A common property experiment with a renewable resource.
I. INTRODUCTIONThis series of experiments demonstrates over-harvesting of a so-called common-property renewable resource. The experiments explore two different solutions to the tragedy of the commons: group communication and the establishment of individual property rights. They can be used in economics courses at all levels, and in courses covering related topics such as environmental studies. The experiments take a total of approximately 50 minutes, depending on class size. Professors have run them in classes as small as seven and large as 75, with all students participating.
II. DESCRIPTION OF THE EXPERIMENTS
In Experiment I, students take the roles of joint owners of a renewable resource. A plate of M&Ms serves as the resource. At the beginning of the first period the plate holds a number of M&Ms equal to ten times the number of students in the class. This amount of M&Ms constitutes the plate's carrying capacity, so the plate can at no point hold more M&Ms. In each period, each student privately writes down the number of M&Ms he or she desires to harvest. It costs nothing to harvest up to 20 M&Ms, but no one may try to harvest more than 20 in one period. In this first experiment, students may not communicate or try to influence each other's desired harvest choice.
Students reveal their desired harvests simultaneously and publicly. If the total desired harvest is less than or equal to the number of M&Ms on the plate, then each person takes his desired harvest. If the total desired harvest exceeds the number of M&Ms on the plate, then each person gets a prorated share of the total harvest, i.e. each gets his desired harvest times the number of M&Ms on the plate, divided by the total desired harvest. The M&Ms left on the plate after the harvest will reproduce, provided there exist enough M&Ms to form a viable population. At least eight M&Ms must remain after harvest to provide enough genetic diversity for reproduction. If there do remain at least eight M&Ms, then each M&M has one offspring, so that the total M&Ms on the plate will double if the carrying capacity permits. After reproduction, another period begins in which students again write down their desired harvest, and so forth. If less than eight M&Ms remain after harvest, then the population crashes, and the experiment ends. Otherwise, the experiment continues for a certain number of periods which the instructor has determined in advance, but which she does not tell the students.
Given the way M&Ms reproduce, if the class as a whole harvested half of the initial M&Ms (five per person), then the plate would return to its carrying capacity in the next period, and students could again harvest five per person. A harvest totaling five M&Ms per person (not necessarily divided equally among class members) represents the maximum sustainable yield for this population. However, a student can try to harvest as many as 20 M&Ms and will likely do so if he believes the others intend to try to harvest 20, thereby crashing the population and ending the experiment. Since students cannot communicate in this first experiment, they have no way of reassuring each other that they will each restrict themselves to a low harvest that allows the population to replenish itself. To try to get around the tragedy of the commons result, Experiment II introduces communication. This second experiment differs from the first only in that students may communicate and attempt to influence each other's harvest choice.
Of course, communication may not solve the tragedy of the commons. Students might still attempt large harvests in order to free-ride off the conservation efforts of others. Moreover, since the agreements students reach don't bind (students still write down their desired harvests privately) a student might not believe others' promises to restrain their harvests. A student who had no faith in others' promises to restrict their harvests would not necessarily feel bound by his promise to restrict his own.
In the third experiment, each student has individual property rights to the M&Ms on a portion of the plate designated as his. Each portion of the plate has its own carrying capacity, equal to ten M&Ms. As long as there remain after harvest at least eight M&Ms in total on the plate, then the M&Ms in each portion will double, subject to the carrying capacity constraint. The experiment continues for the number of periods predetermined by the instructor (but not announced to students). If there do not remain at least eight M&Ms somewhere on the plate, then the population crashes and the experiment ends.
The individual property rights structure of Experiment III gives a student the incentive to conserve his M&Ms. When a student harvests beyond his maximum sustainable yield of five, he bears all of the cost, except in the rare case where the individual's harvest reduces the population below its viable level for reproduction. In contrast, other class members share the costs of every individual's harvest in the first two experiments, with over-harvesting the likely result of this negative externality.
How long each experiment should continue depends on how many M&Ms the instructor has (each one-pound bag contains about 516 M&Ms), and how much time she wants to spend. I suggest predetermining the limit at four to six periods each, in order to get through all three experiments in a 50 minute class. In practice, of course, the population will likely crash earlier in the first experiment and possibly the second also. Not telling students when the experiments will end gives each experiment an indefinite horizon. Unfortunately, time doesn't permit the more realistic infinite horizon. However, the indefinite horizon does limit some of the end-of-experiment effects in which students crash the population in the final period order to extract all possible M&Ms. In Experiments I and II, if people knew which would be the final period, then the fear that the population will crash in that period might cause cooperation (in the form of restrained harvests) to break down in the next-to-last period. In fact, cooperation might unravel back to the first period. Therefore, I suggest leaving the question of which will be the last period up in the air.
I ran this experiment in two of my principles of economics courses (with 16 and 19 students) and in my game theory course (with seven students) at Whitman College. I also ran it with a group of 12 sixth through eighth graders as part of a summer science camp. The principles classes and the middle school kids worked with a carrying capacity often M&Ms per student, as described above, but the game theory class had a carrying capacity of 20 M&Ms per student. In order to simplify the experiment, the middle school kids did not have a lower limit on reproductive viability.
All four groups crashed the resource in Experiment I, although the game theory class did it in the first period, whereas the principles students and the middle school kids took four or five periods. In the second experiment with communication permitted, the game theory group coordinated to harvest the maximum sustainable yield in each of the first three periods. They suspected period four would be the final period, so they jointly agreed to crash the resource that period. They achieved perfect compliance with their group decisions. In contrast, one principles group never managed to agree on a harvesting rule in the second experiment. However, communication in this principles class did lead to more restrained harvests compared to their Experiment I results, so that while the population shrank each period, it was still viable in period six when I ended the experiment. The other principles group had imperfect compliance early on with their rule of taking five each, yet had managed to build the population back up to its carrying capacity when I ended the experiment after four periods. When the middle school kids communicated, they agreed to harvest five each. However, one person defected in the first period, harvesting 19. The kids then revised their group decision to have everyone ask for three or fewer (except the defector, who was told to ask for zero), which they all then did, thereby returning the plate to its carrying capacity. They each took five for the rest of the experiment, which I ended after period six. Under the individual property fights regime of Experiment III, everyone in all four groups harvested the maximum sustainable yield each period, except for a few students who harvested all their M&Ms in the belief that the experiment would end that period.
Because of the amount of time it takes to calculate and distribute harvests in classes larger than about 16 students, the instructor may wish to divide students into groups. Some professors have divided their classes into teams of four to six students, with each team making a harvest decision. In the two cases like this that I know of, both classes crashed the population in the first period of the first experiment, then in the third and sixth periods of the second experiment, and then harvested the maximum sustainable yield in the third. Another option for a large class is to run multiple sections of each experiment concurrently. In one such case, one group of eight students in a class of 32 managed to not crash the population in Experiment I, an example which appeared to have a restraining effect on the other three groups in Experiment II.
Since students might tire of eating M&Ms on the spot, or not want to eat any at all, I suggest bringing plastic baggies for participants to take their extra M&Ms with them. Being able to preserve their M&Ms should keep students motivated to continue collecting M&Ms, preventing the random behavior that Davis and Holt [1993, 87-89] describe from experimental subjects who do not feel motivated by the reward medium. I also suggest keeping the supply of M&Ms hidden so that students can not use the size of the supply to estimate how long each experiment will last.
III. DISCUSSION
Tietenberg's [1996, 271-98] chapter "Renewable Common-Property Resources: Fisheries and Other Species" provides a good reference for the discussion following the experiments. Discussion or a follow-up writing assignment can cover the reasons why a group may overharvest a jointly owned resource, and the effectiveness of communication and individual property rights as solutions to this problem. You can also encourage students to think of other potential solutions. My students volunteered legal limits, peer pressure coupled with publicly recording their desired harvests, taxes and subsidies as possible solutions. Students will likely point out that groups of different sizes and composition may require different solutions. At this point, you may want to discuss the difference between a communal property resource (one with access limited to a particular group) and an open-access resource. We spent some time in my classes talking about the viability of their proposed solutions in the case of international joint ownership of marine life, a case of an open-access resource. To make clear the difference between the communal-property setup of the first two experiments and an open-access set-up, you could ask students to predict the results if the first two experiments had not been limited to the people in the class, but if instead anyone could join.
You might ask students to predict the experimental outcomes if participants could declare their desired harvests anonymously. You could also ask whether the knowledge that each experiment would end soon affected their actions. Along those lines, you could introduce the idea of intergenerational effects by having students predict their behavior in each experiment if they knew that another class would come in after theirs to pick up the experiment where they left off. In fact, an anonymous referee ran a version of this experiment as a guest lecturer in a large general education college course in which the day's topic was intergenerational justice and economics. For each of the experiments, a subset of the class formed a generation whose goal was to pass on to succeeding generations a viable population. The results were striking, in that the populations crashed immediately in Experiments I and II, whereas the students in Experiment III passed on a plate holding its carrying capacity.
I would like to thank Noelwah Netusil, Jan Crouter, Dirk Yandell, Amanda Kernen, Tesa Stegner, Greg Delemeester, Norm Cameron, Betsy Troutt, and an anonymous referee.
APPENDIX I
Discussion Questions
1. What problems came from having the resource owned jointly in Experiment I? Why do you think people behaved as they did in Experiment I?
2. How well did the communication permitted in Experiment II and the individual property rights established in Experiment III each work at overcoming the problems of a jointly owned resource?
3. What other solutions can you suggest for overcoming the problems of a jointly owned resource?
4. If we had a group of a different size, either larger or smaller, how do you think this would have affected our results? How would various group sizes affect the outcomes of the solutions you suggested in Question 3?
5. Suppose participation in Experiments I and II was not limited to people in the class, but that anyone could join. What do you think the results would have been in the first two experiments?
6. If people had been able to keep their desired harvest secret from the rest of the group, so that their actions were anonymous, how do you think that would have affected our results?
7. Did you feel that your actions were influenced by the fact that each experiment was likely to end soon? Do you believe you would have acted differently if each experiment had continued on forever?
8. Suppose there was another class coming in after yours, which would pick up each experiment where your class left off. How do you think our results would have been affected?
APPENDIX II
Common Property Experiments Instructions
We are going to do three experiments in resource economics today.
Experiment I
In the first experiment, the class jointly owns a resource. The resource is the M&Ms on this plate. The plate starts with _____ M&Ms on it. This number of M&Ms is the carrying capacity of the plate. The plate can never hold any more M&Ms than its carrying capacity. In each period of this experiment, each member of the class will be asked to record the number of M&Ms he or she desires to harvest. No one may try to harvest more than 20 M&Ms in one period. No class member may try to influence the desired harvest choice of another class member. The class will reveal their desired harvest numbers simultaneously. If the total desired harvest is less than or equal to the number of M&Ms on the plate, then each person takes their desired harvest. If the total desired harvest exceeds the number of M&Ms on the plate, then each person gets a prorated share of the total harvest, i.e. each person gets: (his desired harvest) x (number of M&Ms on the plate / total desired harvest). The M&Ms left on the plate after the harvest will reproduce, provided there remains enough genetic diversity for viable reproduction, which means at least eight M&Ms left. If there are at least eight left, then each remaining M&M will have one offspring, so that the total M&Ms on the plate will double if the carrying capacity permits. (Keep in mind that the carrying capacity is _____ M&Ms, and that is the most that the plate will hold.) After reproduction occurs, a new period begins in which each class member again writes down his or her desired harvest, and so forth. If there are less than eight M&Ms left after harvest, then there is not a viable population for reproduction. In that case, the population has crashed, and the experiment ends. The experiment will continue a specific number of periods, unless the population crashes and terminates the experiment. I will not tell you how many periods the experiment will continue.
Experiment II
The second experiment is identical to the first, except that class members may talk to each other and try to influence each other's desired harvest choice.
Experiment III
In the third experiment, the plate is divided into equal portions, one for each class member. Each member owns the ten M&Ms on his portion of the plate. The carrying capacity of each portion is ten, so you may never have more than ten M&Ms on your portion of the plate. At the beginning of each period, each person records the number of M&Ms in their portion which they chose to harvest that period. Class members then simultaneously reveal their harvest choices and harvest their M&Ms. If there are at least a total of eight M&Ms left somewhere on the plate after the harvest, then each M&M left in your portion will reproduce itself, up to the carrying capacity often. If there are not eight M&Ms in total left on the plate, then the population crashes, and the experiment ends. The experiment will continue a specific number of periods if the population does not crash. I will not tell you how many periods the experiment will continue.
REFERENCES
Davis, Douglas, and Charles Holt. Experimental Economics. Princeton: Princeton University Press, 1993.
Tietenberg, Tom. Environmental and Natural Resource Economics. New York: Harper Collins, 1996.
Hazlett: Assistant Professor, Whitman College, Walla Walla, Wash., Phone 1-509-527-5155
Fax 1-509-527-5026, E-mail hazlett@whitman.edu
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Author: | Hazlett, Denise |
|---|---|
| Publication: | Economic Inquiry |
| Date: | Oct 1, 1997 |
| Words: | 2970 |
| Previous Article: | An empirical investigation of asymmetric behavior in corporate dividend policy. |
| Next Article: | Profit maximization problem. |
| Topics: | |