Research
Publications
“Resource Allocation, Computational Complexity, and Market Design”
with Peter Bossaerts, Elizabeth Bowman, Felix Fattinger, Harvey Huang, Carsten Murawski, Anirudh Suthakar, Shireen Tang, and Nitin Yadav.
With three experiments, we study the design of financial markets to help spread knowledge about solutions to the 0-1 Knapsack Problem (KP), a combinatorial resource allocation problem. To solve the KP, substantial cognitive effort is required; random sampling is ineffective and humans rarely resort to it. The theory of computational complexity motivates our experiment designs. Complete markets generate noisy prices and knowledge spreads poorly. Instead, one carefully chosen security per problem instance causes accurate pricing and effective knowledge dissemination. This contrasts with information aggregation experiments. There, values depend on solutions to probabilistic problems, which can be solved by random drawing.
Working Papers
“How complex are financial decisions? Evidence from credit card choice”
with Carsten Murawski and Nitin Yadav.
Complexity is increasingly recognized as a barrier to sound financial decisionmaking, yet no theoretical framework currently characterizes it. We propose a framework that quantifies the complexity of financial decisions based on their computational demands and apply it to binary credit card choices. Using a combined behavioral and eye-tracking experiment, we show that complexity, rather than inattention or limited financial literacy, drives mistakes, and that it disproportionately burdens cardholders who carry a balance, for whom the stakes are highest. Our findings suggest that complexity is a fundamental barrier to optimal financial decision-making, even in seemingly simple choices.
Work in Progress
“Improving life insurance decision-making: The role of calculators”
with Carsten Murawski, in collaboration with AIA Australia Ltd.
It is well-documented that people struggle to determine appropriate life insurance (LI) coverage, leading to under-insurance and financial hardship for dependents. This study (N = 624) examines how LI calculators and their recommendations can be used to improve coverage decisions. We show that simply prompting individuals to consider their needs is ineffective, while providing a recommendation produces strong anchoring. When given the choice, most participants opt to use the calculator, indicating strong demand for decision aids. Lastly, we show that coverage choices vary significantly by age, income, and bequest motives, highlighting significant heterogeneity in LI decision-making.
“Complexity undermines resource allocation”
with Michele Garagnani, Carsten Murawski, and Nitin Yadav.
Pre-registered on OSF.