Social Exchange Theory
of
John Thibaut & Harold Kelley
(From the Second Edition of A First Look at Communication Theory by Em Griffin, (1991) 8th.ed. 2014McGraw-Hill, Inc. This text-only version of the article appears on the World Wide Web site www.afirstlook.com. A facsimile of the original article is also available in PDF format.)
In 1959 Perry Smith and Dick Hickock invaded a Kansas farmhouse in a senseless robbery that netted under $50. After tying up the four family members in separate rooms, they blasted them with a shotgun to eliminate all witnesses. Two months later the men were captured and placed in separate interrogation rooms. The police had enough evidence to convict them for parole violation and passing bad checks, but the evidence for murder was thin. The prosecutor needed a confession.
Truman Capote’s book In Cold Blood describes the dilemma faced by Smith and Hickock. Shortly after the crime, they had agreed to stick together so that they could back up their planned alibi if arrested. Yet held in isolation, each doubted the other’s will or ability to hold out. Smith saw Hickock as a convincing liar, but thought his "guts were unreliable." Similarly, Hickock feared dying on the gallows because he thought Smith would lose his nerve. Often in jointly committed homicides, the suspect who turns state’s evidence gains immunity while the other one, who feigns innocence, gets the death penalty. This differential treatment provides a strong temptation to cop out. But admitting guilt doesn’t always achieve leniency. If both suspects confess, the government has more evidence than it needs to convict for murder, and both killers get a life sentence. Yet if both stonewall, the prosecutor can get a conviction only on a lesser charge. This is the classic prisoner’s dilemma: Confess the crime or maintain innocence?
THE OUTCOME MATRIX AS A MIRROR OF LIFE
Figure 16.1 diagrams the interdependence of Smith and Hickock. This 2 x 2 matrix is the central analytical device of John Thibaut and Harold Kelley’s social exchange theory. Thibaut was professor of psychology at the University of North Carolina until his death in 1986; Kelley continues to use the outcome matrix at UCLA to examine how people decide what to do in their relationships. I urge you to work through the potential outcome values shown in each matrix presented in the chapter. The placement of the numbers at the intersection of two behaviors reflects Thibaut and Kelley’s conviction that our relational outcomes are always linked with the actions of others.
The column headings show the two choices open to Smith; the row labels describe Hickock’s same options. The four cells inside the box reveal the consequences of the various behavioral combinations. Smith’s outcomes are in the upper right corner of each cell. Hickock’s are in the lower left portion. For example, if Smith confesses while his partner stands mute, he goes free and Hickock dies.
The numbers in parentheses are attempts to quantify the values of different outcomes. Thibaut and Kelley let a single number represent the rewards minus the costs of a given course of action. For example, Smith might mentally sum up the benefits of a life sentence as a + 14. He’d have the relief of escaping execution, a chance for human contact, plenty of time to watch TV, and perhaps a shot at parole later on. Of course the downside of a life term in jail would cancel out most of those benefits. Smith’s costs, which he might mentally rate as a - 11, would include permanent loss of freedom, guards’ continually telling him what to do, boredom, and fear of violence from other prisoners. Since the outcome of an interaction equals rewards (+ 14) minus costs (- 11), Smith would regard the consequence of a mutual confession as a bleak +3.
The idea of totaling potential benefits and losses to determine behavior isn’t new. Since philosopher John Stuart Mill stated his philosophy of utilitarianism,’ there’s been a compelling logic to the minimax principle of human behavior. The minimax principle claims that people seek to maximize their benefits and minimize their costs. So the higher the number in an outcome matrix, the more attractive the behavior that might make it happen.
It would be nice if every interaction offered both parties a chance to get their optimum outcome at the same time. Unfortunately, the world’s not set up that way. As with the prisoner’s dilemma, there’s the potential for one person’s gain to come at the other’s expense. Thibaut and Kelley describe the prisoner’s dilemma matrix as "bilaterally discordant" and believe it offers a good way to study conflict between people.
Social exchange theory assumes that we can accurately anticipate the payoffs of a variety of interactions. Our minds are like computers, and a computer analysis is only as good as the data that are fed in. Garbage in; garbage out. To the authors of the theory, the data we get are remarkably reliable. Not only can Hickock see that twin confessions will result in a life sentence for him, he can also understand that they’ll produce the same effect for Smith.
The 2 x 2 matrix plots either-or decisions, but most human encounters require a much larger map to represent the multiple options of both parties. Despite this complexity, Thibaut and Kelley believe that members of a dyad, as well as outside observers, can realistically grasp the potential outcomes shown across the grid. In addition, the participants have the sense to choose what’s best.
The best choice is not always the one associated with the highest number on the board. Take a look at Smiths and Hickock’s options. There’s no doubt that freedom is the most desirable outcome, yet if each man confesses in the hope of avoiding prison, they will both end up behind bars for the rest of their lives. As is the case when playing a game of checkers, it’s not enough to know you want to advance your marker to the far side of the board to gain a king; you also must credit your opponent’s desire to do the same. Success requires that you take into account what the other player is likely to do.
The need for a strategy of anticipation leads some to refer to this exchange approach as "game theory." However, Thibaut and Kelley want to avoid the head-butting, win-lose, adversarial tone that goes with seeing the other as an opponent in a game. The internal cost of conflict cuts into the worth of an outcome; its resolution is value added. For this reason, they prefer the term interdependence theory. Whatever label we use, successful players learn to synchronize their moves with the actions of others.
COMPARING THE RESULTS: IS EVERYBODY HAPPY?
It may have occurred to you that a life sentence could mean something quite different to Smith than to Hickock. Social exchange theory presents two standards of comparison by which to evaluate a given outcome, whether in prison or under more normal circumstances. The first benchmark deals with relative satisfaction-how happy or sad an interpersonal outcome makes a participant feel. Thibaut and Kelley call it the comparison level.
A person’s comparison level (CL) is the threshold above which an outcome seems attractive. If your CL for clerical employment is an hourly wage of $8, you would be satisfied working for $9 an hour but feel exploited if you received only $7 for your labor.
Satisfaction depends on expectation, which is shaped by prior experience, especially gripping events of the recent past. A run of bad outcomes can make previously distasteful results more palatable. A string of successes whets the appetite for a gourmet feast.
If Smith’s memory is filled with killers who either go free or merely get token jail terms, his CL will be at + 7 or so, and a life sentence rated at + 3 will seem a cruel joke. In spite of his murderous brutality, even a moderate term in jail will strike him as unfair. Hickock, on the other hand, may see execution as the likely result of committing homicide. As a former convict, he knew men who went to the gallows. With a CL of only +2, a life sentence at +3 might look quite attractive.
COMPARING THE RESULTS: IS EVERYONE GOING TO STICK AROUND?
Thibaut and Kelley suggest that there is a second standard by which we evaluate the outcomes we receive. They call it the comparison level of alternatives (CLalt), and the level is pegged by the best payoffs available outside the current relationship. CLalt is the worst outcome a person will accept and still stay in a relationship. As more attractive outside possibilities become available, or as existent outcomes slide below an established CLalt instability increases. This may sound like a stock market analysis rather than a description of interpersonal relationships. Not surprisingly, some advocates of a social exchange approach refer to it as a "theory of economic behavior."
CLalt, doesn’t speak to the issue of attraction or satisfaction. A woman could be happy in her job, yet leave for a new position which offers even higher pay, better working conditions, or a more interesting assignment.
Relative Value of Outcome, CL, CLalt State of the Relationship
Outcome > CL > CLalt Satisfying
Stable
Dependent
Outcome > CLalt > CL Satisfying
Stable
Nondependent
CLalt > CL > Outcome Not satisfying
Break relationship
Happy elsewhere
CLalt > Outcome > CL Satisfying
Unstable
Happier elsewhere
CL > CLalt > Outcome Not satisfying
Break relationship
Continue unhappy
CL > Outcome > CLalt Highly unsatisfying
Can’t break away
Dependent and unhappy
Figure 16.2
Six Relational Typologies (adopted from Roloff, Interpersonal Communication: The Social Exchange Approach.)
Conversely, she might remain in a loveless marriage because the high cost of exit lowers the attractiveness of the best outside alternative. Social workers describe the plight of a battered wife as "high cost, low rewards." Yet she often remains with her abusive husband because the option of being alone i