Solving Nash Equilibria in Nonlinear Differential Games for Common-Pool Resources

Yongyang Cai · 2025

Read Solving Nash Equilibria in Nonlinear Differential Games for Common-Pool Resources by Yongyang Cai (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-06646.txt.

Introduction

Yongyang Cai, Anastasios Xepapadeas, and Aart de Zeeuw give in a 2025 paper new numeric methods for Nash equilibria in games over shared resources. Such resources sit in ecosystems that can tip past a pollution line, losing services at a stroke. Rival users also overuse the pool when each acts alone.

Two classic non cooperative answers exist for these dynamic games, open loop and feedback equilibria. The methods here derive both for systems of one and two state variables. The test case is the lake game, a standard model of shared water quality. Two dimensional feedback equilibria are the paper's novelty, a first for this class. That solution lands near the cooperative one, a hopeful sign for policy.

Catalog record

CREATORYongyang Cai
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE68 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-06646

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Solving Nash Equilibria in Nonlinear Differential Games for Common-Pool Resources Thanks: Cai acknowledges support from the National Science Foundation grant RISE-2108917. Thanks: Department of Economics, University of Bologna, Italy, and Athens University of Economics and Business, Greece, [email protected] Thanks: Department of Economics, Tilburg University, the Netherlands, and Beijer Institute for Ecological Economics, Royal Swedish Academy of Sciences, [email protected] Yongyang Cai Abstract Many resources are provided by an ecological system that is vulnerable to tipping when exceeding a certain level of pollution, with a sudden big loss of ecosystem services. An ecological system is usually also a common-pool resource and therefore vulnerable to suboptimal use resulting from non-cooperative behavior. An analysis requires methods to derive cooperative and non-cooperative solutions for managing a dynamical system with tipping points. Such a game is a differential game which has two well-defined non-cooperative solutions, the open-loop and feedback Nash equilibria. This paper provides new numerical methods for deriving open-loop and feedback Nash equilibria, for one-dimensional and two-dimensional dynamical systems. The methods are applied to the lake game, which is the classical example for these types of problems. Especially, two-dimensional feedback Nash equilibria are a novelty of this paper. This Nash equilibrium is close to the cooperative solution which has important policy implications. Key words: Dynamical optimization algorithms, differential games, Nash equilibria, tipping points, lake model, two-dimensional dynamics JEL codes: C63, C73, Q53 1 Introduction An important characteristic of managing common-pool resources embedded within ecological systems—whether in the context of pollution control or biomass harvesting—is that individual actions of the economic agents, such as emissions or resource extraction, generally result in accumulation of pollution stocks or depletion of resource availability and that each agent is affected by the future stocks emerging from all the combined individual actions. These impacts are typically damages from the pollution stocks, increase in harvesting costs, or loss of ecosystem services. When each agent takes into account the impact of future stocks on its own well-being only, the outcome is excess pollution or resource overexploitation relative to an outcome in which the aggregate well-being of all agents is considered. Typical examples associated with common-pool or open access resources are climate change with excess carbon emissions, phosphorus pollution of closed water bodies or closed seas (e.g., the semi-enclosed Baltic sea), or overfishing in open access fisheries (e.g., the Northern cod fishery in Newfoundland). The management of these issues involves the integration of economic mechanisms with natural processes describing the evolution of ecosystems which are affected by the actions of economic agents. This usually involves two approaches. In the first, the objective is to maximize the aggregate benefits of all economic agents subject to the constraints emerging from the evolution of the ecosystems. This is the optimal management approach that leads to an outcome which can be regarded as the welfare-maximizing cooperative solution to the management problem. In contrast, in the second approach each agent maximizes own benefits by considering the constraints associated with the evolution of the ecosystems and taking into account the actions of the other agents as they seek to optimize their own benefits. This is the noncooperative solution to the management problem. The solution to the optimal management problem may be obtained by standard optimal control or dynamic programming methods (see, e.g., Rust, 1996; Judd, 1998; Ljungqvist and Sargent, 2000; Miranda and Fackler, 2002; Bertsekas, 2005; Bertsekas 2007; Cai, 2019) while a noncooperative problem may be solved in the context of differential games (Başar and Olsder, 1982; Dockner et al, 2000). The most often used solution concepts for the differential games are the weakly time consistent open-loop Nash equilibrium (OLNE) and the strongly time-consistent feedback Nash equilibrium (FBNE) which possesses the Markov perfect property. These solutions are obtained either by optimal control (for OLNE) or dynamic programming methods (for FBNE) (Mäler et al, 2003; Kossioris et al, 2008; van der Ploeg and Zeeuw, 2016; Jaakkola and Wagener, 2023). 1 1 1 For discrete time dynamic games, OLNE may be numerically solved with an iterative method (see, e.g., Cai et al., 2023), and FBNE may be numerically solved with a time backward iteration method (see, e.g., Cai et al., 2019). An important characteristic of these problems, and one that imposes major challenges to their solution, is the fact that the appropriate description of ecosystem dynamics could include nonlinear feedbacks. A prototype management problem that has been used as a vehicle for studying cooperative and noncooperative solutions under nonlinear ecosystem dynamics, positive feedbacks and strategic interactions is the so-called lake problem (Carpenter and Cottingham, 1997; Scheffer, 1997). In the lake ecosystem, economic agents benefit from emitting phosphorus 2 2 2 Benefits can be associated with agricultural production which generates phosphorus emissions as runoff. The phosphorus runoff is accumulated in the lake. The Baltic Sea provides a real-world example of this problem. into a lake but the accumulation of the phosphorus stock generates collective damages. The positive nonlinear feedback emerges from the release of phosphorus from sediment at the bottom of the lake that has accumulated as mud. 3 3 3 In the study of climate change, positive feedbacks could emerge from the loss of sea ice that reduces the Earth’s albedo, or the thawing of permafrost that releases CO2 and methane. The problem can be analyzed as either a one-dimensional (1D) problem in which there is one transition equation describing the evolution of the phosphorus stock in the lake with positive feedbacks emerging from a fixed amount of mud, or a two-dimensional (2D) problem in which the transition dynamics for both the phosphorus and the mud stocks are included. The economic management of the lake problem has mainly been analyzed in one dimension in which the optimal management and OLNE solutions were obtained using optimal control methods, while the FBNE problems were solved in a dynamic programming framework (Mäler et al, 2003; Kossioris et al, 2008; Jaakkola and Wagener, 2023). Optimal management and OLNE solutions for the 2D problem were acquired by Grass et al. (2017), but to the best of our knowledge there has been no FBNE solution for the 2D lake problem. FBNE solutions for the 1D differential game of the lake problem have been obtained in the dynamic programming context by assuming time stationary equilibrium strategies and a differentiable value function. Differentiating the value function of the associated Hamilton-Jacobi-Bellman equation with respect to the state and using the optimality conditions for the control, a nonlinear ordinary differential equation in the feedback equilibrium strategy with respect to the state is obtained. Solution of this differential equation determines the Markov-perfect equilibrium strategy (Maskin and Tirole, 2001). Kossioris et al. (2008) used this approach to obtain a “partial solution” of the problem. 4 4 4 Partial solution in this context means that the Markov-perfect equilibrium strategies could not be obtained when initial conditions for the phosphorus stock belonged to a certain subset of the state space. Dockner and Wagener (2014) and Jaakkola and Wagener (2023) extended this approach and generated a complete solution that includes discontinuities of the Markov-perfect equilibrium strategies. The main contribution of this paper is to provide a novel method, called strategy function-value function (SFVF) iteration, for obtaining FBNE solutions. For stationary dynamic games with smooth value or policy

Read the complete text