Dynamic models driven by judgment and decision-making from behavioral and experimental economics better capture how people make decisions compared to classical models based on delay time constants and multiplier effects or table functions. The underlying assumptions on table functions are sometimes challenging to justify. Behavioral economists elect to use discrete-time modeling with difference equations. In system dynamics, we build models with the concept of continuous-time and do it on a discrete machine, i.e., a digital computer, and this requires approximating the time unfolding continuously with discrete time steps. This leads to confusion in the system dynamics community on how to properly model discrete-time features in continuous-time models. Even though system dynamicists commonly use continuous-time models with differential equations, discrete-time models are consistent with system dynamics. This issue is important in system dynamics in general and particularly in bringing behavioral economics models with richer, more realistic human decision-making structures and empirically verified with human decision subjects into system dynamics modeling. We first clarify this issue to the system dynamic community. Also, we do this to make sure by replicating correctly behavioral economics models and structures into system dynamics modeling and looking at the issue of discrete-time and continuous-time in a manner that satisfies both behavioral economists and system dynamicists, and we preserve the results published in the behavioral economics literature. We developed a formal, teachable method of bringing behavioral economics into system dynamics. We applied and validated this method using a macro model, the Samuelson multiplier-accelerator model, a micro model, the Cobweb model, and the behavioral economics model, the quasi-hyperbolic discounting model. We also applied this method in two chapters of this dissertation. First, to improve an existing model and paper published in 2019 in the system dynamics review journal and second, to revisit and explore two well-known lifecycle models. We replicated Brumberg’s and Modigliani’s lifecycle hypothesis in system dynamics and built Shefrin’s and Thaler’s behavioral lifecycle hypothesis from the description given in their paper. We also tested a couple of predictions provided by Shefrin and Thaler in their behavioral lifecycle hypothesis paper.

Creator

• Bah, Souleymane

Colaboradores

• Radzicki, Michael J.
• Johnson, Sharon
• Smith, Alexander D.

Degree

• PhD

Unit

• Social Science & Policy Studies

Publisher

• Worcester Polytechnic Institute

Identifier

• etd-110831

Palavra-chave

• Behavioral Lifecycle Hypothesis
• Behavioral Economics
• Discrete-time modeling
• Quasi-hyperbolic Discounting
• System Dynamics
• Continuous-time modeling

Advisor

• Smith, Alexander D.
• Radzicki, Michael J.
• Johnson, Sharon

Committee

• Radzicki, Michael J., Dr.

Defense date

• 2023-05-25

Year

• 2023

Date created

• 2023-05-30

Resource type

• Dissertation

Source

• etd-110831

Rights statement

• In Copyright

Última modificação

• 2023-10-09