COS 68-3 - The evolution of forecasting for decision making in dynamic environments

Abstract: Global change is reshaping ecosystems and societies. This means that people are facing changing environments as they make decisions about their behavior in social-ecological systems. These behaviors often impact the dynamics of the environment (for example by depleting resources or emitting pollution), which in turn shape future strategic settings and may result in persistent environmental oscillations. Economic and evolutionary models of behavior make dramatically different assumptions about how people make decisions as well as divergent predictions about outcomes in dynamic environments. In other words, how people choose behavioral strategies can have important consequences for socio-environmental dynamics. Our objective is to explore a middle ground between economic rationality and evolutionary myopia in models of social-ecological systems and examine the conditions in which different modes of decision making can be favored.

Building from an evolutionary model of myopic agents who consider only the present state of the environment when making decisions, we study the emergence of a new type of agent that forms environmental forecasts when making strategic decisions. We find that forecasting types can have an advantage in changing environments, even when forecasting is costly. Forecasting types can invade but not overtake a population, resulting in coexistence with myopic types. This result provides a theoretical explanation for multiple coexisting modes of decision-making that have been observed in behavioral economic studies. Moreover, forecasters provide a public good because their behavior reduces the amplitude of environmental oscillations and increases mean payoff to forecasting and myopic types alike. This has practical implications for the stewardship of common-pool resources because it suggests that even a small group of forecasting types can dramatically alter the collective dynamics of people and their shared ecological resources.