Utilizing AI for Developing Socially Beneficial Forest Management through Incentive Frameworks

 

Dr. Misti Sharp is my mentor within this project, we are associated with University of Florida in the College of Agriculture and Life Sciences. Within the sub section of Food and Resources Economics Department. ​ I've been Affiliated with the Research Project for 5 Months reading articles regarding Belize and its history of forest degradation. ​

​

Problem Statement:

As of now, there are various ways society has developed to manage forests ecologically and economically. Yet, many forests remain unmanaged or engage in burning behaviors that are detrimental to forest health and/or society (e.g. the Belizean savanna) (Smith 2021). Currently there are organizations (such as non-governmental organizations and producer groups) who choose to burn fires during the dry season instead of the wet season. This causes more unnecessary destruction of the Belizean savannas and forests, while decreasing local living conditions of residents living around the savanna. Residents are aware of the fire being damaging to both their health and livelihoods (in the case of the palmetto harvesters), and yet they lack the ability to affect change in fire behavior (Smith 2021). The residents “have little incentive to care about fires in these areas because they lack clear access rights to the savanna, most do not have crops in the savanna, and most do not directly benefit from the pine industry” (Smith 2021, p. 591). This is a major cause for concern for those within and outside of Belize as radical burning of vegetation within Belizean forests has increased the amount of carbon emissions in the atmosphere. This project aims at addressing fire behavior to ensure optimal forest growth and improved social outcomes in terms of carbon emissions. We plan to use an economic experiment to discern behavior of residents and their reactions to varying incentive structures that promote pro-social behavior (such as a carbon market). The economic experiment is helpful as it allows us to observe the behavior of participants in a short period of time under various market structures. Prior to implementing the economic experiment in Belize, this project will use AI to assess potential behavioral outcomes of the economic experiment. There is currently a proposed market for emissions reductions in Belize as laid out by the Australian government (Frydenberg 2018). This carbon market structure can only be successful if emissions are reduced, which relies heavily on the behavior of market participants. We plan to use iterative AI to explore a multitude of possible outcomes given market parameters and the cultural context of Belize to see if this market would be a viable opportunity for Belizeans. Using economic experiment software such as Otree, we will program potential experimental participants using AI to observe potential behavioral paths of participants in terms of burning behavior and environmental/ecosystem outcomes. The benefit of using this AI driven approach is it allows us to pre-test the experiment to be used in Belize anticipating outcomes.

 

Expected Results:

We expect to have a vast array of scenarios, where after repeated consistent testing, we would know of potential behavior of participants anticipating how best to teach participants about carbon markets and to anticipate likely outcomes from this economic experiment. Using AI will allow participants to have a clear understanding of choices they could make and the possible outcomes of their choices within a carbon market setting. This preliminary data will give us a baseline to work with once we apply the same AI experiment to Belize.

 

Timeline:

First, we will do an extensive literature review during the current spring 2024 semester. During the summer semester, we will develop a clear and concise experimental framework based on the spring literature review. Fall semester we will officially test the experiment using AI, with consistent repeated testing to ensure a broad understanding of potential outcomes. Finally, during the spring semester, Reese will write a thesis describing the outcomes of the AI tested economic experiment. He further plans to present this thesis at various venues including the spring Undergraduate Research Symposium and the Summer Applied Agricultural Economics Association student paper competition.

​​

​​