The concept of power has been central to social exchange theory since its early development. Peter Blau (1964) included “power” in the title of his classic theory of exchange; Richard Emerson's (1972a, 1972b) exchange formulation was built upon his earlier analysis of power-dependence relations; and theories and research on power have dominated the contemporary development of the sociological exchange tradition for the past thirty years (see Molm 2000 for a review).

Although social exchange theories are typically formulated at an abstract level designed to apply to different settings and different levels of analysis, several features make them particularly well suited for application to organizations. The participants in social exchange, called actors, can be either individual persons or collective actors such as teams or organizations, and either specific entities or interchangeable occupants of structural positions (e.g., the CEO of a firm). Actors are connected to one another in dyadic relations, and those relations are typically embedded in larger exchange networks of variable size and complexity, making social exchange theories appropriate for analysis of both intra- and inter-organizational power relations. Finally, the very concept of exchange is central to most organizational analyses.

Social exchange theory is most appropriately viewed as a framework or theoretical orientation that includes numerous specific theories. All social exchange theories share certain concepts and assumptions, but they also differ in various respects.

This chapter challenges the restriction of social exchange theory to the mutual exchange of rewarding outcomes. I begin by examining how and why exchange theorists excluded punishment and coercive power from the scope of their theories. I argue that this scope condition is unnecessarily restrictive, and that it prevents the comparison of power based on both rewards and punishments. Next, I review theory and research from fields that have addressed punishment and coercive power, including macro theories of political power, social psychological theories of conflict and bargaining, and psychological studies of punishment. Finally, I extend the basic assumptions and concepts presented in Chapter 2 to incorporate punishment and coercion. I conclude by discussing the scope of the research program that occupies the remainder of the book.

The exclusion of punishment and coercion from social exchange theory

Although most social exchange theorists included both rewards and costs in their conceptual armament, they restricted costs to those incurred by actors from their own exchange behaviors. Costs that one actor imposes on another (punishment), and power based on the capacity to punish, were omitted from their theories.

Homans ([1961] 1974) and Blau (1964) explicitly excluded punishment and coercive power from the scope of social exchange, and Emerson (1962, 1972a, 1972b) did so implicitly. Only Thibaut and Kelley (1959), whose theory was intended to encompass all patterns of interdependence in dyads (not only exchange relations), included punishment among the potential costs of interaction.

This chapter begins to build and test a theory that explains why reward power and punishment power have such different effects on social exchange. Its focus is one of the most consistent and surprising findings of the previous experiments: the infrequent use of coercion by actors in all structural positions. I assume that the low use of coercion is the immediate cause of the weak effects of punishment power; to explain those effects, we must first explain why actors use coercion so rarely. Chapters 6 through 8 address that question. In this chapter, I consider when and how power use is induced by the structure of power alone; in Chapters 7 and 8, I examine factors that affect strategic power use.

The assumption that structural power is effective only if it is used is consistent with exchange theory; the suggestion that power use is problematic is not. Emerson assumed that if actors possessed a structural power advantage, they would use it. The incentive to use power was inherent in power itself. This chapter shows that this principle holds only for reward power and explains why.

I argue that the coercive use of punishment power, unlike the use of reward power, is not induced by a structural power advantage. This is the primary reason that it is used less frequently. I show that when this difference between reward and punishment power is eliminated, by creating experimental conditions in which a structural advantage on either base of power does induce the use of that power, the hypothesis of structural equivalency is supported: the two bases of power have equivalent effects on the distribution of exchange.

The five preceding chapters developed and tested a theory of coercion in social exchange. Because one of my aims was to show how a program of cumulative experimental research can be used as a tool for building and testing theory, I constructed the theoretical puzzle piece by piece. This chapter provides an overview of the finished picture. Here, I summarize the development of the theory, offer a more formal statement of its logic, and discuss the findings that support it. I then examine the implications of the work for social exchange theory, and revisit the advantages and disadvantages of integrating the study of coercive power within the framework of social exchange.

Logic and development of the theory

Scope conditions

The five conditions that limit the scope of Emerson's theory of reward exchange (Table 2.1) also restrict the scope of coercive exchange theory. The boundaries of the research project were further set by the additional conditions specified in Chapter 3. Several of these are clearly theoretical scope conditions, restricting coercive exchange theory to (1) relations in which actors can reward as well as punish each other, (2) negatively connected networks, (3) nonnegotiated (reciprocal) exchanges, and (4) settings in which actors can neither change the structure nor avoid a partner's rewards or punishments. The logic of the theory rests on the assumption that these conditions are met.

The classical exchange theorists (Blau 1964; Homans [1961] 1974; Thibaut and Kelley 1959) believed that social exchange is governed not only by the actual benefits actors obtain from one another, but by the expectations they bring to the relation as a result of past experiences or social norms. These expectations, like the status quo, provide a reference point for evaluating the rewards received from exchange. Rewards that either exceed or fall short of expectations produce emotional reactions, such as satisfaction or anger, and these emotions can influence behavior.

When expectations are based on social comparisons of some kind – what you receive in comparison to another person or reference group – questions of fairness or justice arise. Because such comparisons are an inevitable result of social exchange, power and justice are considered, by many, to be the two most important concepts of social exchange theory.

Nevertheless, the role of justice in the classical theories is something of an aberration. The concept of justice is found in neither of the root theories of social exchange (behavioral psychology or microeconomics), and its inclusion in exchange theory requires that we import concepts from other theories, either normative or cognitive. While Blau (1964) emphasized the former, contemporary justice theories tend to give more weight to the latter. Principles of justice have a normative base, but because perceptions of justice are in the eye of the beholder, they are subject to some of the same kinds of cognitive biases (e.g., loss aversion) discussed in the preceding chapter.

I began this work with two main objectives: theoretically, to determine whether the scope of exchange theory could be extended to include coercive power, and empirically, to compare the effects of reward power and coercive power on social exchange. Chapter 10 reviewed the results of the first effort, by summarizing the theory of coercive exchange developed in the project and evaluating its implications for social exchange theory.

This chapter evaluates the empirical contributions of the project. Here, I review the major substantive findings of the research, examine their implications for social relationships, and consider their relevance for other structures, forms, and conditions of exchange.

The core findings

Six core findings summarize the main conclusions about coercive power in social exchange. Many of these findings are not only nonintuitive, but directly contrary to common beliefs about the use and effectiveness of coercion.

First, coercion is not structurally induced by a coercive power advantage, but purposively enacted as a strategy to increase rewards. While a reward power advantage directly induces the use of that power, actors who use coercion are motivated, instead, by structural conditions that make acquiring rewards problematic. Their use of punishment is not an unintended side effect of other actions or other exchanges. Rather, it is a purposive effort to change a partner's behavior, by creating contingencies that produce negative consequences for the partner's failure to reward them.

The preceding chapter developed the thesis that punishment power is less effective than reward power because its use is not structurally induced by power advantage. Unlike reward power, coercion is not an unintended by-product of actors seeking exchange with alternative partners. Instead, it must be used strategically, by actors punishing their partners' failure to reward. The effective use of power strategies requires more information, skill, and behavioral monitoring than structurally induced power use. On this basis alone, then, we would expect to observe some differences in the use and effects of the two bases of power.

But we would not expect to find the very low rates of punishment that all subjects, even those with a strong incentive to coerce, have repeatedly displayed in these experiments. Actors might use punishment ineffectively and inconsistently, but why should they use it so infrequently? And why does punishment always decline, before it has any chance to influence the partner's behavior?

This chapter begins to answer these questions, by developing and testing a theory of strategic power use. Although Emerson included voluntary modes of power use within the scope of his theory, he assumed that they were fairly irrelevant; structure would determine power use in any event. Consequently, he made few assumptions about the decision-making processes of individual actors.

In this chapter I draw on theories of choice and decision making to explain how the strategic use of both bases of power – but especially coercive power – is constrained by risk and fear of loss.

Experimental research, especially in laboratory settings, is relatively rare in the social sciences. Common criticisms of laboratory experiments (e.g., artificiality, lack of random sampling, irrelevance for real-world situations) suggest that their role in building and testing theories is not well understood. Consequently, in addition to describing the basic experimental setting used in this program, this chapter explains the theoretical logic that underlies experimentation in general, and this setting in particular. I begin with a brief discussion of the purpose of laboratory experiments and why artificiality is an advantage rather than a disadvantage for the task of this book: constructing and testing a theory of coercive power in social exchange. Then, I turn to a detailed description and explanation of the experimental setting I use for this task.

Laboratory experiments and theory construction: descriptive versus formal theory

Most criticisms of laboratory experiments are based on the assumption that the purpose of social science research is to obtain findings that can be generalized to natural settings and specific populations. Research on power, for example, should tell us something about power in families, in organizations, or in international relations. If so, then the more the research setting resembles famines, organizations, or international relations, the more likely that our findings would apply to those settings. Thus, the obvious and enormous disparities between characteristics of those settings, and characteristics of an experimental setting in which undergraduate subjects sit in small isolated rooms, making choices on computers that produce small increments or decrements of money for other undergraduate subjects (which, as I discuss in a moment, is the kind of setting used for this research), make experiments seem trivial and irrelevant.

The 10-year research project consisted of two distinct phases: an initial phase that compared equivalent structures of reward and punishment power and tested exploratory hypotheses about differences in structural effects, and a second phase that constructed and tested a theory of coercive power in social exchange.

This chapter describes the research and results of the first phase; the next four chapters are devoted to the second. This rather unequal division of labor reflects the greater importance of the second phase of theory construction. Theories are rarely built in a vacuum, however, and deductive theorizing is most productive when it builds upon an initial knowledge base. As Chapter 3 illustrated, that base was small and inconsistent at the beginning of the project. The five experiments in the first phase expanded and strengthened it, and shaped the direction of the subsequent work.

The chapter begins with a discussion of each of the five experiments: their objectives, design, results, and implications. The analyses in this portion of the chapter focus on how the manipulations of structural power affect the frequency and distribution of exchange in the relation. Then, in the second part of the chapter, I use the combined data set from all five experiments to examine the dynamic process of exchange and actors' exchange strategies.

The three preceding chapters have sought to explain why coercive power is used so infrequently in social exchange relations. We have seen that coercion is not structurally induced by a punishment power advantage, and that the strategic use of coercion is constrained by fear of loss, structural dependence, and norms of fairness. As a result, even when actors who are disadvantaged on reward exchange have the power to coerce more rewards from their partners, they seldom use it.

In this chapter I examine whether the low use of coercive power accounts for its weak effects on social exchange. Chapter 6 showed that both bases of power have equal effects when either is the sole source of power. And Chapter 7 showed that coercive power is more effective when the risk of retaliation is removed and its use increases. But an important question remains: how effective is coercion in the networks that are the focus of this research – networks in which all actors have both reward and punishment power, and retaliation is possible? If actors in these networks used coercion more frequently and consistently, would their coercive power be more effective?

As we saw in Chapter 3, many theories – from macro analyses of political power to micro analyses of social interaction – contend that it would not. The exchange theorists who excluded punishment from their analyses, as well as many theorists who explicitly study the use and effects of coercive power, argue that it will be less effective the more it is used.

With the exception of minor modifications required for certain experiments (e.g., those in which subjects controlled either rewards or punishments for one another), the instructions shown here were used for all experiments in the program. Each subject was referred to by a particular letter. To avoid the evaluative connotations that student subjects might attach to the letters normally associated with academic grades (i.e., A, B, C, and D), the four actors in the network were instead labeled W, X, Y, and Z. These instructions are written for Person X; consequently, they refer to the subject's two interaction partners as Persons W and Y – the partners for Person X.

Screen 1

WELCOME TO THE SOCIAL INTERACTION EXPERIMENT!

You are participating in this experiment with 3 other students. Like you, they volunteered for the experiment to earn money. During the experiment, you will interact with one another by using the computers on your desks. The choices you make will affect how much money you earn.

Because we don't want your interaction to be influenced by personal characteristics like sex or appearance, you will not meet or talk to each other either during or after the experiment. You will interact only through the computers, which are connected to a larger computer in the control room. The computer records your responses for future data analysis.

During the experiment, you will use the computer mouse on your desk to make choices that will affect each other's earnings.

The product of nearly a decade of theoretical and empirical work on power in social exchange networks, this book describes the progression of a theoretically driven, cumulative program of experimental research, begun at Emory University in the mid-1980s and completed at the University of Arizona. During that 10-year period, many of the results of the individual experiments in the program were published in journals, the typical outlet for experimental research.

As the program progressed and both empirical findings and theoretical knowledge began to cumulate, however, I became more and more frustrated with the impossibility of conveying the substance of the whole in a series of pieces, each of which had to stand more or less on its own. The logic and significance of each successive experiment depended substantially on what had gone before and, as that earlier body of work became larger and larger, the journal format could no longer accommodate what I wanted to say. This book has given me that opportunity.

In writing this book, I had two different, but related, objectives. One was to compare reward-based and coercive forms of power in social exchange relations: how they are different, why they are different, and the implications of those differences for both theory and social relationships. The results of this analysis offer important insights and challenges for both social exchange theory (which traditionally has ignored the role of punishment and coercion in social relationships) and theories of coercive power (which traditionally have ignored the capacities of actors, in most relations, to reward as well as to punish each other).

Actors. Individual persons or corporate groups who engage in exchange.

Alternatives. Alternative sources (typically, alternative exchange partners) for obtaining valued outcomes within a single exchange domain.

Average power. The average of two actors' dependencies on each other.

Average reward power refers to the average of the actors' dependencies on each other for rewards, and average punishment power to the average of their dependencies on each other for punishments.

Dependence. An actor is dependent on another to the extent that outcomes valued by the actor are contingent on exchange with the other. Dependence is a function of both value and available alternatives.

Exchange. The mutual giving and receiving of valued outcomes by two actors.

Exchange costs. Defined broadly as any negative value obtained through exchange (including punishments), but used specifically to refer to the negative value incurred by the actor who performs an exchange behavior (i.e., opportunity costs, investment costs, actual loss of resources, and costs intrinsic to the behavior).

Exchange domains. A class of outcomes that are functionally equivalent in the sense that the receipt of one outcome in the class reduces the value of all outcomes within that class.

Exchange network. A set of two or more connected exchange relations.

Exchange opportunity. An occasion on which an exchange can be initiated.

Exchange outcomes. The positive or negative value received from another actor through exchange.

This chapter lays the theoretical groundwork for the project. Here I describe the basic concepts, assumptions, and principles of social exchange theory and its conception of power. My discussion is based primarily on the contributions of Richard Emerson (1962, 1972a, 1972b), whose theory of power-dependence relations provides the framework for this work. I also draw on the insights of Thibaut and Kelley (1959), whose early treatment of power-dependence relations strongly influenced Emerson's subsequent work; of Homans ([1961] 1974), who is considered by many to be the founding father of sociological exchange theory; and of Blau (1964), whose conception of social exchange as nonnegotiated – the terms and timing of which are left unspecified – corresponds most closely to my own.

My discussion in this chapter is limited to reward-based exchange, the traditional focus of social exchange theories. In the next chapter I evaluate this restriction and discuss the consequences of removing it for the assumptions and principles introduced here.

The concept of social exchange

The conception of social interaction as social exchange comes, of course, from the example of economic exchange. As anthropologists recognized as early as the 1920s (Malinowski 1922; Mauss 1925), many forms of social interaction outside of the economic marketplace can be conceptualized as an exchange of benefits. “Neighbors exchange favors; children, toys; colleagues, assistance; acquaintances, courtesies; politicians, concessions” (Blau 1964:88).