Gaming is the technique of placing subjects in an environment that requires them to make joint or collective decisions among hypothetical options. The subjects are shown the prospective consequences of their decisions as the game proceeds. Playing the game can improve one’s understanding of how a water system responds to different water management decisions and may reveal changes in operating rules, which improve the performance of the system.
The Colorado River Gaming study was part of a larger effort called the Study of Severe Sustained Drought in the Southwestern United States (the SSD Study), which was conducted by a consortium of western U.S. universities. The purpose of the SSD research program was to identify feasible changes in operating rules for allocating and managing Colorado River water. The SSD Study was supported by funds from the USGS, the Metropolitan Water District of Southern California, the Upper Colorado River Commission, the National Drought Study and the participating universities. The SSD Study grew out of the U.S. State Department’s “Man and Biosphere Program.”
The Colorado Gaming Study was a collaborative effort between the SSD and the National Drought Study teams. Bill Lord (University of Arizona) was the principal architect and proponent for the game. Like the Drought Study, the Gaming Study was designed to connect research and practice, and both kinds of water experts participated. State water officials from the Colorado basin served as study advisors, while students and professors played the roles of officials and applied the research and findings from the SSD Study to the game.
The objectives of this gaming exercise were to screen alternative operating rule sets before a more detailed evaluation and to compare alternative collective choice rule sets. (See page 8 in the main report for definitions of the three levels of rules.) An analysis using game theory helped define the three collective choice rule sets that distinguished the 3 games. (Gaming is distinct from, but can be used in conjunction with the mathematical theory of games. John von Neumann and Oskar Morganstern developed game theory as a method for analyzing competitive situations in economics, warfare, and other areas of conflicting interest [Glicksman.])
The gaming exercise was first done as a graduate seminar in the spring term of 1992 at the University of Arizona. Seven graduate students assumed the roles of state water officials from the seven states of the Colorado River Basin, and chose how they would respond to a severe drought as it unfolded through the simulation of AZCOL. AZCOL was a model of the hydrology, water management facilities, water allocation institutions, and water demands of the basin. AZCOL was developed using STELLA II®. The choices the students could make were limited; some were intrastate, others were interstate and required two or more students to come to an agreement to act. The game was played nine times, each time with unique set of rules.
A second game was played in June 1993 with university professors from each state playing the role of water directors, and a study advisor playing the role of the Secretary of the Interior. The game was based on a very severe 38 year drought (on the order of a 500 year event), and it was played three times under three collective choice alternatives:
The Status Quo Game, in which operating decisions for water management facilities were made by the Department of the Interior in accordance with existing rules. Any changes in the rules required unanimous agreement by DOI and the seven states. This alternative was marked by limited information about consequences, especially to other states. The players used e-mail to conduct this game so as to prevent face-to-face bilateral negotiations.
Colorado River Basin Commission Game. A river basin commission would provide many of the advantages of an interstate compact, but could be established more easily, so this option was simulated in the second game. The rules for the Commission Game were identical to the Status Quo game except that the “Commission” shared more information and analyses with the states, and the players assembled in one location for this game. This meant they could take part in group discussions (no bilateral discussions were allowed) on decision making and alternative decision rules. Given this framework, players developed an alternative to the “equalization rule,” the status quo method of balancing hydropower production from Glen Canyon Dam and water deliveries to the Lower Colorado Basin.
Water Banking and Marketing Game. This differed from the Commission Game in that any two states could exchange unlimited amounts of information and could make bilateral agreements (including interstate sale of water) so long as the other 5 states were not harmed. The “Secretary of the Interior” made the judgment of no harm. In each game, participants had freedom to make changes in operating rules within the applicable collective choice rule set. Players identified the important measures of performance (primarily, amounts of water delivered) and decision criteria. Before the games were played, decision criteria were evaluated for relative importance using the Bureau of Reclamation’s MATS software. In general, the pre-game subjects rated economic impacts twice as important as equity and endangered species preservation more important than maintenance of wetlands and riparian areas. These inferences were important for the game but have little significance out of that context, since these values were elicited from professors playing the role of state water directors.
This study showed that:
In this simulation, water managers were most interested in satisfying diversions for consumptive use and avoiding impacts that would trigger action under the Endangered Species Act. Hydropower production, recreation, salinity, and most non-ESA environmental impacts were less important.
There is sharp competition among water uses on the Colorado, and the Status Quo does not provide clear decision criteria for allocation. The water use priorities of the “Law of the River” are further complicated by independent rules implicit in the Endangered Species Act, the Clean Water Act and federal reserved rights. And, in these three games, the economic value of hydropower was not matched by an equivalent priority for waters. (See page L-6 in the main report for a real world verification of this discrepancy between economic impact and water management priority.)
Consumptive water uses are well protected from drought, but non-consumptive uses are not. Drought risk is the highest in the upper basin.
Only minor changes can be made under existing rules, and only minor improvements can be made with the other two collective choice rule sets, so long as changes in federal legislation or court decisions are ruled out.
Intrastate drought management is more effective in limiting drought losses than are easily adopted changes in interstate water allocation.
The games suggested the value of a compact commission, perhaps similar to the Delaware River Basin Commission, that would examine a re-balancing of consumptive and non-consumptive uses facilitated by water banking and marketing.
Finally, the gaming exercise demonstrated a complete method for testing and reporting on institutional changes.