphasize that action is social in the sense that social values and norms, and social order, are integral to their explanations of human interaction.  Each of these approaches emphasizes shared meanings and ways in which interpretation and action are worked out jointly through a broad range of expressive aspects of human action.  Rational choice theories adopt a different approach to the study of social action, human agency, and social systems and structures. 


Abell defines RCT as “understand[ing] individual actors … as acting, or more likely interacting, in a manner such that they can be deemed to be doing the best they can for themselves, given their objectives, resources, and circumstances, as they seem them” (p. 223).  As with other sociological perspectives, there are many variants of rational choice theory, but each tends to differ from other sociological perspectives in being minimalist and focusing on the individual – a short discussion of each follows. 


a. Individual.  First, RCT begins from the viewpoint of the individual actor, rather than from several individuals interacting together, from social situations, or from groups.  In this sense, RCT stands at the opposite end of the interactionist spectrum from ethnomethodology – for the latter there is always interaction.  But RCT emphasizes the individual social actor, his or her interests as a starting point, and achieving the optimum for him or her.  Different theorists of rational choice may make somewhat different assumptions about the individual and proceed in different ways from the individual to larger social groupings and systems, but each begins with the individual as the basic unit of the theory.  Abell notes that “it is only individuals who ultimately take actions and social actions … individual actions and social actions are optimally chosen” and “individuals’ actions and social actions are entirely concerned with their own welfare” (p. 231).  Each of these basic assumptions show the methodological individualism of RCT – the individual as actor with an initial concern only about him or herself and his or her welfare.  From this basis RCT may show how sharing, cooperation, or norms emerge, but the basis is always the individual social actor.


b. Minimalist.  A second major aspect of RCT is that it is sociologically minimalist.  RCT begins with a few simple assumptions about the individual and the relationship among individuals, and from these builds models of social action and interaction that describe and explain the complexities of larger groups, systems, and whole societies.  This approach is very different from the systems and structural approaches of Durkheim or Parsons, who make social norms and values at the societal level an essential feature of their perspective.  It also differs from most writers in the Marxian and Weberian traditions that emphasize large-scale, global, and historical social forces.  RCT also runs counter to the symbolic interaction, interpretive, and feminist approaches that adopt a more dense and complex view of social actors and social interaction.  The latter consider meaning, interpretation, emotions, experiences, and a wide variety of aspects of human existence, none of which can be reduced to the other, nor is capable of simple explanation.  Much recent sociological theorizing has emphasized the complexity of the individual, the variety of individual experiences, and the diverse ways that individuals develop selves and social action.  The terms thick descriptions and rich explanations are sometimes used in this connection.  Goffman’s detailed analysis of how people present and maintain face is an example of how complex human actions appear to be.  In contrast, RCT adopts a relatively spare and simple model of the individual, one that can be applied across time and space, so that it is a universal model.  RCT also is an example of a more formal theoretical perspective in the “scientific” tradition, with assumptions, concepts, logical deductions from these, and formal models. 


Variety and Uses of RCT. 


As with the other theoretical approaches, RCT has many varieties.  While they all are methodologically individualist and minimalist, each may adopt somewhat different assumptions.  Following this, each develops these in a somewhat different way.  The concerns of various theorists in this tradition are quite different, although all attempt to build a model that explains various social phenomena or society as a whole. 


a. Micro-macro.  In his introduction to Section III of the text, Turner notes that RCT is “the most overt and systematic attempt to resolve the micro-macro relationship … to find an adequate analytical bridge between individual social actions and their structural, macro outcomes” (Turner, first edition, p. 222).  RCT is not modest about its claims, and some theorists in this tradition argue that RCT provides the only overall explanation of social systems and society, and the only solid basis for progress toward a unified social theory.  RCT sociologists tackle issues related to individual action and social interaction, and also construct models explaining social norms and collective behaviour.  While this perspective begins from minimalist assumptions, its scope is very broad – to provide an unified, all-encompassing social theory.  James Coleman ambitiously titles his book Foundations of Social Theory.


b. Individualist and neoliberal?  Because of the emphasis on the individual, RCT is sometimes considered as part of neoliberal ideology – an emphasis on the individual that has been associated with a decline of broad based social programs and social concerns in contemporary society.  While RCT could be turned in this direction, it does not necessarily lead to an individualistic overall model of society.  For example, part of Coleman’s aim is to explain why and how social norms emerge.  In doing this, Coleman does not reject the existence of norms but tackles an issue that we might all agree needs explanation, and one which Durkheim and Parsons took for granted.  


c. Rational choice Marxism.  Some neo-Marxians are rational choice theorists.  While there are many different aspect to Marx’s model of capitalism, in some ways it is a rational choice model, although not as individualistic as most RCT.  Marx begins with  assumptions about the commodity and how humans exercise choice, and builds an explanation for exploitation and class struggle from this.  Some recent Marxist theorists (most notably Jon Elster and John Roemer, and Erik Olin Wright uses some of this approach) have developed new models of exploitation based on models of rational choice.  These have been useful in showing how exploitation can emerge in ways other than the exploitation of labour directly in the labour process – for example, exploitation might occur in the exchange process.  Roemer argues that “exploitation has much more to do with property relations than with the labor market – and that Marxists’ focus on the labor market has been excessive and has given rise to their own fetishism of labor” (Roemer, 1988, p. 10).  Roemer shows how exploitation can exist in socialist societies and how ownership of skills and credentials can be associated with exploitation, just as well as ownership of capital.


d. Usefulness.  As with any other sociological theory, RCT should be considered on the basis of its ability to help us explain and understand the social world.  There is no doubt that each of us is an individual, and if a theory developed from this point of view can help explain aspects of social interaction and social systems, then it has worthwhile aspects to it.  In addition, in our society much social action is explicitly rational and is undertaken by individuals – purchase of consumer durables, choice of a career, and perhaps even choice of a lover or spouse.  Where the choices are not always entirely conscious and rational, it is possible that RCT models may help explain much social action. 


e. Critique.  Critics of RCT note several problems – too individualistic, too minimalist, and too focussed on rational choices in social action.  One tendency that RCT sociologists have is to justify any human action as rational.  For example, we are all involved in sharing and cooperative activities and each of us devotes some time or money assisting others.  RCT tends to argue that in the end, these are all inspired by individual pursuit of self-interest.  As a result, RCT sometimes attempts to explain too much – any theory that tries to explain everything may in the end explain nothing, since there are no standards concerning what factors are to be introduced into the model and no standards concerning how they are to be considered.


If combined with other approaches, RCT can prove to be useful in sociological theory.  While it is not the most common sociological approach, it has active proponents who have developed many interesting ideas and many testable hypotheses.




2. Background


There are many different influences on RCT – utilitarian economics, Weber, Pareto, and recent North American and European theorists. 


a. Weber.  As soon as rationality is mentioned in sociology, Weber’s approach comes to mind.  For Weber, rationality was a driving social force in society, especially in modern society.  He used rationality in several different senses but Holton notes that in all of these “its principal meaning … centers on the calculability, intellectualization, and impersonal logic of goal-directed action.  The instrumental approach to action takes values as given and focusses instead on the efficient choice of means to reach such goals” (Turner, first edition, p. 43).  In such action, a primary focus is on conscious action by the individual social actor, considering others and attempting to achieve his or her own goals in a considered and systematic manner.  Weber regarded this as characteristic of modern society, and tended to regard rationality as an overpowering social force that increasingly affects all aspects of society. 


Abell also notes that RCT can be regarded as one way of working out an explanation of the social world in a Weberian manner.  Weber argued that sociologists should develop an interpretive understanding of social action in order to explain “its course and effects” (p. 223).  For Weber, action is social in that it takes account of the “behavior of others and is thereby oriented in its course” (p. 223).  While most sociologists working in the Weberian tradition adopt a more complex view of meaning and interpretation and how social actors interact, Abell notes that RCT provides one possible way of explaining social action.  RCT provides an interpretation for individual action, it shows what the effects of this are, and it certainly is focussed on goals and orientation – so it satisfies Weber’s conditions for social explanation.  Abell reiterates this on pp. 228-9, outlining the aims of RCT and showing that these parallel the three conditions of Weber: (i) interpretive understanding, (ii) social action, and (iii) causal explanation of its course and effects.  Abell also noted how Weber’s model can be expanded to include social interaction and interdependence among social actors (p. 230). 


See also James Coleman, Foundations of Social Theory, pp. 6-10, for a rational choice analysis and critique of Weber’s argument concerning the influence of the protestant ethic on the emergence of the capitalistic spirit and capitalism.


b. Utilitarian Economics.  Much economic theory, originating from Adam Smith and developed by many nineteenth and twentieth century economists, foreshadows RCT.  In fact, it may be that sociologists, impressed with the overall power and rigour of such economic models, developed RCT as a sociological counterpart to utilitarian economic models.  In general, such models are based on the assumption that economic actors are “rational and as seeking to maximize their utilities or benefits” (J. Turner, p. 303).  These economic models usually begin with an individual who has a set of preferences and who faces constraints (resources, incomes, prices).  Presented with various options, the individual economic actor decides how best to achieve his or her preferences, given the set of constraints and choices available.  Models explaining how individuals decide to purchase certain commodities or supply labour using this approach have proven to be powerful economic models – they can help explain effects of taxes, changes in consumer behaviour, labour force activity, and business operations.


As noted earlier, nineteenth and twentieth century sociologists were concerned that such economic models could not explain social order.  Durkheim argued that the rational economic actions, economic exchange, and contracts by themselves could not work without social norms, conventions, and laws.  Similarly, Parsons argued that action could not be aimed at purely utilitarian ends, but had to be guided by social norms and values.  Adam Smith himself had set his economic models within a framework of balancing  human sentiments which guided economic actors, so that his model was not purely utilitarian.  This aspect of Smith’s approach tended to be ignored by many later economic models which concentrated only on rationality, constraints, choices, and ends of economic action, ignoring the setting in which these occurred. 


A related influence on RCT is that of game theory – how individual actors make decisions in game like situations.  Issues such as strategies, preferences, and decision-making processes come to the fore in these situations. 


c. Pareto.  Vilfredo Pareto (1848-1923) was of Italian-French ancestry who became a professor of economics in Switzerland in 1893, and spent the last part of his life in Italy, where he supported the Fascist regime of Mussolini.  Pareto had many concerns but he is known for distinguishing the maximum utility of the community as a whole from the maximum utility of members of the community (as an aggregate of individuals).  Actions which might be regarded as leading to the former, such as war, nationalism, or specific public policies, might be associated with sacrifice or harm to some individuals.  Pareto was concerned to preserve individual utilities and argued that individual utilities were noncomparable.  As a result, he argued that public policy should be directed toward an optimum that is now referred to as a Pareto optimum – an optimum whereby no one can gain greater benefits without others losing some benefits.  For example, in a social situation it may be possible to make both persons better off through free exchange – presumably the exchange improves the situation of each, otherwise why would they have entered into the exchange.  In other situations, it may be possible to improve the situation of one party without hurting the other.  A Pareto optimum is reached when there are no further possible improvements to the situation of one without harming the other party.  There may be several Pareto optima, with each noncomparable with the other in terms of which is preferable for society, so that there may not be single social optimum.  What Pareto’s method shows is that a methodologically individualist approach can address certain issues of social action.  (see Waters, pp. 61-3).


d. Recent Contributors.  The names most associated with RCT are George Homans and Peter Blau, who examine social exchange and the benefits and costs of alternative courses of action.  Their concerns parallel economic issues, but focus on social rather than economic exchange, with social behaviour as an exchange of activity (Ritzer, p. 405).  Some examples from the sociologist James Coleman, another proponent of RCT, are provided later.


Abell (pp. 227-8) notes how RCT attempts to explain relationships of the macro ® micro, micro ® micro, and micro ® macro type, so that they can help illuminate macro ® macro relationships which may be observed at the system level.  That is, RCT aims to look at individual action and interaction within social systems, with the aim of building models that can explain larger phenomena, perhaps at the societal level.  To the extent that such a sociological approach is successful, this is one means by which the agency-structure debate can be addressed. 


See the example from Coleman, pp. 6-10 (on reserve) to see how Weber’s model of the development of capitalism can be explained with an individual rational choice model.  Coleman notes though that Weber did not adequately address the issue of how the micro ® macro aggregation of capitalistic action causes or furthers the development of capitalism.


3. Assumptions of RCT


Abell notes three assumptions that rational choice theorists make (pp. 230-1).  These are:


·        Individualism – it is individuals who ultimately take actions.  These individual social actions are the ultimate source of larger social outcomes.  This may run contrary to those who adopt particular views of Durkheim concerning social facts as being at the societal level, and in some ways determining individual action through societal level forces such as norms and common consciousness.  Abell notes though that RCT can develop societal level explanations, it is just that RC theorists begin with individual action and interaction and from this build a model of relationships at the system level (see models of p. 228 and 231).  Part of Coleman’s concern was to develop explanations of different types of norms, the demand for norms, how norms emerge, and how they are realized.  RCT does not take these for granted but attempts to develop an explanation of these system or societal level forces.


·        Optimality – actions of individuals are optimally chosen given the preferences of the individual and given the opportunities or constraints the individual faces.  That is, the individual as social actor attempts to achieve the best for himself or herself, given the circumstances of the individual.   Abell defines optimality as taking place when no other course of social action would be preferred by the individual over the chosen course of action.  Note that this does not mean the course that the actor adopts is the best course in terms of some objective, outside judgment.  Frequently we note that other people do not appear to choose the course of action for themselves – but in RCT this would be sociologically misguided.  It may be that the individual is acting in his or her own best interests as far as he or she perceives them.  One possibility is that the individual does not have complete information (see Abell, p. 234), may lack many resources, or may have preferences that others may consider unusual or misguided.  RCT at this stage says nothing about optimal formation of individual preferences, but begins by treating these as given.  Abell notes that RCT does not have a good theory of how wants or affects develop, but would rely on a model of social learning (p. 234).  While this is  a weakness of RCT, it is a weakness that is common to other sociological approaches as well – few or no sociological theories have a model that explains differences in individual preferences and how these emerge.  In the end, RCT assumes that individuals “do the best they can, given their circumstances as they see them” (p. 234).  If individuals do not really do this, for example if individuals were to act randomly or in a way that was clearly contrary to their own interests, as they see them, then RCT would be severely damaged.


Structures.  Abell argues that structures and norms, which dictate a single course of action are merely special cases of RCT.  That is, in other circumstances there is a range of choices, whereas in a strong structural explanation there may be only one choice.  While this may be damaging to RCT, even in situations with few choices, individuals often find a way to exercise agency and do they best they can.  Goffman’s examples of means by which people in prisons and asylums attempt to maintain a self, or Ira Berlin’s history of slavery in the United States South where slaves were able to alter their situation, are examples of extreme structural limitations where seemingly rational choices are exercised.   In terms of inequalities emerging from different resource endowments, rational choice Marxism is certainly able to explain various forms of exploitation – so RCT does not necessarily show harmony, consensus, or equality.  Similarly, what we call structures may not be optimal from the viewpoint of the individual with few resources, what RCT will attempt to explain is how this situation emerges and is maintained through rational choices.


·        Self-Regard – the actions of the individual are concerned entirely with his or her own welfare.  Abell notes that this is a key assumption that forms part of most RCT, but one that is not as essential to it as the optimality assumption.  He notes that various types of group sentiments could exist, but sociologists should then develop an explanation of how they emerge and develop.  These could be sentiments such as cooperation, altruism or self-sacrifice, charity, or self-denial, which may initially seem to be contrary to pursuing the best for the self.  RCT might argue that these sentiments can be incorporated into its models by noting that these may have as an ultimate end the pursuit of some type of self-interest.  Charity, for example, could be to make oneself feel good or be a means or raising one’s social esteem in the eyes of others.  RC theorists are more likely to argue that the simplest assumption is that of self-interest or self-regard.  Begin with this assumption and see if an explanation for group sentiments can be built from this.  If so, then that demonstrates the value of RCT.  For example, cooperation could easily develop from individual self-interest, especially in conditions of limited resources or uncertainty.  Self-denial or asceticism may be more difficult to explain, but again could result from individual pursuit of spiritual salvation – that is, given the individual’s values, it may be that these can be optimally pursued by a life of self-denial.


Rule also argues that self-regard may not be so essential to the models as the fact that “human action is essentially instrumental, so that most social behavior can be explained as efforts to attain one or another, more or less distance, end” (p. 80).  In addition, “these ends or values are organized in relatively stable hierarchies of preference or utility” (p. 80).  That is, the various preferences of individuals are organized so that an individual has more or less important preferences that can be ordered with respect to each other.  In general, these do not change rapidly, so that social processes tend to have a certain regularity or predictability.  People base decisions and actions on these preferences, even if they do not consciously calculate the decision before each action.  So long as these preferences are relatively stable, RCT can build models that help understand social action.


Do individuals actually conform to the RCT model?   Coleman does not consider this as important an issue as whether RCT develops adequate explanations of social systems, that suit the needs of the issue to be addressed.  A RC theorist may be pragmatic in this sense, using a model that is simple and minimal, but which is useful for the purpose intended.  This seems like a reasonable approach, so that RCT can be judged by whether it provides adequate explanations for social phenomena.


4. RCT models and explanations of social phenomena


A major theoretical work on RCT was published in 1990 by James Coleman (1926-1995, United States) just a few years before his death.  Coleman was a sociologist who received his doctorate at Columbia University, working with Robert Merton and Seymour Martin Lipset.  He studied public policy and in the 1960s wrote a report on educational opportunity, termed the Coleman Report.  This advocated busing and other educational policies to improve equality of educational opportunity – a liberal, progressive program.  In later years he concentrated on mathematical sociology and RCT, publishing The Foundations of Social Theory in 1990. (See Ritzer, pp. 427-434).


Coleman argued that sociologists should be concerned with the social settings in which social action occurs.  He notes “Pairwise exchanges in social life do not take place in a vacuum.  They take place in a setting in which there is competition for the resources held by each actor” (Coleman, p. 131).  Systems he cites are dating systems of teenagers and exchanges between grades and performances in a school setting.  In each of these, the

preferences, orientations, and actions of individuals are one aspect of the system and resources another.  The key concepts for Coleman are actors and resources – “interaction and ultimately social organization revolve around transactions between those who have and those who seek resources” (J. Turner, p. 312).  In addition, the initial distribution of control of resources among the actors is an essential starting point.


While focussing on the individual actor, it is the social system as a unit that is the starting point for Coleman’s sociological analysis.  He describes a dating system as follows:

In a high-school dating system the actors are the boys and girls attending the school.  The resource with which each begins is control of his or her own attention.  The interests of various boys in the attention of certain girls and vice versa is the dating patterns, that is, the redistribution of attention.  (p. 132)

In a high school a date between a girl and a boy depends not only on their interest in one another, but also on their interest in others and others’ interests in each of them. (p. 131)

That is, exchange, redistribution, and competition occur in the system, and these along with the orientation, purpose, and mode of action of the individuals form the system.  By examining the individual within systems, the sociologist can build a model of social systems and develop explanations of social activity. 


a. Internalization of Norms.  Based on discussion from Coleman, pp. 292-299. 


Examples of norms.  These may be customs or etiquette, or may relate to forms of behaviour which have broader implications.  For example, throwing garbage in a waste container, not smoking in university buildings, not cheating on examinations, or not entering into physical confrontation with those who one disagrees with.   In each case these are generally accepted, at least among the group of system of which one is part.  Sanctions may be social approval and disapproval, or may be more formal sanctions such as rules and laws which have penalties associated with them.


Purpose.  Difficulty of identifying how purpose emerges in a theory of purposive action.  But Coleman makes some statements and conclusions about this anyway. (292)   The same problem was considered earlier in the semester, when examining Parsons – the problem was the inability of a utilitarian theory to explain how ends or goals are selected.


Norms?  Assuming existence of externally imposed would be easier but this would make the theory less general and weaker, because it make prediction more difficult, especially when interests or norms change. (293)


Internalization of norms occurs when norms have been internalized so that the individual establishes and maintains an internal system of sanctions.  These provide an internal form of punishment if the individual’s actions are inconsistent with the norm. (293)   That is, internalization of norms in others, if it can be accomplished, is an efficient means of exerting influence on the others.


Establishment of Sanctions:


Socialization may be the process by which norms are established in individuals and through which an internal sanctioning system exists.  Note that both external and internal mechanisms may be used, so there may be no pure case of one or the other.  In the case of children, Coleman notes that parents may physically punish children or provide external rewards.  Where words and gestures that express approval or disapproval are used, these may be aimed more at establishing or building upon a set of internal sanctions.   (294).


Internal or external sanctions?  More efficient to establish internal sanctioning systems in others.  That is, the costs of ensuring that the norms are met can be reduced if individuals adopt the norms and police or sanction themselves.  Note the focus on costs or resources as guiding decision-making – the resources involved may be time, money, or other resources that might be required to impose external sanctions.  Compare the cost of establishing an internal norm with the discounted future cost of external policing and the benefits obtained from the internalization.  The relative effectiveness of each would also have to be considered, although Coleman does not discuss that here.  This is a fairly straightforward benefit/cost analysis of the sort that economists carry out with respect to more specifically economic activities – for example, the benefits and costs of building a road.  One difficulty with RCT is that this may sound good in theory, but it is very difficult to estimate such costs, especially with great uncertainty as to the future of the children or the parents.


Modified self and identification with socializer.  One strategy that socializing agents such as parents, religious organizations, or employers may use is to attempt to “get the individual to identify with the socializing agent” (top of p. 295).  If this can be accomplished, this appears to be a relatively efficient way to produce individuals who develop an internal system of regulation.  Coleman notes that this means modification of the self of others – producing a new self which will decide what is right or wrong.  This is efficient since it means that the socializing agent need not dictate specific rules, but can rely on the modified self to produce proper actions in the socialized individual.  (295).  The aim is to “align the agent’s interest so fully with those of the principal that the agent’s self-interest comes to coincide with the principal’s interest” (top of p. 296).  As Coleman notes, this may go quite deep and result in a major change in the interests of the individual, so that the self of the socialized really does take on the interests of the socializer.


Corporate selves who identify with the goals, interests, and norms of the corporation may be provided with monetary rewards (tied to the success of the corporation), long-term employment, and collective activities that attempt to build a sense of community.


Efficiency of internalization:


Diversity of actions.  The more diverse the set of actions to be governed by norms, the greater the interests in internalization.    This implies


·        That when the aim of the socializer is to govern a broad set of actions, greater attempts will be made to create a new self (296, bottom) – e.g. religious orders or parents who wish to govern a broad set of actions of children. 

·        Costs devoted to internalization are likely to be related to potential benefits.  This implies that greater efforts will be made to have children internalize norms in the home, but less effort may be devoted to internalizing norms that do not relate to the home.  (297, top)

·        Investment in internalization will depend on the length of time children stay at home.  That is, intergenerational families of previous eras may have devoted greater efforts to internalization of norms.  (297, bottom)

·        Parents can increase returns to themselves by identifying with child later in life.  Note the increased satisfaction to parents by getting approval from others.  (298)

·        Community status may affect investment in internalization.

·        Cost of supervision and observation may affect method of socialization. 


Note the emphasis on individuality, optimality, and self-interest.  The interests of the socializer are to ensure that others conform to the socializer’s norms.  That is, the socializer pursues self-interest and attempts to do so optimally – to select a socializing strategy which reduces costs as much as possible.  In terms of the establishment of the norms internally into the socialized individual, the aim is to change the self-interest of the socialized individual in a particular direction. 


Coleman accepts the idea that norms can be established and maintained, his theoretical explanation provides some clues concerning how this will be done.  That is, they emerge out of specific forms of rational action, and may change depending on the circumstances.  Also note the extensive comparisons with economic situations and extensive use of economic concepts (e.g. middle of p. 297). 


Some of Coleman’s specific comments concerning egalitarian parents or children of the 1960s may be misleading.  Here Coleman appears to use specific interpretations of their interests, abstract some aspects of their interests, and make comparisons with other situations that may not be exactly comparable.  At the same time, if other things are more or less the same, some of these predictions may turn out to be correct.  Note that Coleman does not say these are necessarily correct, but that these are testable propositions that come from his model. 


In terms of norms, what Coleman has done here is to establish a model of self-interest and from this argue that more macro level social aspects can be explained from micro models.  His systems, such as families as socializers or corporations as trainers, use the model of actors using the resources at their disposal to work in their own self-interests to establish norms in others.  Broader aspects of this model include socialization, training, and the modification and development of the self.


b. Collective Behaviour.  Based on discussion by Coleman, pp. 220-229.  This is an example of what Abell refers to as strategic social action (pp. 267 and 269), in which Coleman analyzes “the tactics of the actors in arriving at rates of exchange” (p. 269).


The issue addressed here is how a crowd engages in “expressive acts which no member would have engaged in alone” (220).  In this example, from 1968, Columbia University proposed to build a gymnasium on public land rented from the city of New York.  Little access was to be provided to these facilities for community members.  These were primarily African-American and Hispanic, since Columbia University is in the part of New York City known as Harlem.  Community leaders and students became involved and protests against the actions of the University developed on the campus.  SDS is Students for a Democratic Society (with leaders Ted Gold and Mark Rudd), an organization of predominantly white students who were opposed to the Vietnam war and various activities of the administration of Columbia University.  SAS was a black student organization, the Student Afro-American Society.  While the original demonstration was organized, the manner in which the demonstration proceeded was somewhat spontaneous, with the crowd responding to calls and suggestions from unidentified individuals.  The collective marched to the gymnasium site, attempted to tear down a fence and, after returning to the main campus, eventually occupied Hamilton Hall – actions which would not have made much sense for individuals acting alone.  Coleman notes though that “many students had an interest in some action against the university” (222). 


Coleman makes the argument that individuals acting alone were essentially barred from these radical actions because they would have been severely punished if they had carried out these actions as individuals.  As individuals, rights of control over student actions existed for authorities of the university and the city.  When the students began to act as a collective, the students turned their rights of control over to the collective, and to those who suggested radical action.  Coleman argues that this shift began with what he refers to as a “milling” period, where individuals attempted to “determine the degree of common sentiment” (223).  By gaining the information from each other that there was a sentiment to proceed to more radical action, the crowd became a collective, with the individual members turning their rights of control over to the collective.  The collective was then able to take on certain actions which the individuals could not, as individuals.  


Coleman notes that this was a means of solving the free rider problem – without collective action, the negative consequences of action, or the costs of such action, would be entirely borne by individuals who took individual action.  Those who do not take such action but might have benefited from such action, do not bear the costs, so long as it is up to each individual to act.  This can be referred to as externalities associated with actions and “the consequences for the individual are highly dependent on what others do” (223).


Coleman notes that the number of people involved in such actions alters the situation – with increasing numbers of those with common sentiments leading to individuals being more likely to turn authority over to the collective.  In such situations, the potential gains of collective action are likely to be greater and the potential costs less than in the case of individual action.  In addition, Coleman notes that increased numbers may be associated with increased courage – exactly how this relates to the assumption of RCT though is not clear. 


Table 9.8 is an example that shows how rewards are increased (or costs decreased) as the number of actors involved increases.  In this table, collective action would be unlikely to occur before five actors are involved.  Any increase in the number of actors beyond this would increase the likelihood of collective action.  Figure 9.3 demonstrates that it is in the interest of the actor to encourage others to act.  By increasing the number of actors in the collective, individual benefits rise with the number of actors. 


In Table 9.9, a reward structure is such that A1 will not act unless A2 does.  That is, for A1 there is no positive payoff unless A2 acts (and a large negative result if A1 acts while A2 does not).  But if A2 acts, then A1 is better off by joining in the collective action.  The optimum result is when both A1 and A2 act.


On p. 225, it is also noted that when a diverse or heterogeneous set of individuals are involved, with quite different reward structures, collective action is more likely to emerge.  That is, it may be worthwhile for one person to participate in collective action when as few as 2 people are involved, and for another it may be worthwhile at 3 people.  As a result, the critical minimum for the transfer of authority to the collective may be reduced if these heterogeneous reward structures exist.  Note that this is contrary to common sense, in that we often attribute collective action to common sentiments and solidarity, rather than diversity. 


In this section Coleman shows how collective action can emerge from individual rational decision making.  The students assembled each had grievances against the University, so their preferences or values were to carry out some action that could stop the construction of the gymnasium, change the University’s plans, or embarrass the University.  Individually they could not do this very effectively, if at all, but collectively they could.  The action of the collective could be considered rational in that sense, although Coleman might not want to refer to the collective as having a rationality.  But the actions taken were the result of rational choices for the individuals involved in the collective.  And as the payoff matrix shows, everyone is better off if collective action occurs.


Whether all the different aspects of crowd behaviour can be examined with this single model is questionable.  But Coleman’s model is useful in providing some unexpected predictions, and some explanation of how rational choice can have wider effects, even though it is rational choice at the individual level that produces this.


5. Analytical Marxism or rational choice Marxism (RCM)


One branch of RCT theory is analytical Marxism or rational choice Marxism (RCM), although Abell only mentions this in passing (p. 235).  It has primarily been social scientists in the Marxian tradition who have attempted to develop Marxism in this manner, so that there is a considerable gap between the theoretical approaches of writers like Coleman and RCM.  Writers who consider themselves analytical Marxists ask many of the same questions and address the same issues as did Marx – transitions from one mode of production to another, class structure, class consciousness, exploitation, socialism.  Some of the answers of RCM are the same as Marx and some are different, but generally RCM is committed to many of the same ideals as those to which Marx was committed – democratic socialism and human freedom and creativity.  (See Wright in Carver and Thomas, pp. 23-24).


Wright notes that analytical Marxism has four characteristics (in Carver and Thomas, p. 14):


·        Commitment to conventional scientific norms concerning construction of theory and research.

·        Emphasis on systematic conceptualization of concepts that are at the core of Marxian theory, for example, mode of production, surplus value, exploitation.

·        Careful specification of the steps involved in constructing explanations and models of society.

·        Emphasizing the importance of the intentional action of individuals.  In RCM, individuals are self-seeking, face choices, and generally attempt to obtain the best they can, given the circumstances.


RCM often begins with individuals, considering them as having the ability to work and having access to various amounts of resources, both natural resources and those created by humans.  Resources may be land, tools, capital equipment, buildings, skills and abilities, and organizational resources.  Labour, as that which is part of humans and an aspect of their creativity, plays a key role in RCM, and it is the loss of labour by some and appropriation of labour by others that is key to understanding production and exchange.  While the method and starting point of RCM is very similar to other RCT, the different set of questions asked lead it in a different direction.  The emphasis of RCM is more likely to be on the inequality of resources among different social actors, and the consequences of this for the social position of the individual, with emphasis on the use of human labour and exploitation.  Much of the detailed work of RCM is economic theory, with an examination of labour, capital, surplus value, and exploitation.


As an example of RCM, the following notes review Erik Olin Wright’s discussion of social classes in Classes, Chapter 3, pp. 64-104.  Wright’s approach comes from Roemer, but is developed in a more sociological manner.


a. Labour Transfer Approach (see Wright, Classes, pp. 65-67)


2 people – one industrious, one not ® no exploitation.  That is, one does not gain at expense of other (Pareto).

Exploitation – “one person’s welfare is obtained at the expense of the other” (65)

Transfer of products of labour is one way that this can be done, and this is the typical form of exploitation in capitalism.  That is, the capitalist is wealthy because of expropriation of labour power of those who have no resources other than their own labour power. 

Roemer – exploitation can occur if all own means of production, but size of assets differ.  So it is not the relationship to the means of production that is key, but distribution of assets.  As a result of differential quantities of assets owned, so that some need to work more hours to obtain subsistence, free trade leads to exploitation of those with few assets by those with many.  That is, the asset rich can work less hours because the asset poor work more.

Exploitation can thus emerge merely from inequalities of asset ownership, not just from deprivation of ownership of means of production.


b. Game Theory Approach (see Wright, Classes, pp. 67-71)


Actors bring different labour and other assets (means of production, skills) together, and abide by certain rules.  Would actors be better off if they stayed in game or withdrew from the game? 



A coalition of actors S exploited by another coalition S’ if


            i. alternative exists where S would be better off than they are in game

            ii. if S withdraws, then S’ would be worse off than they are in the game.

            iii. S’ prevent S from withdrawing – dominance


Capitalist exploitation – if each individual worker were able to take their per capita share of society’s productive assets, they would be better off to withdraw from capitalism.  This implies that workers are exploited. 


Feudal exploitation – if each serf were to withdraw from the system with only their personal assets, they would be better off and the lord would be worse off.  This is a definition of feudal exploitation.  In this system, the serf has some assets of his own, but provides certain products to the lord.  As a result, the serf would be better off if freed and the lord would be worse off if serfs were freed. 


Socialist exploitation is a situation where members would be better off if they could withdraw with their per capita share of inalienable assets – skills and talents.    That is, the highly skilled exploit the less skilled so that the highly skilled would be worse off than at present if skills were equally shared.  In this case, skilled workers attempt to preserve their privilege through limiting the number of people who obtain such skills.  In order to maintain such an exploitative position, more highly skilled workers attempt to maintain skill differences and may make it difficult for those without valuable credentials to obtain these credentials.


Assets, Exploitation, and Classes


Type of Class Structure

Principal Asset that is Unequally Distributed

Mechanism of Exploitation



Labour power

Coercive extraction of surplus labour

Lords and serfs


Means of production

Market exchanges of labour power and commodities

Capitalists and workers



Planned appropriation and distribution of surplus based on hierarchy


bureaucrats and non-management



Negotiated redistribution of surplus from workers to experts

Experts and workers


Source: Erik Olin Wright. 1985. Classes.  London: Verso.  p. 83, Table 3.2




6. Weakness of RCT


In addition to the difficulties associated with accepting the three basic assumptions, there are a number of other problems associated with RCT.  Some of these are:


·        Problems associated with inadequate information and uncertainty.  This may make it difficult for individuals to make rational decisions.  As a result, they may rely on other ways of making decisions.


·        Human social action and interaction are complex, and many of the theories examined earlier may provide better guides to how these take place.


·        Theorists of rational choice argue that macro level structures and institutions can be explained from the models of individual social action.  But there are problems of aggregation of individual to societal level phenomena.  These same difficulties exist in well developed economic models.


·        Norms and habits may guide much action, and once these take root people may not question them but use them to pursue meaningful social action.


·        One problem of RCT is that some theorists argue that almost everything humans do is rational, even altruism and self-sacrifice.  By expanding to include all forms of action as rational, action that is nonrational or irrational becomes part of the model.  By including every possible form of action in rational choice, it is not clear how the standards of what is rational and what is not are constructed.


7. Conclusion


It is always worth considering whether action is rational or not.  That is, it is not sociologically wise to assume that actions taken by others are irrational when we disagree with them.  It may be that in the eyes of the individual taking the action, the action is rational, and RCT suggests we should look for such reasons.  Even where other theories may work well, RCT may be useful because there is a rational element to much human social action.  Whether it can explain all human action and provide a means of uniting and developing sociological theory is less clear.




Coleman, James S. 1990.  Foundations of Social Theory.  Cambridge, Massachusetts: Belknap Press of Harvard University Press.

Ritzer, George.  1996. Sociological Theory.  New York: McGraw-Hill.

Roemer, John E.  1988. Free to Lose: An Introduction to Marxist Economic Philosophy.  Cambridge, Massachusetts: Harvard University Press.

Rule, James B.  1997. Theory and Progress in Social Science. Cambridge: Cambridge University Press.

Waters, Malcolm.  1994. Modern Sociological Theory.  London: Sage Publications.

Wright, Erik Olin. 1985. Classes.  London: Verso.


Last edited March 11, 2003

Return to Sociology 319