Research
Dissertation: Part 1
The theme in my Ph.D. Dissertation is firm’s competitive strategies in food retail sector. By working with my advisor, Dr. Steven Wu, we examine the profitability of price discrimination (PD) schemes under different information environments by combining nonlinear pricing theory with machine learning (ML) techniques in the first part of my Ph.D. dissertation. Theoretical PD models typically exogenously impose consumer segments using assumptions that may or may not be reasonable for specific applications. Given advances in computing power, ML methods can now replace exogenous assumptions in forming consumer clusters that can proxy different consumer segments, making theoretical models more applicable in practice.
The specific question we explore is how informationally robust PD techniques are. Using Nielsen HomeScan data, we construct a full information set including consumers’ demographics and purchase history, and a limited information set including only demographics. We then apply K-Means clustering, which is one of the unsupervised learning methods, to segment the market. Using this method, consumers with similar characteristics are assigned to the same group, helping firms target each group in an effective way. By selecting households from the 30 most populated counties in Nielsen HomeScan data, we estimate the consumer demand for yogurt products using full and limited consumer information. Applying demand estimation results to PD schemes, we show that using more information helps increase the store’s profit under second-degree PD. This paper titled “Estimating Consumer Segments and Choices from Limited Information: The Application of Machine Learning Methods” will be presented in a lightening session at 2022 AAEA conference. Our next step is to investigate whether additional information from consumers affects the firm’s profit under third-degree PD.
Qin, Fei, and Steven Y. Wu. “Estimating Consumer Segments and Choices from Limited Information: The Application of Machine Learning Methods.”
Dissertation: Part 2
In the second part of my dissertation, I worked with my advisor Dr. Meilin Ma, to investigate retail stores’ product assortment responses to the intensive price competition brought by Unit Pricing Regulation (UPR). Retail stores selling differentiated products are incentivized to obfuscate consumers by making direct price comparison complicated. UPR was introduced to help consumers make more informed decisions by requiring the store to display the unit price along with the item price for products on the shelf. Despite extensive research on UPR effects on consumers’ perceptions and decisions, much less is known about the impacts on retail stores’ price and non-price responses under UPR. Relying on the geographic variation in UPR implementation across states and using Nielsen scanner data, we empirically show that mass merchandizers remove products without changing prices, whereas grocery stores without same-chain stores not under UPR add brands and charge higher prices on average. Using a structural demand model to estimate consumer welfare changes, we find that, on average, consumer welfare falls for both retail formats, highlighting an unintended policy effect. This paper is selected and will be presented at 2022 AAEA conference.
Qin, Fei, and Meilin Ma. “Unit Pricing Regulation and Retailer Assortment: Evidence from the U.S. Yogurt Market.”
Other projects
Besides my dissertation, I have been involved in several research projects related to the food supply chain in the U.S. and China as well as the impact evaluation of interventions on education in rural China.
Published papers
Balagtas, Joseph, Joseph Cooper, Patrick McLaughlin, and Fei Qin. 2023. “Consumer Demand for Food at Home and Food away from Home: Understanding Economic Linkages during the Pandemic” Applied Economic Perspectives and Policy https://doi.org/10.1002/aepp.13361
Household food expenditure has shifted away from Food at Home (FAH) and towards Food Away from Home (FAFH). Prior to the COVID-19 pandemic, FAFH’s share of food expenditure surpassed that of FAH, reaching 55% in 2019. Yet economic research on FAFH and the interaction of FAFH and FAH has been limited. Combining scanner data for meat sales in grocery stores with data for FAFH expenditure, we estimate a model of demand for at-home meat, incorporating FAFH expenditure as a demand shifter. We quantify substitution between FAFH expenditure and FAH meat and quantify the impact of the COVID-19 disruptions to the food service sector on retail prices of FAH meat.
Qin, Fei, Huanmin Hu, Prashant Loyalka, Sarah-Eve Dill, and Scott Rozelle. 2022. “Stuck in the Middle School Rut: Can Anything Improve Academic Achievement in Rural Chinese Middle Schools?.” Journal of Development Effectiveness 14(4): 306-39 https://doi.org/10.1080/19439342.2022.2067890
Academic achievement in middle schools in rural China remains poor for many students. This study examines whether programmes and interventions can improve academic achievement by reviewing rigorous experimental evaluations of nine programmes (11 interventions) on 47,480 rural middle school students in China. The results find none of the interventions improved academic achievement. Moreover, we find no evidence for heterogeneous treatment effects by student gender, age or previous academic achievement. These results may be due in part to the academically-demanding nature of the middle school curriculum, which is applied universally to students with varying levels of cognitive ability.
Ma, Meilin, H. Holly Wang, Yizhou Hua, Fei Qin, and Jing Yang. 2021. “African Swine Fever in China: Impacts, Responses, and Policy Implications.” Food Policy 102: 102065. https://doi.org/10.1016/j.foodpol.2021.102065
African Swine Fever broke out in China in August 2018 and has caused a substantial loss to China’s hog industry. Pork is the dominant meat in the Chinese diet with its price being a critical component of China’s Consumer Price Index. In 2019, large increases in the pork price caused by the sharp reduction in pork supply incentivized the government to suppress the price by subsidizing large-scale hog farms. With an updated estimation of China’s meat demand, we argue that the ongoing policy interventions may not be the most efficient for achieving short-run reductions in the pork price. Subsidizing the production of chicken, a major substitute for pork and currently accounting for a relatively small share of meat consumption in China, could help suppress the pork price faster and at lower government costs. We estimate price dynamics and compute consumer surplus of multiple subsidy plans over a 30-month window from the third quarter of 2019 to the last quarter of 2021. Simulation outcomes suggest that allocating some subsidies from hog to chicken farms is likely to benefit consumers, producers, and the government. Our novel proposal of expanding production of a substitute meat to help lower the pork price after a large loss of the hog stock may be useful to other countries that suffer or may suffer from severe livestock losses due to animal epidemics.
Working paper
Ma, Meilin, Fei Qin, and Jayson L. Lusk. “Retailers’ Product Assortment Decisions during the Great Recession: Evidence from the U.S. Yogurt Market” Available at SSRN: https://ssrn.com/abstract=3927353
Despite extensive research on retailers’ price responses to demand shocks, much less is known about their non-price adjustments. Using heterogeneity in timing, location, and magnitude of income and wealth shocks associated with the 2008 Great Recession, we explore how U.S. retail stores adjusted product offerings in response to the shocks in local markets. Evidence shows that stores reduce product variety and change product sizes besides lowering prices. Using a structural demand model, we quantify the net welfare impact of the price and assortment adjustments. On average, the consumer welfare losses from variety reduction more than offset the welfare gains from price reductions.
My complete research statement can be found here