Description

In the face of risk and uncertainty, how does new information affect economic decisions? The course examines this question both within the workhorse model of Bayesian belief updating and in more recent belief-free models. Participants in the course will obtain insights into questions, methods, and applications, preparing well for further research.

In the Bayesian model, we first review properties of updating beliefs in single-player learning environments. In applications to binary hypothesis testing, emphasis is placed on martingale properties of beliefs and likelihood ratios. We also discuss the comparative statics of actions with respect to beliefs and information, including the monotone likelihood ratio property. We turn to two more advanced topics. One is the comparison of information structures as first proposed by Blackwell, and later extended by Lehmann. The other is application to multi-player settings, emphasizing social learning and herding with learning cascades.

In the belief-free approach, we start from static choice problems. This part starts from alternative interpretations of belief elicitation outcomes. The part continues with a theory of beating the play, reacting to out-of-equilibrium play in a game without beliefs, including an application to robust bidding in auctions. In this context, minimax regret has interesting properties. The remaining part considers dynamic choice problems. Social learning can be effective through imitation, without any need for beliefs – reinforcement learning (well-known also from large language models) provides a flexible belief-free tool for learning and responding to new information. The non-guessing approach similarly allows for good outcomes in the true state without any need for beliefs.

> course description

Course format

The two course organizers provide lectures on these topics. We assume that students have already learned game theory at a level that is typical for programs in Economics. Students are expected to prepare before the lectures, both to obtain a basic understanding of the required reading material, and to think of questions.

Both days will have course lectures in the morning and afternoon (9:00-12:00, 14:00-17:00 on Day 1, 8:45-10:15, 13:00-16:30 on Day 2). Moreover, on the second day, a session includes presentations of current research by both lecturers (10:30-12:00). All lectures will take place at Aarhus BSS, Aarhus University, Universitetsbyen 51, DK-8000 Aarhus C, building 1811, room 111.

Towards the end of the course, we briefly list some topics for further work. In order to obtain credit for the course, each student must subsequently write a short paper on one such topic, or a similar topic approved by the organizers. More information is provided during the course.

ECTS

In order to obtain credit for the course, each student must subsequently write a short paper on one such topic, or a similar topic approved by the organizers. More information is provided during the course.

For PhD students at AU (ECON), this course has been pre-approved as an internal BSS PhD course equivalent to 2 ECTS.

Lecturers

Professor Peter Norman Sørensen, University of Copenhagen, https://www.econ.ku.dk/ansatte/vip/?pure=da/persons/5862
Professor Karl Schlag, University of Vienna, https://homepage.univie.ac.at/karl.schlag/

Registration

Registration deadline: 12 June 2026 via https://event.au.dk/events/dgpe-learning-information-in-economic-settings/registration

The course is free of charge for:

1. DGPE members from AU, KU, CBS, AAU and SDU.
2. PhD students from Economics Departments at Nordic universities outside Denmark.

For other participants, the course fee is EUR 100.

See other DGPE courses here: https://econ.au.dk/talent-development/phd-programme/phdcourses/dgpe-courses

Contact

Academic: Nicola Maaser, nmaaser@econ.au.dk
Administrative: Susanne Christensen, sch@econ.au.dk