1 Introduction

Climate resilience in the face of drought is the ability of policies, practices, infrastructure, communities, and users to withstand and adapt to the impacts of extreme weather events, such as drought and flooding, as well as more subtle changes such as rainfall variability and inconsistent snowpack [1]. Climate resilient management of water resources can be supported by using tools and technologies to improve understanding of community water conservation practices.

In recent years, the Western United States has experienced a significant decline in groundwater levels. This depletion is primarily attributed to human-induced extraction and natural factors affecting recharge rates. For communities dependent on irrigated agriculture, this trend poses substantial challenges to local economies and traditional ways of life. However, these challenges also present opportunities to develop community resilience and deepen our understanding of effective conservation practices. Policy interventions have long been employed as tools to encourage resource conservation within communities or to curb excessive resource use. A notable example of such an intervention can be found in Southern Colorado, a region that exemplifies the challenges faced by rural, agricultural communities grappling with declining groundwater resources. Here, the Rio Grande Water Conservation District spearheaded the development and implementation of a groundbreaking initiative: a voluntary fee tied to groundwater pumping allocations. This first-of-its-kind program in the United States aims to restore the aquifer system to sustainable levels [2, 3].

In response to alarmingly low groundwater levels and looming state intervention, farmers in the San Luis Valley (SLV) came together to implement a self-imposed fee on groundwater use, which has incentivized a one-third reduction in groundwater use since 2013 [4]. The fee is used to help irrigators buy supplemental surface water or to pay them to let their acreage fallow in dry years and encourages irrigators to plant less thirsty crops and use water more efficiently. The long-term implications of groundwater depletion for agriculture and the environment are complex and multifaceted. Some potential consequences include reduced crop yields, land subsidence, and increased energy costs for pumping water from deeper wells. Groundwater depletion can also have negative impacts on ecosystems and wildlife that rely on groundwater. The local governance measure in the SLV has served as an example as lawmakers in California, Texas, and other states consider similar ways to regulate groundwater and curb over-pumping [4]. While using monetary tools such as increasing local taxation and imposing fees on groundwater withdrawals can serve as an effective tool for conserving groundwater in some agricultural communities, it may not be effective everywhere and across all groups of users. Finding ways to mitigate groundwater usage requires a combination of policy changes and technological innovations, both of which aim to encourage behavioral shifts among farmers and other community stakeholders.

Hardin’s influential “tragedy of the commons” theory [5] posited that individual actions motivated by self-interest would inevitably lead to the overexploitation and eventual depletion of shared resources, or common pool resources (CPRs). Hardin argued that this outcome could only be avoided through strong government regulation or privatization. However, Ostrom and others have challenged this notion through extensive field studies, demonstrating that small, local communities can effectively self-manage and self-regulate CPRs without external interference [6, 7]. CPRs, such as forests, fisheries, or irrigation systems, are often subject to collective action problems. In these situations, individual resource users may have an incentive to overuse the resource, even if it is not in their best collective interest to do so [8]. This occurs because the benefits of overuse accrue to the individual user, while the costs are shared by the group as a whole. Consequently, CPRs are frequently at risk of overexploitation, potentially leading to resource depletion or degradation. Ostrom’s work revealed that under certain conditions, such as community trust and effective communication, local communities can devise rules and effective mechanisms to avoid resource over-exploitation. This research challenged the conventional belief that top-down intervention was the only solution to the “tragedy of the commons.” Instead, Ostrom’s findings highlighted the importance of strong local institutions in the successful governance of CPRs. By demonstrating the potential for community-based resource management, Ostrom’s work has significantly contributed to the understanding of sustainable resource governance. It has shown that local knowledge, social norms, and community-driven enforcement mechanisms can play crucial roles in maintaining the long-term viability of shared resources.

Robust local institutions can help address these collective action problems by creating rules and norms that encourage resource users to cooperate and to manage the resource sustainably. By doing so, they can help to ensure that the benefits of the resource are shared fairly among all users and that the resource is managed in a way that is sustainable over the long term. Voluntary local leaders play an important role in the initiation of self-governance institutions because such leaders can directly affect local users’ perceived costs and benefits associated with self-rule. Andersson et al. propose that unselfish behavior and leading by example are key to facilitating a cooperative process of local rule creation (2020). The emergence of self-governance institutions may be influenced by a variety of factors, including the characteristics of the resource, the characteristics of the user group, and the broader political and economic context in which the resource is situated. For example, the emergence of self-governance institutions may be more likely in situations where the resource is relatively small and homogeneous, where the user group is relatively small and cohesive, and where there is a high degree of trust and social capital among resource users [9, 10].

There is little scientific evidence on which combination of policy and technology interventions are most likely to be successful in promoting the joint goals of economic performance and environmental protection [11]. There has been scholarly exploration of some groundwater regulations: well spacing requirements were introduced in Kansas [12], some California basins created (non-tradeable) groundwater rights [13], portions of Nebraska created cap-and-trade systems [14], and SLV implemented a groundwater pumping fee combined with subsidies for crop fallowing and a Payment for Ecosystem Services (PES) program [4]. There are potentially dozens of other monetary interventions that could be effective in curbing over-extraction in places that face groundwater shortages; however, the challenge is that it is hard to predict which particular intervention is likely to be most effective in local contexts and what role non-monetary incentives like information sharing, voluntary leadership, and public recognition have.

This knowledge gap underscores the need for more scientific research on the emergence and effectiveness of various management rules, particularly economic-based incentives, which have been infrequently implemented in the Western US. While researchers have conducted lab and field experiments to understand groundwater user behavior [15–17] and the interaction between regulation sources and their effects [18, 19], recent studies emphasize that successful groundwater governance often stems from local collective action [20]. This participatory approach, where stakeholders collaboratively design rules tailored to their specific conditions, can lead to more equitable and effective governance outcomes that balance economic and environmental objectives. Thus, future research should focus on understanding how these locally-driven, collaborative approaches can be effectively implemented and scaled to address groundwater management challenges across diverse contexts.

In spite of efforts to study behavioral approaches to water use [21–23], the uptake of technological approaches and tools for supporting collective action is understudied. Behavioral research tools are needed to better understand how communities in the Western US are responding to water stress and the combined role that data collection, communication, and monetary and non-monetary incentives can play in motivating community water management. This principle of integrating technological solutions with human behavior is particularly relevant in systems where technologies require complementary human actions to achieve desired outcomes. Such integration is common in environmental health and is equally applicable to sustainable water management, where the goal is to benefit communities and agricultural producers both now and in the future [24]. Technological solutions alone are insufficient; their effectiveness is heavily dependent on human adoption, understanding, and consistent implementation.

Common pool resource games have been used to study collective action under differing governance strategies. While some prior studies have shown correlations between real-world behavior and behavior within games designed to simulate those situations [25–27], ultimately, these results hinge on the degree to which real-world teamwork requires the same critical components necessary to achieve success in our game: effective communication and coordination with others who have different access to knowledge and resources [25]. Individual actions in groundwater management can lead to increased dependence on groundwater and incentivize continued and expanded groundwater use [28]. This can create tension between individual and collective adaptation efforts. However, incremental changes can lead to larger, collective adaptations over time [28]. Behavioral games can be engaging and versatile tools with the potential to be used to inform local jurisdictions on groundwater regulation, as has been demonstrated by other behavioral games that have provided a framework for public policy measures [29]. However, while behavioral games are often used in studies of natural resource management in international research [30, 31], they are not frequently used in similar methods in studies of the Western US.

Traditional social science methods, such as choice experiments, large N-surveys, and key informant interviews, have been widely used to study farmers’ water extraction behaviors [32–34]. While valuable, these methods are susceptible to biases like acquiescence, moderacy, and satisficing, which can distort data accuracy [35]. Holland’s 2022 study, using interviews and surveys to identify challenges facing Colorado water managers, provided insights but did not fully capture the complex interactions between stakeholder groups’ management practices [36]. Data-informed behavioral games offer an alternative approach that can potentially overcome these limitations. By simulating local agricultural community dynamics, these games can better illuminate farmers’ and producers’ motivations and decision-making processes without the risk of response biases. This approach allows for observation of stakeholder behaviors in response to simulated policy changes or environmental pressures, revealing unexpected outcomes that might not be apparent through conventional research tools. This is particularly crucial when considering novel policy instruments to address groundwater depletion, such as tradable permits or subsidies for regenerative farming. By enabling participants to interact within a simulated ecosystem, these games can provide insights into the complex, interdependent nature of stakeholder decisions and their cumulative effects on water resources, potentially informing more effective and tailored policy responses to groundwater management challenges.

Through collaboration with local water users, policymakers, and other relevant stakeholders in Colorado, this work examines the design and evaluation of a data-informed digital behavioral economics game that could improve the study of incentives and mechanisms that lead to groundwater conservation within agricultural communities compared to traditional methods. Also included is an exploration of the utility of this game as a tool to evaluate the potential effectiveness of local governance measures related to water policy.

2 Methods

A behavioral economics game was designed and evaluated as a digital study tool to advance knowledge of behaviors and practices associated with groundwater conservation and governance in contrast to existing qualitative and quantitative social science methods. This stakeholder and data-informed behavioral game was designed by integrating groundwater pumping data from representative geographies of an agricultural community reliant on groundwater in Southern Colorado. The potential effectiveness of this tool as a means to simulate local governance measures related to water usage and improve understanding of farmers’ behaviors around groundwater was evaluated by game sessions played with key stakeholders followed by focus group discussions.

2.1 In-situ data sources

The data-informed behavioral game was designed to simulate crop choice and groundwater withdrawals for agricultural producers based on conditions similar to those in Southern Colorado. The SLV receives just 6–9 inches of rain annually. To grow crops, farmers irrigate 95% of the cropland. There are complex interactions between groundwater and surface water and the two layers of the aquifer itself. In general, groundwater extraction, especially to the north of the Rio Grande, affects both the river and the aquifer. Impacts to one well due to pumping at a nearby well have been noted but are hard to disentangle from the overall drop of groundwater levels where wells are dense [37].

In situ groundwater pumping data was collected from June 2022 through November 2023 over two consecutive growing seasons from the partnering communities to better inform and simulate groundwater pumping behaviors and incentives in the game. Five groundwater pumping sensors were installed in the SLV to monitor groundwater extraction from the Rio Grande aquifer. Measured trends in groundwater pumping from these sensors were used to generate plausible groundwater pumping forecast simulations for the digital behavioral game. These groundwater monitoring sites were installed on three representative stakeholders’ alfalfa farms in Alamosa, Colorado, seen in Fig 1. This municipality is part of the region’s largest irrigation district, Subdistrict 1, under local jurisdiction from the Rio Grande Water Conservation District where self-imposed groundwater pumping taxes have increased from $45 per acre-foot in 2013 to $500 per acre-foot in 2024 [38]. These tax disincentives and monitored pumping behaviors informed the design of the simulated groundwater irrigation game, wherein the game design attempted to provide realistic pumping and tax scenarios that stakeholders would react to.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

A on the left is topographic map of Colorado highlighting Subdistrict 1 irrigation district in the SLV surrounding the town of Alamosa. B, the inset on the right, shows Subdistrict 1 in detail, marking the monitoring locations of groundwater pumping on three alfalfa where data was recorded from June 2022 through November 2023.

https://doi.org/10.1371/journal.pwat.0000298.g001

Along with data collected on groundwater pumping, stream height data along the Rio Grande River was retrieved from the Colorado Department of Water Resources gauge at Thirty Mile Bridge in Creede, Colorado; surface water from the Rio Grande River is used as a primary irrigation source for approximately 80% of farmers in the SLV supplemented with groundwater use [38]. Stream height and pumping usage data is overlaid in Fig 2 displaying the average groundwater pumping usage divided by maximum level each week and average stream level divided by maximum level each week. The pumping frequency validates key informant insights from SLV stakeholders that as the river level lowers throughout the irrigation season (approximately April to October), groundwater pumping increases and then remains steady.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Displayed is average groundwater pumping usage divided by maximum level each week and average stream level divided by maximum level each week.

The pumping data was from three alfalfa farms in Alamosa, Colorado, that all utilize both surface and groundwater rights. Stream height data from the Rio Grande River was retrieved from the Colorado Department of Water Resources gauge at Thirty Mile Bridge in Creede, Colorado. There is no data for groundwater pumping outside of the growing season period from April through October.

https://doi.org/10.1371/journal.pwat.0000298.g002

2.2 Game scenario choices

Agricultural practices including crop choice, expected water use, and common regenerative agricultural techniques developed in the game design were also informed from regional practices in the SLV through discussions with local, representative stakeholders as well as subsidy data from the USDA [39]. Table 1 describes five production packages—choosing between them represent the primary decision players make in the game. These production packages were reviewed and validated as accurate representations of local practices from agricultural producers in the SLV. Along with the five options listed here, another version of the game included five identical packages that did not include subsidized income for regenerative agricultural practices. Crop type is the central component of each package (and consequently the key decision to be made by players) it plays a critical role in net water extraction, has a direct impact on potential revenue (referenced in the game as “Maximum Income”), and is also readily visible, so local practices relating to crop choice are a viable way to limit groundwater extraction.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Agricultural production packages.

https://doi.org/10.1371/journal.pwat.0000298.t001

Regenerative agricultural practices have long-term advantages in arid farming communities including the potential to increase soil water retention, improve crop yield, sequester carbon in the soil, and decrease erosion. Cover cropping, for example, can increase nitrogen levels in the soil in some cases reducing fertilizer application by 50% [40]. After implementing these practices over time, farmers may create more drought resilient soil that requires reduced water inputs. As cover cropping and no/low tillage may necessitate costs for seeds, equipment, and labor farmers often require incentives to integrate regenerative practices. In Colorado, the Conservation Reserve Enhancement Program (CREP) provides funding for this through the Department of Natural Resources [41]. CREP was used to inform subsidy incomes used in the behavioral game while also forecasting reduced water use for implementation of regenerative practices.

2.3 Digital game design

The system dynamics model behind the behavioral game is shown in detail in Fig 3. Participants in the game all draw from a common groundwater resource that is regenerated consistently at the end of each round (representative of an agricultural season or year) with random drought events where recharge varies. Changes in the groundwater table resulting from the water pumped by the participants, affect the cost to pumping in future rounds: as groundwater levels decrease, it becomes less efficient and more expensive to withdraw water (described by Eqs 3 and 4). These factors contribute to a crop yield estimate (Eq 2) which in turn affects participants’ income in each round. As pumping effectiveness (Eq 1) decreases with the decreasing water table, it is assumed that crops will not receive irrigation levels needed for a full harvest and crop losses can occur. These nonlinear relationships between key variables are derived from real-world scenarios including the collected in-situ pumping data and represent an important improvement over the models used for dynamic groundwater games so far [30] since they better represent the typical CPR dynamics where declining resource base gradually (rather than abruptly) manifests in reduced appropriation effectiveness and/or higher costs. Players’ individual decisions regarding which crops to plant (and the corresponding water use) at the beginning of each round of the game affect the common pool groundwater resource and their income. The game is won by the player with the highest accumulated income at the end of all rounds.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. The system dynamics model behind the behavioral game is displayed here where all players’ decisions at the beginning of each round of the game affect the common pool groundwater resource level (described in “units”) and their own accumulating income (in “dollars”).

Pumping cost, pumping effectiveness, and crop yield all interact with decisions made by participants and the level of the groundwater table; their nonlinear relationships to current groundwater level are presented in Fig 4. Players’ production package decisions affects water pumped from a common groundwater resource. As the water level (which begins at a set maximum of 100 units) decreases after each round it becomes less efficient and more expensive to withdraw water and crop yield may decrease, affecting player’s net income.

https://doi.org/10.1371/journal.pwat.0000298.g003

The interrelated parameters to determine costs of crop yield, pumping effectiveness, and groundwater withdrawals were set to simulate a realistic environment for participants based on the in-situ data collected in the SLV and research from the University of California, Davis, and the USGS which has studied the economic impacts of groundwater pumping in agricultural regions and studies on groundwater depletion and its impacts on crop yields in the US [42, 43]. These parameter relationships are displayed in Fig 4. The pumping cost multiplier (Eq 4 describes the effect of available groundwater. As the available groundwater level decreases from all players’ water use it becomes less efficient and more expensive to withdraw water and crop yield may decrease, affecting player’s net income. (1) (2) (3) (4) (5) Where:

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Pumping cost multiplier function represents increasing costs of pumping water with lowering groundwater level.

Pumping effectiveness function represents increasing difficulties of pumping water with lowering groundwater table. Crop yield represents a fraction of the crop yield as a function of a fraction of required water based on individual participant’s production decisions.

https://doi.org/10.1371/journal.pwat.0000298.g004

Fig 5 presents one of the screens of the graphic user interface that is accessible to participants during game play displaying current current conditions at the beginning of each round or “year”. The “available groundwater” icon will visually display the groundwater table increasing or decreasing based on the net result of groundwater pumping from all participants and groundwater regeneration combined.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Graphic user interface displaying environmental conditions for players to view at the end of each game round or “year”.

The “available groundwater” icon will visually display the groundwater table increasing or decreasing based on the net result of groundwater pumping from all participants and groundwater regeneration combined.

https://doi.org/10.1371/journal.pwat.0000298.g005

Three versions of the behavioral game were designed following Ostrom’s Common Pool Resources appropriation game. The experiment, moderated by a trained facilitator,consists of groups of five irrigators making individual appropriation decisions in multiple experimental stages, as outlined below. By the use of actual groundwater pumping data collected, the game is designed to test the behavioral effects of three different treatments on groundwater conservation:

1. Piguovian Tax: replicates the baseline or current conditions in the rural agricultural communities in the Western US like the SLV, where fees may be imposed on groundwater pumping by local governance measures;
2. Data Transparency: displays players’ crop choice and water consumption to all participants at the end of each round, elucidating the effects of rural community dynamics;
3. Subsidies: includes financial compensation for fallowing land or incorporating regenerative agricultural practices including cover cropping and low/no tillage.

The baseline version of the game was expected to elicit decisions on groundwater pumping that generally reflect typical stakeholder behavior. An intervention with total data transparency would share all individual groundwater pumping data within the community, providing possible personal incentive to curb resource use through community pressure. This intervention may lead to moderate changes in groundwater sustainability among participants. As seen with some utility companies [44], the inclusion of neighbor comparisons as a data sharing mechanism can have an immediate effect on resource use. Finally, the inclusion of financial subsidies for regenerative practices may facilitate the adoption of these regenerative practices, including fallowing, leading to greater groundwater conservation in the game.

Using common pool resource games to collect data can help inform policies especially related to shared resources. When gathering data through CPR games, internal validity is difficult to assess, so the design and validation of our tool with stakeholders will help make the simulation as realistic as possible and presumably more effective [25, 45–49]. To use such data from CPR games effectively for policy learning requires local decision-makers to be actively engaged in joint research activities, or the research results are not likely to be widely used. Therefore, behavioral game sessions were run with a variety of stakeholders representing agriculture and local policy implementers to validate the usefulness of this tool and provide feedback on its potential utility in simulating local policy changes.

Throughout the fall of 2023 across seven sessions, thirty expert stakeholders in Colorado including 15 agricultural producers and ranchers, 11 conservation district managers, 5 academics, 2 water lawyers, 3 water economists and a state official (with some stakeholders fitting into multiple categories) were recruited to play and validate the behavioral game as a tool for improved understanding of groundwater conservation practices within agricultural communities. In the current game design, all participants are assumed to operate the same acreage and therefore every player’s decision is equally impactful; however, this could be modified in future versions to reflect more realistic conditions based on local stakeholder’s acreage in production by weighting water requirements.

3 Results

Overall, game sessions were characterized by strong player engagement, fostering rich discussion between farmers, ranchers, and other relevant community members. When unexpected drought events occurred in the game and the groundwater table dropped to its lowest levels, players would work together to develop effective strategies to address the challenges and stabilize the groundwater table demonstrating the usefulness of this behavioral game as an effective tool to observe community conservation practices.

Similar behavioral game methods to the tool described in this work have been used in a pilot study in Andhra Pradesh, India where collective action games were used to simulate crop choice and consequences for the aquifer [30]. The game sessions allowed community members to discuss different mechanisms to find an appropriate middle ground and triggered discussions about the linkages between agricultural practices and the status of groundwater, and about what steps could be taken in order to slow aquifer decline. Of interest here is the long-term outcome this study had on community behavior: communities that participated in game playing were more likely to develop new rules about ground water extraction than in communities where the game was not played. The game can help stimulate deliberations within groups, a first step toward self-governance.

In all sessions played with farmers who irrigated crops with groundwater, participants shared that the simulated game design accurately represented key aspects of local agricultural practices, validating the game design. Specifically, farmers noted the realism of economic values for crop payments and water pumping costs, crop choices, subsidy values, and regenerative practices. They also recognized the accuracy of groundwater parameter dynamics, including recharge rates, pumping efficiency, and crop yields for the objective of the tool. While the game simplified real-life decision-making processes, particularly regarding detailed irrigation scheduling, it focused on eliciting specific behaviors and decision-making factors related to long-term resource management. Of all stakeholder groups that played the game, farmers were most excited to provide direct feedback into their decision-making processes, often sharing their motivations to continue irrigated agricultural practices despite extreme economic variability. Most farmers participating in game sessions had additional sources of income to supplement revenue from agriculture, with one player notably comparing farming to gambling. The game’s design, which introduced dynamic parameters such as a finite groundwater resource, crop choice and water use, regenerative practices, subsidy influence, and community pressure, engaged farmers in a unique way. These discussions demonstrate the usefulness of games in elucidating insights into behaviors and motivations around groundwater pumping for irrigated agriculture that may have been difficult to determine through more traditional survey methods alone. By focusing on specific behavioral factors rather than detailed irrigation scheduling, the game provided a versatile tool for studying a busy demographic in a more engaging and time-efficient manner than traditional methods or full simulations. This approach allowed for valuable insights into farmers’ decision-making processes regarding long-term water resource management and conservation strategies.

One of the greatest advantages of using behavioral games as a study method is the tendency to generate lively discussions and debates among participants, something that is very rarely observed with traditional methods [30]. This can further illuminate motivations to behavior patterns and community dynamics. Future work should be done to quantify this difference through monitoring dialogue frequency and sentiment in games compared to standard interview methods.

Fig 6 illustrates the impact of introducing subsidies for regenerative agricultural practices on groundwater table levels over the course of the game. The underlying data reveals significant changes in participants’ crop choices and water use patterns when financial incentives were introduced. In the baseline version without subsidies, participants consistently chose to plant alfalfa, steadily depleting the groundwater table. Without subsidies, participants chose water-intensive Alfalfa and Alfalfa with regenerative practices 17 times, while opting for more water-conservative choices (Barley with regenerative practices and fallowing) 13 times. When subsidies were introduced, this pattern reversed: water-intensive choices decreased to 16, while water-conservative choices increased to 24. Notably, fallowing increased from 3 to 10 instances with financial subsidies. This shift towards more water-conservative choices resulted in higher groundwater table levels, as depicted in Fig 6. The data suggests that financial incentives encouraged participants to opt for fallowing or less water-intensive crops more frequently, contributing to improved groundwater conservation demonstrating how economic incentives can effectively influence agricultural decision-making towards more sustainable water use practices. Behavioral games incorporating information sharing mechanisms led to dynamic interactions among participants, with notable instances of peer accountability for water-intensive decisions. However, despite these engaging discussions and apparent social pressures, the final groundwater levels remained comparable to those observed in baseline games without information sharing. This outcome suggests that while increased transparency and peer interaction may influence short-term behavior, they may not be sufficient to significantly alter long-term groundwater management outcomes in the simulated environment.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Results of game played with and without subsidies for regenerative agricultural practices with key stakeholders illustrating changes in the groundwater table over the course of eight rounds of decision-making.

https://doi.org/10.1371/journal.pwat.0000298.g006

Qualitative assessments of gameplay and debriefing sessions following the behavioral economics study provided valuable insights into the relevance of the experimental game to the challenges faced by rural, agricultural communities in Colorado. These discussions explored how the outcomes of the game could shed light on the real groundwater situation within stakeholders’ localities and explored potential strategies to address existing issues. Specifically, participants acknowledged the state pressure in the SLV to restore aquifer levels in the coming years, recognizing the critical importance of conservation efforts. The discussions highlighted the local economy’s heavy reliance on irrigated agriculture, particularly alfalfa production, and the desire of generational farmers to pass their farms to future generations, underscoring the long-term significance of sustainable water management. Participants engaged in substantive conversations about potential conservation strategies, including the controversial option of shutting down wells and the need for enhanced education programs in these localities to improve groundwater conservation practices. While existing subsidy programs were acknowledged, discussions illuminated that these are often ineffective in changing long-standing agricultural practices, suggesting a need for more innovative or targeted approaches. These sessions also included discussions of the versatility of the game for uses beyond the goal of groundwater conservation, such as modeling other resource management scenarios or policy impacts. Finally, participants were further invited to discuss the advantages and disadvantages of the tool as a study mechanism compared to traditional methods such as surveys and choice experiments.

The intervention opportunities identified through these discussions described innovative approaches for using this tool to improve understanding of water resource conservation challenges and behaviors effectively. Several variations of this tool that could be used in future study include:

• Multi-year Allocation: Implementing a set amount of water allocation for each user with the flexibility for unused portions to be traded for credits as a means to optimize water usage more efficiency.
• Conservation Benefits: Recognizing and celebrating positive outcomes that result in aquifer conservation or improvement with financial incentives could serve as a motivational factor for sustainable water usage practices. This could also explore the tool as a means for conservation education with education sessions integrating demonstration of positive soil and water effects of regenerative agricultural practices such as cover cropping or low till practices.
• Governance Measures: considering state or local interventions including shutting down wells when no tangible improvement in the aquifer’s condition is observed could contribute to changes in observed pumping behavior. Here this behavioral economic game could aid in the development of thresholds for payments for ecosystem services (PES) and other financial and non-financial incentives that would pay local ranchers and farmers to fallow their land.

4 Discussion and conclusion

The data-informed digital behavioral economic game developed in this study offers a novel and effective tool for identifying technological, financial, and social mechanisms to reduce groundwater pumping in agricultural communities across the Western US. By simulating complex interactions, the game provides a unique platform for testing and refining policy approaches before their implementation in real-world settings, potentially leading to more effective and tailored groundwater management strategies. The game’s design allows for the evaluation of various scenarios and interventions, providing insights into their relative effectiveness. It demonstrated the impact of data transparency by allowing participants to compare their irrigation practices locally, fostering a sense of community accountability. Financial incentives were explored through varying subsidy thresholds and baseline Pigouvian tax scenarios, with results showing reduced water pumping when participants were incentivized with financial subsidies. Moreover, the game revealed the development of collaborative behaviors among participants, as players often coordinated their decisions to conserve water as a group, suggesting the potential for community-wide adaptive strategies.

Experimental results from the series of games run with key stakeholders validated the tool’s utility in determining how data-sharing mechanisms and financial incentives may influence groundwater use and conservation within agricultural communities. This study tool represents a departure from more commonly used survey and interview methods, which may present challenging biases and obscure the complex factors motivating behavior change and collective action towards improved water resource management. The game’s potential extends beyond its immediate research applications. It can be used to develop and validate local policy measures that are effective in conserving shared resources. By testing these measures in a simulated setting before implementation, communities have the opportunity to discuss concerns and make necessary adjustments, fostering greater community orientation towards practical self-regulation. This process can help communities address the challenges of managing common pool resources and avoid the “tragedy of the commons” by developing informal or local institutions tailored to their specific needs and circumstances.

While the game offers numerous advantages, it is important to acknowledge its limitations. The simplified nature of the game, particularly in terms of irrigation decision-making processes, may not capture all the nuances of real-world agricultural practices. Additionally, the game’s effectiveness may vary depending on the participants’ engagement levels and their ability to translate game experiences to real-world scenarios. Future iterations of the game could address these limitations by incorporating more detailed irrigation scheduling or by developing mechanisms to better bridge the gap between game play and real-world decision-making.

In conclusion, this digital behavioral economic game represents a powerful tool for informing and empowering communities to actively engage in collaborative problem-solving processes. By providing a platform to explore, experiment, and pilot new institutional innovations, it has the potential to significantly improve the community sustainability of groundwater pumping. As water scarcity continues to be a pressing issue in the Western US and globally, such innovative approaches to resource management and community engagement will be crucial in developing effective, locally-tailored solutions for sustainable water use.

Acknowledgments

The authors thank the expert stakeholders and game participants. This study was funded in part through support from Deloitte Consulting LLP and The National Science Foundation Convergence Accelerator (Award 24C0011). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsors. The authors also acknowledge valuable contributions to this work in programming and graphic design of the game from Łukasz Kowalski, Pawel Wiercinski, Bartosz Naprawa, Vladyslav Zoloto, and Michalina Kulakowska from the Centre for Systems Solutions. The groundwater game platform can be accessed at https://play.socialsimulations.org/ by Centre for Systems Solutions, Play Social Simulations Platform.