Matthew Kovach - Research

Working Papers (My arXiv page)

Inertial Updating with General Information (with Adam Dominiak and Gerelt Tserenjigmid) - Revise and Resubmit at American Economic Review (slides)
- idea: you are told that some outcome of interest will occur with some range of probabilities. how do you update your beliefs? the answer to this and a characterization of f-divergences inside.
  - This paper was previously titled Minimum Distance Updating With General Information.

Inertial Updating (with Adam Dominiak and Gerelt Tserenjigmid) - NEW!
- idea: a framework for belief updating unifying Bayesian and non-Bayesian updating rules, along with rules for updating on null-events.

Misspecified Model Estimation and Its Impact on Predictions (with Junnan He, Lin Hu and Anqi Li) - Revise and Resubmit at American Economic Review
- idea: how do you learn about media sources when you misperceive the bias in news reports? learn about that and more here. [arXiv]

Can an LLM Learn Preferences from Choice Data? (with Jeongbin Kim, Kyu-Min Lee, Euncheol Shin, and Hector Tzavellas) - Revised and Resubmitted to The Review of Economics and Statistics Slides
- idea: some people

Conservative Updating - under revision
- idea: some people do not change their beliefs enough when provided information. here's a preference based characterization of such behavior.

Selling to Wishful Thinkers (with Tommy Chan) - coming soon, preliminary draft available upon request.
- idea: some people are overestimate the chances of good events. you can extract surplus from such wishful thinkers by introducing uncertainty.

The Focal Quantal Response Equilibrium (with Gerelt Tserenjigmid) - under review - NEW!
- idea: some options draw more attention than others, even after controlling for their utilities. we incorporate this idea into a strategic setting and show accounting for focality explains some experimental findings.

Learning from an Unknown DGP: Experimental Evidence on Belief Updating with AI Recommendations (with Daniel Martin and Gerelt Tserenjigmid) - NEW!
- idea: how do people learn from opaque AI recommendations? experimental evidence inside. [arXiv]

Published and Accepted Papers

Ambiguity and Partial Bayesian Updating - Economic Theory. Vol. 78: 155–180, August 2024. [arXiv]
- Idea: How should a set of beliefs be updated? One way is to first select those that are "pretty likely" given the data, then use Bayes' rule. You can find a behavioral analysis of this and more inside.

Reference Dependence and Random Attention (with Elchin Suleymanov) - Journal of Economic Behavior and Organization. Vol. 215: 421-441, November 2023. [arXiv]
- - Idea: How does a status quo interact with random attention?
  - This paper combines and supersedes "Thinking Inside the Box: Status Quo Bias and Stochastic Consideration" (Kovach 2016) and "Stochastic Attention and Search" (Suleymanov 2018).

Behavioral Foundations of Nested Stochastic Choice and Nested Logit (with Gerelt Tserenjigmid) - Journal of Political Economy, Vol. 130(9): 2411-2461, September 2022. [arXiv]
- Idea: Nested logit is the most widely applied generalization of multinomial logit, but it has not been behaviorally characterized. Until now.
- Finalist, 2023 Exeter Prize

The Focal Luce Model (with Gerelt Tserenjigmid) - American Economic Journal: Microeconomics, Vol. 14(3): 378-413, August 2022. - online appendix
- Idea: Multinomial logit is great, but what if a decision maker pays more attention to some alternatives than others? It turns out you can identify these "focal" alternatives and distinguish utility from bias.

A Model of Stochastic Choice from Lists (with Levent Ulku and Yuhta Ishii) - Journal of Mathematical Economics, Vol. 96: October 2021. - slides
- Idea: Menu variation is great, but what if all you can change is the order of alternatives? Can we still learn about preferences? Yes.

Twisting the Truth: Foundations of Wishful Thinking - Theoretical Economics, Vol. 15(3): 989–1022, July 2020.
- Idea: Wishful thinking based on an endowment "twists" beliefs a bit. Behavioral consequences inside.

Satisficing with a Variable Threshold (with Levent Ulku) - Journal of Mathematical Economics, Vol. 87: 67-76, March 2020. slides
- Idea: What can we say about a satisficer with an unobserved and changing notion of "good enough?"

Select Work in Progress

Trading with AI (with Jeongbin Kim, Kyu-Min Lee, Euncheol Shin, and Hector Tzavellas) - in IRB / programming stage
ChatGPT Plays the 10 Little Treasures of Game Theory (with Sudipta Sarangi, Hector Tzavellas, and Michael Wagnon) - in data collection stage
Impressionistic Luce Rules (with R. Sonal and P. Cheung)

Permanent Working Papers

Thinking Inside the Box: Status Quo Bias and Stochastic Consideration - working paper - these results have been incorporated into "Reference Dependence and Random Attention."
Ordered Surprises and Conditional Probability Systems (with Adam Dominiak and Gerelt Tserenjigmid) - major results have been incorporated into "Inertial Updating."
- Idea: How do you revise your beliefs when you're surprised? We introduce a notion of ordered (im)possibility and show that it provides a simpler foundation for CPS.