Overview of the Game

“The Collective Action Game” divides a class into a number of teams, each of which represents an “aluminum ingot authority” for an aluminum producing nation.¹

Each team is charged with determining how many units of aluminum ingots to produce in its country over each of five rounds, with the rounds representing production years. For the sake of simplicity, teams are restricted to five choices as to the amount produced. The instructions state explicitly that each team is to attempt to make as much money as possible during the course of the game; nothing is specified about the desirability of doing better than other teams. The instructor distributes the rules, takes the aluminum ingot production orders, and calculates the world price for the material based on total class production according to a schedule distributed to students. The instructor may also calculate the amount each team has earned over every round based on tonnage produced and world price, less a fixed cost of production for each unit produced. This provides a running tally of how the groups are doing and facilitates inferences about the effects of different production decisions. The rules for the game and schedule for the determining the market price are included as Appendices A and B.

Under the market price schedule provided to students, all teams are advantaged if they act together as a cartel, restricting production to less than the maximum allowed under the rules. This is because reduced production can lead to a substantially higher world market price. However, any individual team maximizes profits as a consequence of producing as much as possible while the other teams cooperate to reduce production. Yet if every group follows such a strategy there is a glut of aluminum ingots, the world market price falls, and profits are low. In short, the game creates the classic N-person prisoner's dilemma in which the incentive for individual players to “defect” can lead to a collective result that is inferior for each team than mutual cooperation.

Aside from illustrating the collective action problem, one of the virtues of the game is its simplicity. The rules cover a single page and students can easily digest enough to play the game (if not to understand its subtleties) within a few minutes. No special equipment is required. The groups spend minimal time on secretarial tasks since their only obligation is to indicate how much they will produce each round. It also places little burden on students to perform well, unlike some other in-class exercises.²

I have used the exercise in both a three-hour graduate seminar and a one and a half hour undergraduate class. For the former, I take about an hour to run the exercise, after which I typically devote the remainder of the class to debriefing, reinforcing connections to the literature, and discussing the larger implications of the collective action problem. For the latter, I devote the entire class period to the exercise and debriefing, then devote a portion of the subsequent class to lecture about the collective action problem and its implications.

Students tend to find the exercise fun. I believe this is a good thing in itself, but the main advantage of students' positive experience is that it leads to a high level of engagement in subsequent discussion of collective action related issues.

Patterns and Lessons

I have used the exercise for about a dozen years and have observed a number of patterns in how students play the collective action game. First, it is common for students to be confused or highly uncertain about strategy through the first round. After the first round's results are announced, most students understand that they will do better if they work together to restrict production than if they all produce as much as they want. Naturally this is more commonly true for graduate students, but it may be true for undergraduates as well, in part because the more astute in each group will correctly diagnose the situation for others.

Second, usually at least one team elects to produce the maximum amount possible in the first round, leaving that team with the best relative score. Sometimes the team has made a deliberate choice to be a “defector;” at other times the choice appears to be less strategic. In any event, much of the subsequent dynamic involves how other teams react to the defecting team(s). Do they attempt to convince the group to follow a different path in the future? Do they attempt to punish the defector(s)? Do they ignore the defector(s) and do the best they can? I have seen each of these approaches, sometimes in a single playing of the game.

Third, at some point a student inevitably asks me whether it is okay to talk to the other groups; the exact timing of such an inquiry varies. I then ask the student if he or she observes a rule prohibiting such conversation. Finding no such rule, most students then begin to interact with other groups. However, classes vary dramatically with respect to how intense and systematic these efforts are. In conducting the game himself, Professor Rosenstone has been surprised at how long it often takes teams to talk to one another. He commonly tells students that given the difficulty of achieving cooperation with a highly educated group in a relatively non-stressful situation, they should avoid unrealistic expectations about how long it takes other groups to cooperate.³

Fourth, as the game progresses it is common for students to become aware that the last round will be different, since there is no subsequent opportunity to make a different production choice based on what other groups have done in the prior round. Sometimes this awareness contributes to across the board defection in the final round, but this is far from inevitable.

The main lesson from the game probably involves a deeper appreciation of how individual incentives may undermine group interests. The game also offers the following specific lessons:

Associated Readings

The lessons from the collective action game can be enhanced with appropriately chosen readings. While I have tried a number of different ones, two seem to work especially well: selections from Robert Axelrod's The Evolution of Cooperation (1984), especially the first chapter, and selections from Robert Frank's Luxury Fever (1999). Both books use non-technical language and are uncommonly clear in explaining how collective action problems work. The Axelrod reading has the further advantages of explaining well the connection between the collective action problem and the prisoner's dilemma representation, introducing the distinction between the single and iterated game, and being widely cited in the literature. The Frank reading has the further advantage of using interesting examples from the public policy world. It should be noted that I have also used portions of Mancur Olson's influential book. However, students tend to find that book tough going. I have found it works as well to lecture about the importance of Olson's arguments while using the other readings as grounding for the game and sources of discussion.

But in the end it is the game that makes a potentially difficult and abstract topic come alive. I urge other instructors to give it a try.

Author Bio

Edward L. Lascher, Jr. is professor and chair, department of public policy and administration, California State University, Sacramento. He is the author of The Politics of Automobile Insurance Reform and his work on direct democracy, legislative decision making, political institutions, elections, and other topics has appeared in various academic journals.