Hello! I am Yiqi Li (pronounced YEE-chee), a PhD candidate in quantitative marketing at the Ross School of Business, University of Michigan. I apply structural modeling, causal inference, and experimental methods to examine how information disclosure policies affect consumer decision-making, with the focus on nutrition labeling.I obtained a Master's degree in Economics from Duke University and a Bachelor's degree in Economics from Peking University. Outside of research, I enjoy marathon running, triathlons, classic Chinese fiction, and food.I am on the the 2025-2026 academic job market.

Education

PhD in Quantitative Marketing, University of Michigan Ann Arbor, Expected 2026
MA in Economics, Duke University
Bachelor in Economics, Peking University

Text

Teaching Experience

Instructor:
• Marketing Management, Core course for undergraduate business minors, University of Michigan, Spring 2022
Course evaluation: 4.8/5

Teaching Assistant:
• Marketing Strategy for the Digital Age (MBA), University of Michigan, Winter 2023, 2024
• Marketing Core (EMBA), University of Michigan, Fall 2022, 2023, 2024
• Marketing Research and Analytics (Master), University of Michigan, Fall 2020, 2021, 2022, 2023
• New Product and Innovation Management (MBA), University of Michigan, Fall 2021
• Advanced Microeconomic Theory (Master), Duke University, Spring 2019
• Microeconomics (Master), Duke University, Fall 2018
• Principles of Economics II (Undergrad), Peking University, Winter 2016

Testimonials from Marketing Management students
• "Really great professor. She is awesome, kind, and a great educator. Give her a raise!"
• "I loved learning about concepts and then seeing real life examples of them. It really helped me to understand the concepts deeply.
Furthermore, the examples used were very good ones, I found that I could recall them pretty easily when studying and taking my
exam."
• "The professor was very helpful and was always open for questions at any time. This course had group projects which was very fun
and helpful"

Research

The Tale of Two Nutrition Labels: Rational Inattention and Consumer Learning (JMP)

with Anocha Aribarg and Matthew Osborne
Summary: To promote healthy diets, the FDA recently proposed front-of-package (FOP) labels—"Healthy Claim" and "Nutrition Info Box" (NIB)—in addition to existing back-of-package Nutrition Facts Labels (NFL). We examine how FOP labels affect consumers' use of the NFL in their food choices using a randomized incentive-aligned experiment that directly measures attention to the NFL, nutrient belief and food choice. To provide policy recommendations, we develop a structural choice model with rational inattention that allows consumers to tradeoff utility gains from information with cognitive costs while relying on fewer assumptions of models based on observational data. We find that Healthy Claim and NIB reduce attention time on the NFL by 17% and 45%, with limited improvement in learning. While FOP labels increase healthy product choices, their effects on choice optimality are smaller. Suboptimal choices are driven by incorrect beliefs about nutrients. Across FOP labels, we find that NIB outperforms Healthy Claim in terms of learning and choice due to its higher informational value. These findings suggest policymakers prioritize comprehensive FOP labels that provide substantive nutritional information rather than relying on simple health cues.

Quantifying the Heterogeneous Impact of Recent FDA Changes to the Nutrition Facts Label

with Anocha Aribarg and Matthew Osborne
Summary: In 2020, the FDA updated the Nutrition Facts Label to emphasize added sugar while de-emphasizing total fat. Using combined store scanner and consumer panel data with longitudinal label information, we show that the new label decreased demand for products with high added sugar and increased demand for products with higher fat, with stronger effects in healthier categories such as yogurt. Using discrete-choice modeling of household purchase behavior, we find the strongest responses among consumers with an overweight BMI or those who self-perceive as overweight, as well as those who frequently read nutrition labels and are on a diet.

Sharing Car-sharing and Bike-sharing: An Empirical Analysis

with Puneet Manchanda, Emanuel Schuster and Martin Spann
Summary: We investigate cross-platform usage in shared mobility services by applying machine learning with graph-based methods to analyze user behavior patterns. Using data from a large European transportation company that operates both car-sharing and bike-sharing, we construct user-to-user networks based on shared station usage and temporal patterns. We then apply random forest classification to predict which single-mode users are most likely to adopt additional sharing services. Our approach integrates spatial usage patterns with temporal similarity measures to model user relationships, providing insights to inform targeted cross-selling strategies across multiple channels.

Impact of the Post-Roe v. Wade on Migration Patterns in the United States

This study examines how the 2022 overturn of Roe v. Wade affected migration patterns at the county level across America, leveraging consumer address history data tracking 13 million of US adults. Employing a difference-in-differences design for county-level net migration rates, a four-arm causal forest for move flows between counties, and a two-stage nested logit structural model, I find that migration patterns following the Dobbs decision varied systematically with local political composition. Counties with higher Democratic vote shares experienced higher net migration compared to Republican-leaning counties, though these patterns appear related to broader political sorting trends rather than abortion policy alone.