Published in ENVIRONMENTAL PSYCHOLOGY AND NONVERBAL BEHAVIOR, 1979, Volume 3, pp.252-254. Also described by the author in PSYCHOLOGY TODAY, August 1979, p.80

    Concern over depletion of the world’s vital resources has provoked a limited number of experimental investigations by social and environmental psychologists to determine what psychological factors are involved in the mismanagement of resource pools (see Edney & Harper, 1978) for a review). Although laboratory work on this topic is recent, indications are strong that the processes of the tragedy of the commons (Hardin, 1968) can be captured in experimental form using small groups and contrived tasks. Computerized analogs (e.g., Brechner, 1978; Cass & Edney, 1978) have been developed as well as non-computer methods (e.g., Edney & Harper, 1978a, 1978b) but the latter require nonportable apparatus. Other, simpler apparatus have also been reported (e.g., Watzke et al., 1972), some of which are time-consuming to run. This note describes an apparatus for a game which embodies the dynamics of the commons problem. It is simple, portable, easy to operate, and allows manipulation of a number of independent variables as well as the collection of a number of dependent variables. Experimental trials are also short.

    Essentially a commons problem occurs when a community of consumers consumes resources at a rate high enough to endanger the regenerating resource itself. Most resources that are living, and many that are not, are liable to overexploitation, and the consumers face the choice of restricting their own consumption for the good of the pool and the community, or continuing to consume at self-satisfying rates to face dire consequences at a later time.

    This problem can be operationalized in the following form, paralleling the commons dilemma in many ways. A small number of subjects (three or more) sit around a shallow, nonbreakable, open bowl (convenient diameter 12 inches) which initially contains 10 hardware hexagonal nuts (half-inch diameter is convenient). An experimenter sits with the group and introduces the exercise as one where the player’s goal is to get as many of the nuts as possible. Players can take nuts from the bowl at any time and in any quantities after the start of the trial. The experimenter also explains that the number of nuts remaining in the bowl after each 10-second interval is automatically doubled from an outside source (operated by the experimenter, who manually replenishes the nuts
from a separate container next to him). This replenishment cycle continues until either an arbitrary time limit is reached or the bowl is emptied by the players. The experimenter can also set a ceiling to the number of nuts in the bowl throughout the trial (10 is convenient). Subjects can be asked not to communicate.

    To maximize their individual "harvests" of nuts, one would expect that each subject would restrain himself to taking one or two nuts out of the bowl each 10-second period: this allows the replenishment cycles to continue for some time (a typical game runs 2 minutes) and each subject eventually would end up with a sizeable score. In pilot work I have found that approximately 65% of groups never in practice reach the first replenishment stage because they exhaust the pool by taking all the nuts out in the first few moments of the game.

    In this apparatus the bowl symbolizes a resource pool (such as an ocean of whales); the nuts, the resources themselves (these are given real value to the subjects by making them redeemable for money or credit); and the replenishment cycles, natural resource regeneration rates. Taking the nuts is obviously harvesting behavior. As in real-world situations, subjects frequently make the mistake of exhausting the pool by overharvesting, resulting in unnecessarily low scores on each trial. Whatever factors enable subjects to develop a slower harvesting rate (e.g., communication) benefits them all in the long run, and these factors are the focus of experiments as analog methods of determining what prevents resource crises and they form independent variables. Parametric variations of pool size, replenishment rates, value of the nuts, etc., should be easy to explore.

    Dependent measures can include (a) individual and group scores (harvest size), (b) number of replenishments or duration of the trial, (c) variance in individuals’ harvest sizes, (d) questionnaire items asking subjects for their reactions to the experience, etc., all subject to conventional statistical analysis over a number of trials and/or groups.

    Recent pilot work has also shown that the game is unusually involving for subjects. As in the Prisoners Dilemma game (mathematically related to Commons Dilemma games) a fundamental aspect of a good outcome is cooperation among players. Cooperation requires some form of trust in the exercise of mutual restraint, and the evolution and breaking of trusts over the task appears to be an important part of the social dynamics of the exercise.

    In pilot testing I have explored a method of running the game in which groups first play the basic game (a majority end up with low scores), then each group is allowed a 5-minute period of free discussion to invent their own rules in order to increase their harvests on a second game, played after the discussion period. The focus here has been on the kinds of rules (methods of play) groups spontaneously generate: these are essentially group-generated attempts at solutions to the commons dilemma. So far, results have suggested two main types of solution: (a) those involving numbers (such as the group which decides systematically to take only "one" or "two" nuts per person per 10-second interval; this type of solution is quite effective in preserving the pool) and (b) nonnumerical solutions. An interesting illustration of the latter was a group which decided deliberately to use a rather complicated system of harvesting. Each player had to hook each nut out of the bowl with a pencil, place it on his nose, walk over to a nearby chalkboard, and deposit the nut in the tray before returning for another nut. Harvesting was thus slowed down enough to prevent pool depletion, increasing individual scores, and incidentally making the game more entertaining to players. The evolution of both kinds of solutions can be regarded as analogs to community-generated laws and practices for direct and indirect governance and management of resources in real-world situations.

    In sum the Nuts Game is a practical, manageable, concise, and conceptually defensible form of analog for the study of behavior in resource shortages, and it should lend itself to research in a variety of settings. It might also be conceived of as a teaching device for teaching individuals (such as school children) the value of cooperation in resource management situations. The game’s inherent flexibility allows wide variation of parameters and independent variables and makes practical a variety of dependent measures to suit the researcher.

Notes