Chapter 4- Group Conflict


1. The nature of conflict.

2. The use of experimental games to study conflict.

3. How people usually behave in group conflict situations.

4. How communication influences these behaviors.


A great deal of literature and research exists regarding the concept of conflict, and in recent years in particular, the topic has been gaining popularity. More and more universities are offering courses that are devoted entirely to the subject of conflict. Scholarly journals such as the Journal of Conflict Resolution are dedicated to studying the topic, and other journals (Journal of Peace Research, Journal of Conflict Management and Peace Science) discuss the subject as it relates to various fields. For centuries people have been studying conflict as it relates to all human conditions, including conflict within and between people and conflict within and between groups, organizations, or nations.

Such a widely researched topic cannot be covered exhaustively in a book such as this. Therefore, this discussion will mostly be limited to the work of scientists who are interested in group conflict situations. Their research seeks to discover how people act when there is conflict within groups. Do people tend to compete or to cooperate? In order to examine this topic researchers often employ games in their work. Our goal in this chapter is to describe how experimenters have performed these studies and the major findings that their research has provided.


Each of the perspectives toward small groups that is described in Chapter 1 has its own definition of "conflict." Rather than attempting to cover all these definitions (see Fink, 1968, for a review), we will focus on the definition created by Morton Deutsch (1973). Deutsch was a student of Lewin and an important contributor to our knowledge of conflict. His concept of conflict, using the relational perspective, is the most relevant definition for this book. As Chapter 3 explained, Lewin's field theory concerns the ways in which "forces" move the "self" around its "life-space." Deutsch felt that two people are interdependent when each can affect the other's life-space and the field of forces within it.

In Chapter 1, we saw the difference between cooperative situations and competitive situations. We can now restate that difference using the terminology of field theory. Imagine that Person A and Person B both attempt to reach goals in their individual life-spaces. Cooperation and competition are relevant when A and B, in trying to reach their own goals, can affect each other's progress. In cooperative situations, if one reaches his or her goal, the other is more likely to reach his or hers. In competitive situations, if one reaches his or her own goal, the other is less likely to reach his or hers. Harold and Leonard are playing Frisbee. The goal is to throw and catch the Frisbee 50 times in a row without it dropping to the ground. Each can reach his goal only if the other does. This is a cooperative situation. In contrast, Amy and Lisa are playing chess. The goal of each is to win the game. Each can reach her goal only if the other does not. This is a competitive situation.

Conflict can occur in both competitive and cooperative situations. To show this, let us review the matrix game presented in the description of Deutsch's relational definition for "group" in Chapter 1. Two people play the matrix game. Each must choose between two possible moves: "Yes" and "No." After both move each receives a certain number of points, which depends both on their own move and on the other's move.

Scientists who examine the implications of matrix games are called "game theorists." Game theorists make several assumptions about how people act in game situations. They assume that each person knows all the moves that both they themselves and their opponent can make, along with the number of points that will result from all possible combinations of moves. They assume that people consider only the number of points each will make in the short run when deciding which move to make. Finally, they assume that the goal of game players is to win as many points as possible for themselves, while not caring about how many points their opponent wins. In other words, game players are seen as having "individualistic" orientations.

Conflict, according to Deutsch, exists whenever Person A and Person B choose actions that are incompatible. In other words, conflict exists when Person A makes moves that make it harder for Person B to reach his or her goal.

In a competitive situation, as shown in Figure 4.1, Deutsch's version of conflict is most apparent. The normal goal is to win as many points as possible in the game. Any move by Person A to reach his or her goal of having the most points will make it more difficult for Person B to reach his or her goal of having the most points, and vice versa.


Person A

Person B


Yes -10


No 10

Figure 4.1 seems to present a clear picture of a situation that leads to conflict. However, Deutsch distinguished between "competition" and "conflict." For him, conflict was a broader concept than competition. Conflict can be found in competitive situations, but the two concepts are not simply equal. As has been stated, conflict occurs when two people choose actions that are incompatible. They may have the same goals, but because they don't agree on what action to take, it is difficult for them to reach their goals.

According to Deutsch, conflict can also occur in a cooperative situation. The matrix in Figure 4.2 illustrates a cooperative game matrix in which supposedly a move by Person A should help Person B move toward his or her goal also. However, this is true provided that both players make the same move. If they are unable to work out just which moves to make, conflict will result.


Person A

Person B


Yes 10


No -10

For an example outside of the numbered matrix, consider two people who have a goal of crossing a river. Both persons should be helping each other to reach the same goal cooperatively. However, Person A may vote "Yes" for a plan to use a canoe when Person B votes "Yes" for a plan to use a powerboat. Their actions are then incompatible, and they are in conflict.

Much conflict within small groups occurs in a cooperative setting. Each member may have the same goal but prefer different courses of action to reach that goal. Agreement on the group's course of action is imperative to avoid conflict. An important point to stress here is that conflict has many constructive and positive aspects in a cooperative group setting. This is particularly true within task-oriented groups. Conflict usually increases the involvement level of group members. Arguments over courses of action often lead to better decisions. Also, successful resolutions of conflict can lead to increased member satisfaction and group cohesiveness.

However, much group conflict is competitive. Different group members often have individual goals that are inconsistent with one another. Members may fight with each other for power over the group decision, or compete for the friendship or respect of other group members. Group activity may lead to the group gaining some profit or reward, and group members may compete over that reward. For example, Mario and Guido have worked together to make a hundred dollars shoveling snow for their neighbors. Now they have to decide how to divide that money between them. Mario might claim that he deserves most of it because he did most of the shoveling, but Guido might argue that he was the one that got most of the business for them.

To summarize, we can make a distinction between cooperative conflict and competitive conflict. In cooperative conflict, the attempt by one person to reach their goal helps the others reach theirs. However, one person's routes toward their goals interferes with the others' routes to theirs. In competitive conflict, both the goals and the routes toward the goals are inconsistent. Different situations can influence people toward different types of conflict. The game matrix shown in Figure 4.1 will lead people toward competitive conflict, while the game matrix in Figure 4.2 could bring about cooperative conflict. Some situations can be responded to either competitively or cooperatively. Because people can act either way, they have come to be known as mixed-motive situations. Next, we will examine some games that represent mixed-motive situations.


The Prisoner's Dilemma Game

Imagine the following situation:

The police arrest two people. They suspect that the two have committed a serious crime. However, the authorities only have enough evidence to convict the pair of a minor offense. The police decide to conspire with the district attorney to try to get confessions from one or both prisoners. The officials approach the criminals in a certain way. They take the prisoners into two different rooms. The district attorney tells them separately that they each have two options. They can either confess or not confess to the serious crime. If neither confesses, the officials will go ahead and charge and convict them of the minor crime. Each will receive a light sentence of one year. If both criminals confess, the authorities will charge each equally for the serious crime. The district attorney will recommend the relatively lenient sentence of five years apiece. However, if only one confesses, the situation changes. In that case, the one who confesses will receive a very light sentence of only three months as a reward for turning "state's evidence." On the other hand, the criminal who did not confess will get the maximum sentence of 10 years. The criminals find themselves in a dilemma.

The circumstances outlined above are equivalent to the game matrix shown in Figure 4.3. The negative numbers refer to the time, in years, that the players would lose due to being in prison if they make a certain move.


Prisoner One


Two Yes -1


No -1/4

As you can see, the prisoners are in a difficult situation. It is to their mutual benefit not to confess. They will both receive light sentences of a year apiece if they do not. However, each is tempted to betray his or her partner because the one who turns state's evidence will escape with the lightest possible sentence. There are further complications. Each criminal does not know what his or her partner is saying to the district attorney. If one does not confess, he or she would worry that perhaps the other partner had confessed. Each criminal knows that the one who did not admit to the serious crime would receive the maximum possible penalty.

Therefore, the game poses a difficult problem for the prisoners. Should they confess? It is mutually advantageous for both not to do so. However, there are also strong pressures for each to admit guilt and turn state's evidence. In addition, there is no way for the prisoners to tell whether either has betrayed the other. There is no clear solution to the problem.

We call the problem the "Prisoner's Dilemma Game." This name has come to cover various games of this type. These games have become the principal device for experimental research in interpersonal conflict.

Weaknesses of the Game

Games such as the Prisoner's Dilemma Game have both strengths and weaknesses when scientists use them as experimental devices. The major weakness is that they are based on game theory assumptions about conflict. For example, game players are believed to know all the actions that they themselves and their opponents can take. Game players are also assumed to know all the outcomes that will occur for these actions. For example, players who participate in the game we described know that they have two options, "confess" and "not confess." If they confess, they will spend either three months or five years in prison, depending on whether the other player confesses. They have no other options.

To some extent this situation is not realistic. In "real world" conflicts we may not know all the actions we or our opponents could take. Further, we may have no idea what the results of our decisions might be. We may not know what kind of result a confession will bring us. Hence, game players have more knowledge than people in real circumstances have. This difference may make the Prisoner's Dilemma Game problematic as a valid representation of "real world" conflict. However, most scientists believe that the game's positive features outweigh this negative feature.

Positive Features of the Game

The Prisoner's Dilemma Game is an example of what scientists call a mixed motive game which can be played cooperatively or competitively. The most important positive feature of the Prisoner's Dilemma Game is the internal conflict that the dilemma creates in players. Participants feel torn between the desire to be cooperative and the temptation to compete. The game allows us to observe this conflict. We can discover which option the players tend to follow most. In addition, because it is a controlled situation, we can manipulate the conditions under which the participants play the game. By doing so, we are able to discover if a certain condition can change the player's tendency to cooperate or compete.

Scientists using the game measure their experiments through the "tendencies" that they observe. Will a person have competitive tendencies in a conflict situation? Or will he or she usually cooperate? Only over a period of time can we see what players "tend" to do. Therefore, the experiments usually consist of a series of repetitions of the game, called "trials." Experimenters can then add manipulations to the game to see if these have any impact on game players' cooperativeness or competitiveness over time.

Minas et al. study. An example of a manipulation can be seen in a study by Minas, Scodel, Marlowe, and Ranson (1960). Their participants played 50 trials, using one of several Prisoner's Dilemma Game matrices. Here are two of the matrices used in the study.






These matrices differ in whether they reward cooperative or competitive game playing. Look at the matrix on the top. If one of the players plays competitively, then the other player will get one point for playing competitively and zero points for playing cooperatively. If, instead, one of the players plays cooperatively, then the other player will get five points for playing competitively and three points for playing cooperatively. Therefore, no matter how one player plays, the other would always make more points by playing competitively.

Now look at the matrix on the bottom. In this case, if one player plays competitively, the other player gets zero points for playing competitively but three points for playing cooperatively. If one player plays cooperatively, then the other gets three points for playing competitively but four points for playing cooperatively. Here, no matter how one player plays, the other always makes more points by playing cooperatively.

We would then expect that players would be more likely to cooperate when playing the Prisoner's Dilemma Game matrix on the top than when playing the game matrix on the bottom. The participants in Minas et al.'s study responded consistently with this expectation. When playing the top matrix, participants competed in 62 percent of the trials. When playing the bottom matrix, they competed in 47 percent of the trials. Although the difference between these percentages is not huge, it is large enough for us to conclude that people's behavior in conflict situations is affected by the extent to which the situation rewards competitive or cooperative action.

Dozens of studies have been performed using the Prisoner's Dilemma Game, and these studies have greatly contributed to our knowledge about conflict. However, the Prisoner's Dilemma Game is an example of conflict between people or between groups. We need different types of experiment games to examine conflict within groups. We will now turn our attention to two different types of mixed-motive conflict situations that can exist within groups, and how these can be studied using experimental games.

Resource and Public Goods Dilemmas

Resource Dilemmas

The residents of Everytown have a problem. Like all other communities, they have access to enough electrical power for most situations. Sometimes, however, they cannot get as much electricity as they need. An example might occur when a heat wave hits. Each resident of Everytown would then be tempted to turn their air conditioners to the "high" position. They would expect that this would help them stay cool. However, they might have forgotten that if everyone does the same, the demands of all of the machines working at the same time may overtax their power reserves. This would result in a "brownout" and resulting discomfort for everyone. If, instead, everybody left their air conditioners at a lower setting, they would be less comfortable in the short term, but the power demands would be less. This means that all of the residents would be able to stay somewhat content for the long term.

During a heat wave, Everytown experiences an example of a resource dilemma. Resource dilemmas occur when a group of people share a limited "pool" of a slowly-renewing resource. Group members can individually decide how quickly they will "harvest" from the resource pool. It is to each member's short-term advantage to harvest as much as they can from the resource pool. However, they cannot escape the fact that everyone in the group is competing for that resource, and if everybody acts competitively by harvesting large quantities, they will use up the resource faster than it can be renewed. The resource pool will be used up much more quickly than it needs to be. If, instead, each person acts cooperatively and harvests a limited amount, the resource pool can maintain itself much longer. Each group member would then be able to take advantage of the pool for the long term.

You can see why Everytown is one example of a resource dilemma. It takes a long time to generate a significant amount of electricity, so electrical power is an instance of a slowly-renewing resource. The "resource pool" of electricity can be only be maintained for long periods of time if usage is not too great. There are many other examples of resource dilemmas. Consider the fish industry. A particular type of fish, for example tuna, is a slowly-renewing resource, and if fishing boats do not harvest more tuna than are reproduced, then people will have it to eat for decades to come. If, instead, the fishing boats overharvest, people will have a lot of tuna for a short time, but before long tuna will become rare.

The resource dilemma situation is a very relevant model for the world as a whole. It illustrates the serious problems that the earth's population faces. As the population grows and continues to deplete the world's natural resources, it will face more and more resource dilemma situations.

A Resource Dilemma Game

In order to examine how the members of groups act when faced with a resource dilemma, researchers have developed various games. One example of a resource dilemma game was invented by Brechner (1977). He had groups face a situation in which they had a pool of resources. Each group member could take from the pool whenever he or she wanted. The players kept track of the resource pool by looking at a bank of electric lights. Brechner created one game with 24 lights and another with 48 lights. Each light represented part of the supply pool. A bulb was lit when its part of the resource was "available." The bulbs would go off, one by one, as the players took from the supply. However, the pool would replenish itself and the lights would go on again after a time. The researcher gauged the amount of pool replenishment according to how much of the resource the players left in the supply pool. The less the amount of supply, the fewer lights would go back on.

Consider the example of a fish pond. The fish are akin to the lights in Brechner's experiment. Every time someone takes a fish, there is one less of the "resource" to reproduce more fish in the future. Let us say that there is a pond with 25 fish in it, and five people fishing from it. In the course of a week, these original 25 can reproduce 5 more fish. If each person were to take five of the fish, 20 of the 25 would be gone, and the five that are left could only reproduce one more fish in a week. There would only be six in the pool for the people to share during the next week. If, instead, each person harvested only one fish a week, the five would have 52 fish apiece over the course of a year. In addition, they would still have the original 25 in the pond.

Brechner's game situation was like the fish pond example. It is clear that there is a great incentive for each player to take a large amount of resources before the other players have a chance to do so. However, the fish pond example showed how far-sighted people can end up with many more fish than those who fished quickly. Similarly, a shrewd player in Brechner's game realizes that he or she will have a greater portion of supplies eventually if the pool can replenish itself steadily. The optimum way to play the game would be for the players to cooperate by taking out resources slowly. This would allow the supply pool to replenish itself as quickly as possible. Players must decide between long-term, greater gain or short-term, lesser gain.

How did the players respond to this conflict situation? Results showed that players who played the 24-light version of the game usually succumbed to immediate pressures and decided upon the short-term, lesser gain. They used up the entire pool very quickly. In contrast, test participants who played the 48-light game responded differently. They were far more likely to cooperate and keep the pool replenished. Each participant in the group ended up much closer to the optimum amount of resources.

Public Goods Dilemmas

Everytown has another problem. A foundation has donated enough money to start a public radio station. It is, however, up to the townspeople to contribute enough money every year to keep the station on the air. Now, consider one radio listener, Jeannie. Once the station is on the air, Jeannie can listen to it as much as she wants. It does not matter whether or not she decides to contribute money to the station. In other words, Jeannie has the option to "free ride"; enjoy the station without contributing to it. In fact, it would be to her personal advantage to free ride, because she would be saving her money. However, what happens if too many people also decide to free ride? Not enough money will be contributed in total to maintain the station. It will go off the air, to everybody's detriment.

This is an example of a public goods dilemma. A public good is some product or service that can only exist if a group of people choose to contribute a sufficient amount to it. As long as it exists, all group members can take advantage of it, whether they choose to contribute or free ride. It is to each group member's short-term advantage to free ride and take advantage of others' generosity, in effect acting competitively. However, if too many people free ride, then the public good will not be able to continue. Rather, if enough people act cooperatively and contribute a sufficient amount, the public good can maintain itself indefinitely. In this case, everybody can enjoy the benefits of the public good for the long-term.

A public radio station is just one example of a public good. There are many others. A second good example is a charity. People can choose whether or not to contribute, and their choice has no impact on whether or not they get help from the charity when they need it. Therefore, it is to everyone's advantage in the short run not to contribute, but unless some people give money the charity cannot help those in need. A third example is a voluntary labor union. Labor unions require dues to exist, and in many cases it is up to the individual worker whether or not to join the union and pay the dues. However, if the union is successful at getting the workers a raise, everybody gets the benefits of it, whether or not they have joined. Therefore, one saves money by not joining, but if too few people join, the union will not be effective.

A Public Goods Game

As with resource dilemmas, researchers have invented different games to examine how people act when faced with a public goods dilemma. One example of a public goods game was created by Braver and Wilson (1986). Research participants were assigned to membership in a nine person group. Each of the participants was given five dollars. They were then given a choice of whether to keep the money or contribute it to the group. If at least half of the group members contributed, then all nine group members would be given ten dollars. This ten dollars is the public good in the game. If less than half of the group members contributed, then everybody would get only the money they had.

Let us look at the possible outcomes of this game. Cecilia has the choice of keeping the five dollars or contributing it. If enough other group members contribute to obtain the public good, Cecilia will end up with fifteen dollars (her five plus the public good) if she keeps her five dollars and ten dollars if she contributes. If too many group members keep their money, Cecilia will end up with five dollars if she keeps her money and no money at all if she contributes. As you can see, Cecilia's best outcome is keeping her money if enough other people contribute to secure the ten dollar public good. However, she is better off contributing the money if she thinks it will help the group achieve the public good than keeping it if the group is unsuccessful.

This game is analogous to the earlier public radio station example. Jeannie's best outcome is for the radio station to stay on the air while she keeps her money, but she is better off contributing the money to keep the station on the air than keeping the money but losing the station. Cecilia's predicament in Braver and Wilson's game is similar. What do people choose to do when faced with this dilemma? In this study, 48 percent of the people decided to contribute their money, leading to 60 percent of the groups attaining the ten dollar public good for its members.


It should be obvious that resource dilemmas and public goods dilemmas are opposite sides of the same coin. Resource dilemmas start with a public pool, which group members can choose to take from and put in an individual "account." Public goods dilemmas start with individual accounts, from which group members can choose to contribute into a public pool. In both cases, if one group member acts competitively, it works to that member's short-term advantage, but if too many members acts competitively, everybody is hurt in the long term. In both cases, if one group member acts cooperatively, it will hurt that member in the short term, but if enough members act cooperatively, everybody is helped in the long term. Although there are certainly differences between these two types of situations, they are similar enough for our purposes for us to consider them together as two examples of social dilemmas.

As we saw in the Brechner and Braver and Wilson studies, when faced with these social dilemmas, some people act competitively and some people act cooperatively. We saw the same result for the Minas et al. (1960) study of Prisoner's Dilemma Game behavior earlier. As with the Prisoner's Dilemma Game, researchers can manipulate various aspects of public goods and resource dilemma games to see if these changes have an impact on game players' cooperativeness or competitiveness. Researchers can also measure people's responses to these situations to determine if these also have a significant effect. In the next section, we describe five important factors that affect people's actions in these games.



As we discussed in Chapter One, as they interact over time, group members develop shared expectations and evaluations about correct ways of acting. We call these expectations norms. Some norms have become so widely shared that they have become part of our culture. Because they are so universally accepted, people often apply them when entering into new groups. Two of these norms are particular relevant to how people act when involved in social dilemmas. These are the "equality norm" and the "equity norm." The equality norm states that all members of a group are equally important. Therefore, if the group gets a certain reward for its work, everyone in the group should receive an equal share of that reward. The equity norm states that the "value" of each member of the group is based on what they put into the group. In this case, if the group has a reward to share, the proportion of rewards that each member gets should equal the proportion of resources that he or she put into getting the reward.

As you can see, the equality and equity norms can lead to inconsistent expectations. Imagine that Irene and Todd built a bridge over a river, and charged a toll to people driving over their bridge. Irene put in sixty percent of the money for building the bridge, and Todd contributed forty percent of the money. If they were to divide the money they receive from tolls based on the equality norm, each would get half. If, instead, they used the equity norm, then Irene would get sixty percent of the money and Todd would get forty percent. The amount of money each gets depends on which norm they use.

Social dilemmas are situations in which the equality and equity norms can be relevant. In a resource dilemmas, should the pool be divided up equally or in some other fashion? In a public goods dilemmas, should everyone contribute the same amount or not? It seems that, all else being equal, people use equality norms in social dilemmas. Samuelson and Allison (1994) devised a resource dilemma game in which each group member had only one opportunity to harvest from the resource pool, and that if everyone's request summed to more than the pool contained, nobody would get any points. Each was told that they would be the first in their group to harvest from the pool. Some members were told that the reason they were first was the result of either their birth date or their performance on what was obviously a phony test of their skill on a task. As there were six members of each group, if each participant were to take one-sixth of the pool, they would be following the equality norm. It turned out that the average harvest taken by these participants was almost exactly one-sixth of the pool.

However, sometimes all things are not equal. If some group members believe that, for some reason, they deserve more points than other group members, they will act as if they are following the equity norm instead. Samuelson and Allison told other participants that the reason they were first to harvest was the result of either a coin flip or the result of what looked like a real test of their skill. In these cases, participants took much more than one-sixth of the resource pool. Samuelson and Allison then asked another group of participants to judge whether the reasons given for being allowed to harvest first were good or bad. The participants felt the phony test and birth date were bad reasons for being allowed to harvest first, whereas the coin flip and real test were good reasons. Given these judgments, Samuelson and Allison concluded that participants who went first because of either the real test or the coin flip believed that they "deserved" to go first and take more than one-sixth of the resource, but participants who went first because of either the phony test or their birth date believed those to be unfair reasons for going first and took an equal share of the pool.

Analogous results have been found in studies of public goods dilemmas. The point is that group members in resource dilemmas will take more than an equal share, and group members in public goods dilemmas will give less than an equal share, if they believe that their unequal share is for some good reason justified.


One possible justification for using an equity norm is what we will call "asymmetry." In public goods situations, asymmetry can occur in two ways. One way is if group members have unequal endowments that they can contribute. Going back to Everytown, if Glenda has much more money than Jeannie, Glenda is able to contribute more to the public radio station. A second type of asymmetry in public goods situations is if group members receive unequal benefits from the public good. For example, Jeannie may have much more time to listen to the public radio station than does Glenda. Research has shown that, with either type of asymmetry, group members follow equity norms. Beginning with the first type of asymmetry, van Dijk & Wilke (1994) gave some group members three times as many points as the other group members. In half of the groups, they told the participants that the reason for the inequality was that the "wealthier" members had worked three times as long on an earlier task than the "poorer" members had worked. In the other half of the groups, they told the participants that all had worked on the task the same amount of time. In both cases, the wealthier members contributed many more points than the poorer members, as the equity norm would suggest.

However, the difference between the contributions was particularly large for those groups told that they had worked the same amount of time. In other words, when told that they had done an equal amount of work, the wealthier members gave more and the poorer members gave less than when they were told that they done an unequal amount of work. As with Samuelson and Allison's study, it seems that giving a good reason for having an advantage leads to different effects than giving a bad reason. Those group members told that they had worked an equal amount of time apparently thought that this was a poor reason for the difference in wealth. As a result, the wealthier members believed that they ought to give more to the public good, and the poorer members believed they were justified in giving less, than when they had been told that they had worked an unequal amount.

The second type of asymmetry in public goods dilemmas is inequality in benefits. van Dijk & Wilke (1993) performed a study that was similar to the one we just described, and found that group members who expect to benefit more from a public good will contribute more to obtain it. Analogously with their other study, this tendency was greater when all of the group members had worked an equal amount of time for their benefits than when the wealthier people had worked longer for their endowment than had the poorer people.

Turning to resource dilemmas, there are, once again, two types of asymmetry. In one type, different members of a group are allowed to harvest different amounts of the resource pool. It is as if some people can take four fish out of a pond each week, but others are only allowed to take two. Samuelson and Messick (1986) allowed half of each group to take up to 30 points from a resource pool, but the other half to take no more than 10. Not surprisingly, the people allowed to take more resources generally did so. In the other type of asymmetry in resource dilemmas, group members can take out the same amount, but the "value" of their harvest differs. The analogy here is to people who can all take four fish out of a pond each week, but while half are paid two dollars for each fish, the other half are only paid one dollar per fish. Budescu, Rapoport, and Suleiman (1990) had five-member groups play a resource dilemma game, in which members got different percentages of their harvest. The group member with the worst "exchange rate" only got 1/4 of the value of their harvest, whereas the member with the best rate got 4 times the value of their harvest. Unlike the other types of asymmetry, in this case the group members attempted to maintain an equality norm in their harvesting. The better the exchange rate, the less the group member harvested.

Group Size

Consider the following two resource dilemmas. In order to continue on the air, Everytown's public radio station must receive a total of 100,000 dollars from the town's 20,000 residents. To keep broadcasting, Largertown's public radio station needs 200,000 dollars from the town's 40,000 citizens. In each case, the stations need an average contribution of five dollars per person. Which town's radio station is more likely to survive?

According to research, the answer is clear. In smaller sized groups, members contribute more and are better at obtaining public goods than in larger sized groups. Smaller sized groups also lead to less harvesting and are better at maintaining resource pools than larger sized groups. The more complicated question is why this is so. There have been two proposed explanations for this tendency.

The first proposal was made in a very influential book about public goods dilemmas written by Olson (1965). Olson believed that, when in a public goods dilemma, most people see the value of obtaining the public good as large enough to justify making a contribution to it. However, they will always be uncertain that enough will be contributed by everyone in their group to obtain the public good. Nonetheless, anything that increases their confidence in the other members' generosity will make it more likely that they will contribute. Consider Glenda's decision on whether to contribute to Everytown's radio station. Glenda is wealthy, and can afford to contribute 10,000 dollars, but she does not want to do so unless she feels that the funding goal will be reached. Her contribution alone would bring the station one-tenth of the way to its goal of 100,000 dollars. If, however, she lived in Largertown, her contribution would only bring its station one-twentieth of the way to the 200,000 dollars it needs. Her contribution would increase the likelihood of goal attainment much more in Everytown than in Largertown. Therefore, she would be more confident of attaining the funding goal in Everytown, and more likely to contribute. In Kerr's (1989) terminology, her "self-efficacy," or her belief in her own personal control over the situation, would be greater in the Everytown situation than in the Largertown situation. Kerr's own research suggests that people in smaller groups have greater feelings of self-efficacy, which increases the odds that they will contribute to a public good. A similar explanation can be made for resource dilemmas. In smaller sized groups, people feel more personal control over the resource pool, and so they will be more confident that their actions can lead to the pool's maintenance. As a result, they take less from it.

The second proposal was made in a thoughtful essay about resource dilemmas by Messick and McClelland (1983). They felt that group members focus on the size of a resource pool and forget about how many people are harvesting from it. Suppose there is one pond with 25 fish in it, and five people fishing from it. There is a second pond with 50 fish in it, and ten people fishing from it. In both cases, there are five fish per person. However, if they only notice the number of fish, then the people at the second pond will harvest more, and so deplete their resource pool faster than those at the first pond. An analogous explanation can be made for public goods dilemmas. Glenda may be more likely to contribute in Everytown than in Largertown just because Everytown's required amount of money is smaller, while forgetting the fact that there are twice as many people in Largertown that can contribute to its station.

Both of these explanations have some validity. It is likely that group members' increased self-efficacy, and their tendency to focus on the size of the resource pool or public goods goal while ignoring the size of their group, work together to increase cooperation in smaller sized groups.

Environmental Uncertainty

We have already examined the situation in which five people share a pond with 25 fish, that can replenish itself with 5 more fish a week. Let us now compare this situation to situations with "environmental uncertainty." For example, the five people may have been told that the number of fish in the pond is somewhere between 15 and 35, or that the replenishment rate is somewhere between 2 and 8 fish a week. How does that change harvesting behavior?

Rapoport and his associates have performed several studies of the first of these types of environmental uncertainty (e.g., Budescu, Rapoport, & Suleiman, 1990). In these studies, members of five-person groups were allowed to harvest from a resource pool up to thirty times if the pool lasted that long. Some of the participants were told that the pool contained 500 points (no uncertainty), others were told that it contained somewhere between 250 and 750 points (intermediate uncertainty), and still others were told that it contained somewhere between 0 and 1000 points (high uncertainty). The greater the range and thus the uncertainty, the greater the harvest and the worse groups were at maintaining the pool over time. Further, when the range was greater, group members made higher estimates of the size of the pool. It seems that, when there is environmental uncertainty, people tend to overestimate the size of the pool. They then use that overestimate in deciding on their harvest size, leading to overharvesting.

Research has also shown that uncertainty in the replenishment rate of a resource pool also leads to overestimating the size of the pool and greater harvesting. In addition, making the amount needed to achieve a public good uncertain leads to smaller contributions. In this case, people probably underestimate the needed amount and contribute accordingly.

Social Values

Back in 1958, Deutsch proposed the idea that there are three "motivational orientations" that a person can take toward conflict. To describe these, let us imagine a conflict between Wendy and Jeff. Wendy could enter this conflict with the attitude that she will try to reach her goals while also helping Jeff satisfy his. Deutsch called this attitude a "cooperative motivational orientation." In contrast, Wendy could enter this conflict with the attitude that she will try to reach her goals while preventing Jeff from obtaining his. In other words, Wendy wants to defeat Jeff in the conflict. Deutsch called this attitude a "competitive motivational orientation." In addition, Deutsch proposed that Wendy could enter this conflict with the attitude that she will try to reach her goals, while ignoring Jeff's goals altogether. Deutsch called this attitude a "individualistic motivational orientation." As you can see, the difference among the three lies in whether Wendy wants to help Jeff, do better than Jeff, or does not care about what Jeff wants.

Deutsch was primarily interested in how the conflict situation affects people's motivational orientations. Not surprisingly, a situation such as the one shown at the beginning of this chapter in Figure 4.1 would usually lead people to adopt a competitive orientation. Analogously, a situation like the one in Figure 4.2 would influence people toward a cooperative orientation. Deutsch did, however, describe other factors that might impact on motivational orientations toward conflict. One of these factors was a person's personality. Some people might characteristically adopt one of the orientations in most situations. This would influence their actions in mixed-motive conflict situations, such as the Prisoner's Dilemma Game, resource dilemmas, and public goods dilemmas.

Starting in the 1970s, researchers began to study people's characteristic "social values" toward conflict. They adopted a special game for the purpose. This game has come to be called the "decomposed Prisoner's Dilemma Game." Figure 4.5 is an example of a decomposed Prisoner's Dilemma Game. It is taken from Kuhlman and Marshello (1975), which has a good discussion of how this research is performed. If you were a study participant, you would be told to imagine that you were playing the game against another person. There are three options you can choose, labeled A, B, and C. Under each option is listed the number of points both you and the other person would receive if that option were chosen. You are then asked to indicate which of the options you prefer.


Self 50 40 40
Other 20 0 40

Each of these options can be paired with one of the three motivational orientations proposed by Deutsch. Let us see how. Suppose that Ed, Ted, and Ned play this game. Ed has an individualistic orientation, which means that he is interested in doing as well as he can, and does not care about how well the other person does. He would choose option A, because it provides "Self" with more points than the other two options. Ted has a competitive orientation, which means he wants to do better than the other person. Ted will choose option B, because he "wins" by more points from that option (40 points) than from the other two options (30 points for A, 0 points for C). Ned has a cooperative orientation, which means he wants both himself and his conflict partner to do as well as possible. Ned will choose option C, because the number of points "Self" and Other" receive together (80 points) is greater than the number of points both receive from the other two options (70 points for A, 40 points for B).

Figure 4.5 is only one example of a decomposed Prisoner's Dilemma Game, and there are countless other combinations of points for "Self" and "Other" that could also serve as decomposed Games. One should not take a player's response to one particular game as indicative of a characteristic social value, because the player might respond much differently to a different game. For that reason, researchers have generally asked participants to play fifteen different versions, and only considered a person to have a characteristic social value toward conflict if they showed the same orientation in at least nine of the fifteen games. In other words, Ted would be considered to have a competitive social value if he choose the competitive option nine or more times in the fifteen games.

It turns out that most people have characteristic social values toward conflict. If 100 people were to take the decomposed game test, probably 20 or 25 of them would turn out to be cooperatives, another 20 or 25 of them would turn out to be competitives, and still another 20 or 25 of them would turn out to be individualists. Interestingly enough, perhaps 5 or 10 of the others would display a fourth social value, by consistently choosing the option that helps the other person without considering their own outcomes. We would call these people "altruists." The rest of the 100 would not show a consistent orientation, so we would conclude that those people did not have a characteristic social value toward conflict.

Dozens of studies have been performed which have compared how people with these four types of social values toward conflict respond to regular Prisoner's Dilemma Games, resource dilemma games, and public goods dilemma games. Rather than describe any in particular, we will mention some of the more important findings (see Komorita & Parks, 1996, for a review). In many of these studies, altruists and cooperatives were described together as "prosocials" and individualists and competitives as "proselfs." We will use these descriptions in our account when it is relevant.

1. Prosocials are more cooperative game players than proselfs, harvesting less in resource dilemma games and contributing more in public goods dilemma games. Some researchers have compared people with all four types of social values. They have consistently found altruists to be the most cooperative game players, followed by cooperatives and then individualists, with competitives the least.

2. The difference between prosocials and proselfs in resource dilemma game behavior is particularly pronounced when the size of the resource pool is uncertain or is getting very small. In those cases, proselfs behave as if they are trying to maintain the pool, whereas proselfs try to grab as much as they can quickly.

3. When playing Prisoner's Dilemma Games, competitives seem to be unaffected in their game play by whom they are playing. They will play competitively no matter if their opponent is playing cooperatively or competitively. Cooperators, in contrast, base their behavior on their opponent's tendencies. They prefer to play cooperatively, but if they find that their opponents consistently compete, cooperatives will begin to compete too. Individualists are intermediate in this regard.

4. Prosocials attach more importance to long-term group outcomes and less importance to short-term individual outcomes than proselfs.

5. Prosocials expect other group members to be more cooperative than proselfs.

In this section, we have discussed the impact of five factors that influence whether people compete or cooperate when involved in social dilemmas. We chose these five because they have been well researched, but there are actually many other factors that can also affect how people act in these situations. Sally (1995) has reviewed many of these studies in order to determine the relative importance of these factors. Although several of them are critical, there is one more factor that is the most influential of all; the opportunity to communicate.


The research evidence shows very clearly that the opportunity for game players to communicate about the game they are playing increases cooperation among the players substantially. This is true for the Prisoner's Dilemma Game that we studied earlier in this chapter, and it is also true for social dilemma games. For one example, earlier we described Brechner's (1977) resource dilemma experiment, in which game players could harvest from a "resource pool" of either 24 or 48 electric lights. Players tended to overharvest and had difficulty maintaining the resource pool, particularly if there were only 24 lights in the resource pool. When they had the opportunity to communicate, the players were far more successful. In fact, the total replenishment for resource pool in the 24-light version of the game in which there was no communication was only 40.50 lights. When communication was allowed, this rate increased all the way to 212.84. In other words, communication led to such better resource pool maintenance that the players had more than five times as much total resource from which to harvest. For another example, we also described Braver and Wilson's (1986) public goods dilemma game, in which players were given five dollars, and promised an additional ten dollars if at least half of them contributed to the public good. As we said earlier, without communication, 48 percent of the people decided to contribute their money, leading to 60 percent of the groups attaining the ten dollar public good for its members. With communication, both the cooperation rate and the percent of groups attaining the public good went up to 75 percent.

As always, scientists were not satisfied with only demonstrating that communication has a big impact on social dilemma game playing. They also wanted to know why communication has that effect. Many researchers have performed studies in an attempt to discover the factors that produce this effect. We will discuss some of their findings now.

First, communication within the group must be relevant to the task for it to have an impact on cooperation. Dawes, McTavish, and Shaklee (1977) compared situations in which resource game players either talked about the game for ten minutes before playing it, talked about an irrelevant issue for ten minutes before playing it, or did not talk. Cooperation rate with no communication was 30 percent. With irrelevant communication, cooperation rate was substantially the same, at 33 percent. Relevant communication increased cooperation all the way up to 72 percent.

Second, although it must be task relevant, communication's effect is not just to help the players come to understand the game. Edney and Harper (1978) compared a situation in which resource game players could communication with one another with three other situations. These included a situation in which the experimenters gave game players a good strategy for playing the game, a situation in which the experimenters described to the game players what would happen if they overharvested and depleted the resource poll, and a situation in which there was neither communication nor information from the experimenters. Providing the warning led to the game lasting considerably more rounds (12) than the strategy (7.6) or nothing at all (6.4), but allowing communication was far more effective yet (18.90). Similarly, in a public goods situation, Kerr and Kaufman-Gilliland (1994) found that showing public goods game players comments supposedly made by previous game players, or allowing game players to overhear previous game players discuss the game, had no impact on cooperation, but that actual communication among the game players did.

Third, communication seems to increase cooperation when it leads to an expectation among members that attempts to cooperate by one group member will be reciprocated by the others. Chen and Komorita (1994) found evidence for this generalization in two studies in which group members made pledges to contribute to a public good. The amount contributed to the public good did not differ very much between situations in which the pledge was binding (3.61), nonbinding (3.06) and a situation with no pledges at all (3.81). However, when game players pledged a given amount and were bound to give the lowest amount pledged by any member, contributions rose to 6.14 in one study and 5.10 in the other. Further, when game players pledged a given amount and were bound to give the average amount pledged by all members, contributions rose further to 6.80. Chen and Komorita concluded that in these latter two situations, players knew that everyone was going to contribute the same amount. This knowledge motivated them to be willing to contribute more than the first three situations, in which the players knew that each of them would be giving a different amount.

Fourth, some theorists have proposed that communication's impact is to increase group cohesiveness and a feeling of group identity, which leads in turn to cooperation among the members. However, research has suggested that this proposal may be false. Chen (1996), in a study following up from the ones just discussed, discovered that communication occurring before pledges were made increased group identity but did not increase contributions to the public good. In the Kerr and Kaufman-Gilliland (1994) experiment we discussed earlier, communication increased both perceptions of group identity and contributions to the public good, but that perceptions of group identity and contributions were not related with one another.

Findings such as these suggest that the opportunity to communicate allows group members to form agreements and commitments with one another to cooperate. If these agreements and commitments are kept, a norm for cooperation emerges within the group, enhancing trust among the members (this process is described by Bornstein, 1992). Although this chain of events has not been directly examined in research, there is evidence about some of its links. For example, Ostrom and Walker (1991) watched the interaction among groups of public goods game players, and saw them reach agreements about how much to contribute that were usually adhered to. Members who did not give what had agreed upon were publicly described as "traitors" and "scumbuckets" in later discussions.


Experiments that examine how people behave when playing Prisoner's Dilemma, resource, and public goods games may seem to have little obvious relevance to how group members act when faced with a conflict about a decision they need to make. In fact, as we said earlier, one of the disadvantages of using games such as these in scientific research is that they make assumptions that are unrealistic. However, we also said that games have some advantages as tools for researchers. In particular, the way in which they lead to internal conflict between the desire to cooperate and the temptation to compete is very much like what often occurs in "real life" conflict. We believe that the results of research using these games does provide us with some lessons that can help us cope better with conflict within our groups.

First, the Minas et al. (1960) study showed how people are more likely to cooperate when the situations rewards them for cooperation than when the situation benefits them to compete. It follows that one way to help manage conflict in groups is to make the situation such that everybody wins through cooperation. Sometimes a group gets in a predicament in which it seems that, no matter what decision it makes, some members will be better off while others will be worse. It is as if there is a pie that the group has to divide up, and the bigger some members' pieces of the pie, the smaller the other members' pieces must be. Sometimes, this is the case. If a group has fifty dollars that it can split up among the members, then if we pay Paul more we will only be robbing Peter. At other times, however, if the group works really hard, it can find a way to make the pie bigger, so that everybody ends up ahead. The way to do this is to find out what each member really needs and why. We can imagine two cooks who each need an egg, but there is only one egg for them to use. Perhaps they will argue endlessly over who gets to use it, in which neither will finish preparing their meal. If, however, they discuss what each needs rather than fight, they might discover that one needs the yolk while the other needs the white. One egg is sufficient for both.

Second, as we discussed earlier, the equality and equity norms have become so widely shared that they have become part of our culture. As studies such as Samuelson and Allison's (1994) have shown, people naturally adopt these norms in game situations. There is good reason to believe that people adopt them just as naturally during real-world conflicts. The best way to manage a conflict is to make sure that members feel that they are getting a fair deal. Often, that means dividing up responsibilities and rewards equally. To use the example from the last paragraph, if there is fifty dollars to split up among the members of a group, then our first thought would be to give everyone an equal amount. However, we need to keep in mind that some members may have contributed more toward getting the reward, so that equity is sometimes a better solution. As the van Dijk and Wilke (1993, 1994) studies showed, people who have more to give to a group are usually willing to make larger contributions to it, but they may feel entitled to getting a greater share of any winnings the group achieves from its work. The research we have described implies that other group members will often see this entitlement as justified. The main point here is that conflicts among group members can often be managed by applying either the equality or equity norm, depending on which of the two seems more relevant to the situation.

Third, the studies on environmental uncertainty show that people tend to overestimate the size of resource pools or underestimate what is needed to achieve a public good. In short, they tend to be overly optimistic in both cases, and as a result they do not act as cooperatively as the situation requires. Given that uncertainty about the impact of people's behaviors is a hallmark of the real world, these findings are particularly important. As we mentioned when discussing the disadvantages of using games in research earlier, in real-life conflicts we may have no idea what the results of our decisions might be. If people tend to be overly optimistic about those results, then members of groups facing conflict in uncertain situations will tend not to contribute as much to their groups as is needed. It is then critical that group members be reminded of the uncertain nature of the situation, and instructed to consider the worse-case scenario as seriously as the best-case.

Fourth, the opportunity to communicate is a valuable tool for managing group conflicts. As described above, it provides the potential for forming agreements, coordinating actions, and gaining trust in other members. However, the opportunity for communication does not solve all conflict problems as well as it has solved social dilemmas in experimental research. Keep in mind that the experiments we have described included participants who have not worked together before and have no reason to distrust one another. Real-world conflict usually includes people who have worked together before and may have had bad experiences with one another. When previous events have led to a great amount of mistrust among members, the conflicting parties may interpret any attempts to bring reconciliation as a sign of weakness or as attempted trickery. Communication alone cannot bring trust and solve this problem. In addition, social dilemmas provide the potential for mutual benefit, but other conflicts are inherently competitive. In other words, sometimes groups have no option but to pay Paul by robbing Peter. Even honest and trusting communication can backfire when people truly have different interests causing the conflict. It will only clarify these differences and sharpen the conflict. Thus, communication is not a panacea for disagreement. Communication requires trust and some mutuality of actual interests between sides before it can bear fruit.


Conflict occurs when two people or two groups choose actions that are incompatible. In this way each action by one makes it more difficult for the other to reach his or her goal. Conflict can occur in competitive situations when both parties have incompatible goals. It can also occur in cooperative situations when both parties have the same goals but prefer incompatible ways of reaching these goals.

Experimenters have used games in order to research conflict. The Prisoner's Dilemma Game has been the principal device for research about interpersonal and intergroup conflict. Through the use of this game, we have learned a lot about interpersonal conflict; for example, how people's actions can be influenced by the extent to which the situation rewards competitive or cooperative behavior.

To study conflict within groups, researchers can use games designed to represent resource and public goods dilemmas. A resource dilemma exists when a group of people share a limited "pool" of a slowly-renewing resource, and each group member can individually decide how quickly they will "harvest" from the resource pool. A public good is some product or service that can only exist if a group of people choose to contribute a sufficient amount to it, from which all group members can profit from, whether or not they have contributed. In both of these dilemmas, it is to each group member's short-term advantage to compete by "harvesting" heavily from a resource pool or not contributing to a public good, but to everybody's long-term advantage to cooperate by restraining harvesting from a pool or contributing to a public good.

Because they are two sides of the same coin, research findings can usually be generalized across both of these types of "social dilemmas." For example, in both types, group members usually follow an "equality norm," by attempting to harvest or contribute an equal amount. However, when group members believe they are justified for harvesting more or contributing less than the average, they follow an "equity norm," by attempting to get out of the situation the same proportion they put in. This can occur when they have more to contribute or expect to get more out of a public good, or when they can take more out of a resource pool, than other group members. Smaller groups tend to harvest less from a resource pool and contribute more to a public good. Uncertainty about the size of a resource pool, the rate at which it replenishes, or the amount needed to achieve a public good, leads to greater harvesting and less contributing. Finally, there are individual differences between people who are inherently competitive, individualistic, cooperative, and altruistic, resulting in consistent differences in harvests and contributions.

The opportunity to communicate has been shown to substantially increase cooperation rates in social dilemma games. It seems that communication provides an opportunity for group members to form agreements and commitments with one another to cooperate. If these agreements and commitments are kept, a norm for cooperation emerges within the group, enhancing trust among the members.