Two new projects at DTU Management have gotten funding from The Independent Research Fund Denmark DFF-Research Project2!
The Independent Research Fund Denmark finances groundbreaking research of the highest international quality, and for the DFF-Research Project 2, Associate Professor Toke Reinholt Fosgaard and Professor Stefan Røpke from DTU Management are two of the amazing researchers who have received funding.
The project “Who are the Tax Evaders? Understanding, Predicting, and Decreasing Tax Evasion” by Toke Reinholt Fosgaard aims to provide a more detailed understanding of the motivations of individuals who have been caught evading taxes or attempting to do so. Tax collection is crucial for a well-functioning society, but some citizens attempt to avoid paying taxes. Despite this longstanding issue, we lack precise knowledge about who the tax evaders are and what drives their behavior. In collaboration with the Danish tax authorities, both tax evaders and taxpayers are invited to participate in an online experiment with resulting data being used to identify behavioral profiles associated with tax evasion.
Existing research frequently relies on indirect measures, laboratory experiments, self-reported motivations, or aggregated data, which offer only partial insights. Building on these insights the project aims to design and implement an intervention aimed at promoting tax compliance. Furthermore, the insights gained will be used, in collaboration with the tax authorities, to design an intervention aimed at increasing tax compliance and test it in a natural field experiment among taxable citizens.
The project” Learning-guided optimization” by Stefan Røpke is about solving optimization problems that play a critical role in improving decision-making in both business and broader societal contexts. Poor decisions can lead to significant costs, delays, or environmental impacts. While existing methods that produce optimal solutions are often computationally intensive and lack scalability for larger problem sizes, heuristic approaches offer faster, more scalable alternatives.
Learning-Guided Optimization (LEGO) seeks to address these limitations by integrating Machine Learning (ML) algorithms to improve both exact and heuristic methods, with the goal of making both approaches more practical and applicable in real-world settings.
Congratulations to Associate Professor Toke Reinholt Fosgaard and Professor Stefan Røpke!
