Dibya Mishra
Economist studying decision-making, markets, and AI
Chicago, Illinois
About
I am an economist with a PhD in Economics from Rice University and an MS in Computer Science from Georgia Tech. I study how people, firms, and increasingly AI systems make decisions when information and attention are imperfect.
My work combines economic theory, structural and causal econometrics, machine learning, and computational methods. My economics research studies consumer choice, healthcare markets, and allocation; my recent AI work examines model decision-making, interpretability, evaluation, and incentives.
A recurring theme across my work is the gap between what a decision-maker could optimize and what it actually considers, observes, or learns.
This site is my public research page and journal. The research page lists my papers, the research agenda page describes the questions tying that work together, the journal holds notes on what I am reading and working through, and the projects page covers things I build to learn.
Research themes
Decision-making & representation
Consumer consideration, discrete choice, model representations, interpretability, and how human or artificial agents form and use internal signals when making choices.
Markets, incentives & allocation
Healthcare, digital markets, market design, AI economics, monitoring, incentives, matching, and resource allocation.
Empirical & computational methods
Structural econometrics, causal inference, machine learning, optimal transport, and computational methods for studying economic and AI systems.
Selected Research
Quality Considerations and Universal Healthcare
How patients form limited consideration sets over hospitals, and whether information and empanelment incentives can expand them — a structural model estimated on a large public insurance scheme in India.
Pennies for Pads: Subsidized Menstrual Health Products and Adoption in India
How price and availability shape adoption of a health product, using the staggered rollout of subsidized pharmacies as a natural experiment in access.
Current Directions
AI decision-making & interpretability
How language models represent preferences, comparisons, utilities, or other decision-relevant information, and when those representations actually affect behavior.
Economics of AI & incentives
How ideas from principal-agent theory, information economics, mechanism design, and evaluation can inform the design and governance of increasingly capable AI agents.
Choice, allocation & optimal transport
Using optimal transport and related tools to study choice under changing environments, matching, allocation, and distribution shift.
From the journal
Alignment × Econ
What can economics teach us about building and evaluating AI agents? A personal essay series in my journal, connecting classic ideas from economics — incentives, information, contracts, mechanism design, and strategic behavior — to the problem of monitoring and evaluating AI systems.
You Can't Reward What You Can't See
What a classic 1979 economics paper can teach us about monitoring and aligning modern AI systems.
Additional essays in the series are forthcoming.