Hi!
I am a fifth-year Economics PhD-student at Heidelberg University under supervision of Jörg Oechssler, you can find my CV here.
I use theory and (natural and laboratory) experiments to address questions about the role of (AI) technology in human capital formation, and integrating information into beliefs when multiple interpretations are available.
I completed the PhD-courses at Mannheim University and spent part of my fourth year visiting Ran Spiegler at University College London.
Contact me at julian.matthes [at] awi.uni-heidelberg.de.
Advanced Work in Progress (ask me for a draft)
On the Demand for Mental Models (with Katharina Momsen)
- Theory+laboratory experiments: Do news impact beliefs if they contain no new information, but only interpretations of publicly available data? How do people value such news relative to new information?
- Presented at: Heidelberg Economics Seminar (2024), VfS Jahrestagung (Berlin, 2024), Zeuthen Workshop on Cognitive Foundations of Economic Behavior (Copenhagen, 2025), HeiKaMaX (Mannheim, 2025).
Tempted by AI: Skill formation, present bias, and performance standards (with David Piazolo)
- Theory: What individual characteristics determine the choice of delegating ability-forming tasks to AI?
- Presented at: ESA (Barcelona, 2026), VfS Jahrestagung (Innsbruck, 2026), CMID (York, 2026).
Methodological Norms and Research Bias (with Clément Staner)
- Theory: Can scientific results be unbiased if researchers have a preference for commonly used specifications?
Publications
Don't Put All Your Legs in One Basket - Theory and Evidence on Coopetition in Road Cycling (with David Piazolo)
European Economic Review: Volume 170 (2024)
- Theory and natural experiment: How do characteristics of competing groups impact cooperation within groups? Applications in industrial organisation. Empirical analyses using data from road cycling races.
Finding Large Rainbow Trees in Colourings of 
The Electronic Journal of Combinatorics: Volume 30, Issue 4 (2023)
- I prove that asymptotically, the bipartite graph with 2n vertices can be partitioned into n copies of any tree with n edges.
Work in Progress
Repeated Contests with Performance Enhancing Technology (with David Piazolo)
Wiggle Room - Uncertainty Attitudes and Motivated Reasoning (with Clément Staner)
Economic Decision Making during Moderate Physical Exercise (with Carlo Dindorf and Sebastian König)