AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Optimizing Step By Step: A Theoretical Framework for Behavioural Economics Based on the Variational Free Energy Principle

Author

Term

4. term

Education

Publication year

2019

Submitted on

Pages

70

Abstract

Dette speciale foreslår en ny, mere realistisk måde at forstå handling, optimering og valg i økonomi. Det trækker på variational free energy-princippet (et matematisk rammeværk, der beskriver, hvordan systemer handler for at reducere uventede udfald og usikkerhed) for at nuancere rational choice-teorien ved at omformulere, hvordan aktører optimerer. Med denne tilgang opstår adfærd, der naturligt rummer mange af de kognitive bias, som er beskrevet i adfærdsøkonomi. I stedet for at behandle disse bias som beviser på markedsineffektivitet eller -fejl undersøger specialet, hvilke grænser der er for, i hvilket omfang de kan informere eller kritisere standardøkonomisk teori. Resultatet er en mellemposition: Den neoklassiske aktør er hverken så fuldkommen rationel, som rational choice-teori antager, eller så gennemgående irrationel, som adfærdsøkonomi ofte antyder.

This thesis proposes a new, more realistic way to think about action, optimization, and choice in economics. It draws on the variational free energy principle (a mathematical framework that describes how systems act to reduce unexpected outcomes and uncertainty) to temper rational choice theory by reformulating how agents optimize. Framing choice in this way yields behavior that naturally includes many cognitive biases described in behavioral economics. Rather than treating these biases as evidence of market inefficiencies or failures, the thesis examines the limits of how far they can inform or critique standard economic theory. The result is a middle-ground perspective: the neoclassical agent is neither as perfectly rational as rational choice theory assumes nor as broadly irrational as behavioral economics sometimes suggests.

[This abstract was generated with the help of AI]