Biased beliefs in dynamic decisions

Current Project

Department

Public Economics

Project Management

Peter Haan

Project Period

January 1, 2017 - December 31, 2020

Commissioned by

Deutsche Forschungsgemeinschaft (DFG)

In Cooperation With

Humboldt-Universität zu Berlin
Ludwig-Maximilians-Universität München

In many dynamic decision problems, the relevant market prices and budget constraints are random variables whose realizations lie in the future. This makes optimal decision-making difficult as optimality requires an accurate quantification and interpretation of the dynamic processes operating in the market. The studies in this project model and test the agents’ expectations about relevant future variables in competitive markets, especially in labor markets and in markets for savings products. We build on the existing behavioral literature on biased decision making. In its discussions of forward-looking decisions, this literature mostly formulates preference-based biases whereas it only rarely formulates belief-based biases. This is prominently true in contexts of intertemporal financial decisions where discussions of hyperbolic preferences dominate the literature. In contrast, we focus on belief-based biases. Most of our contexts involve decisions that are relevant over the decision-maker’s life-cycle and involve dynamic optimization and/or inferences from probabilistic updating. We combine theoretical, experimental and econometric methods, including dynamic structural econometrics. A recurring theme in our empirical strategies is to use a mix between laboratory experiments and survey evidence from the longitudinal SOEP data set. Together with the projects by Engelmann and Winter (A1) and by Menkhoff (A3), we will conduct a household survey module where we ask for expectations regarding outcomes in labor markets and financial markets, among other things.

DIW Team

Contact

Peter Haan

Head of Department in the Public Economics Department

Partners