Juin Kuan Chong | National University of Singapore (original) (raw)

Papers by Juin Kuan Chong

1This research was supported by NSF grant SES-0078911. This draft is not intended for publication... more 1This research was supported by NSF grant SES-0078911. This draft is not intended for publication. It simply summarizes a variety of results in a format accessible to interested readers through web-based archiving. Thanks to C. M¶onica Capra, Haitao Cui, Paul Glimcher, and Roger Myerson. Former math nerd Matthew Rabin directed our attention to the golden ratio. Ming Hsu and Brian Rogers provided excellent research assistance. Useful comments were received from seminars at Caltech, Chicago, New

Behavioral & Experimental Economics eJournal, 2021

This paper examines the sunk cost fallacy as a self-commitment device in mitigating self-control ... more This paper examines the sunk cost fallacy as a self-commitment device in mitigating self-control problems and analyzes its implications for contract design. The sunk cost fallacy can lead to over-consumption and escalation of commitment. We show that consumers anticipate the fallacy ex-ante, and can strategically use it to mitigate their self-control problem. Therefore, a firm's optimal pricing contract has to balance the demand for flexibility due to the sunk cost fallacy and the demand for commitment due to the self-control problem. We find that the optimal fixed fee for investment goods (e.g., gym attendance) has a U-shape relationship with the fallacy when the consumer has self-control problems: i.e., the optimal fixed fee first decreases and then increases with the sunk cost fallacy. We compare the optimal fixed-fee contract with a pay-per-use contract which does not induce the sunk cost effect. We also investigate two commonly-observed pricing schemes: a contract menu incl...

We are especially grateful to John Lynch for helping us to develop a behavioral underpinning for ... more We are especially grateful to John Lynch for helping us to develop a behavioral underpinning for our model. David Bell, Pete Fader, and Bruce Hardie generously provided the data. We also thank the editor and 3 anonymous reviewers for their invaluable suggestions. Direct correspondence to Teck-Hua Summary We develop a highly parsimonious model to describe and predict stock-keeping unit (SKU) choice in any frequently-bought consumer packaged good category. The model posits that a product category consists of a few salient attributes and that an SKU is an attribute-level combination. The proposed model has 21 parameters for a category with 3 salient attributes, and in general has 5 + 4 ¢ (K + 1) parameters for a K-attribute product category. This model parsimony is accomplished by neither discarding part of the data nor aggregating the level of analysis beyond the level of SKU. Since the number of parameters is neither a function of the number of SKUs nor a function of the number of at...

International Series in Operations Research & Management Science, 1998

Most brand managers believe that a brand that has more products should have a higher brand share ... more Most brand managers believe that a brand that has more products should have a higher brand share because it offers more options to customers. We utilize the underlying structure of the products (within a brand) to develop three measures of brand width: the number of stock keeping units (SKUs), the number of distinct feature levels, and the number of distinct products. To examine the impact of brand width on brand share, we develop a logit model and estimate the sensitivities of brand share to our brand width measures using panel data of eight food product categories. Our logit model suggests that the brand width measures provide explanatory and predictive power. In addition, our latent-class analysis implies that different segments have different (brand choice) responses to different measures of brand width. We also use the estimated model to simulate the impact of stock-out and delayed new product introduction on brand share. Our simulation results suggest: (1) stock-out will lower the brand share in a long run and its impact is more severe when the stock-out duration lengthens; and (2) delayed new product introduction will lower the brand share initially but has minimal impact on the brand share in a long run.

The Quarterly Journal of Economics, 2004

American Economic Review, 2003

Manufacturing & Service Operations Management, 2001

T he complexity of managing a category assortment has grown tremendously in recent years due to t... more T he complexity of managing a category assortment has grown tremendously in recent years due to the increased product turnover and proliferation rates in most categories. It is an increasingly difficult task for managers to find an effective assortment due to uncertain consumer preferences and the exponential number of possible assortments. This paper presents an empirically based modeling framework for managers to assess the revenue and lost sales implication of alternative category assortments. Coupled with a local improvement heuristic, the modeling framework generates an alternative category assortment with higher revenue. This framework, which consists of a category-purchase-incidence model and a brand-share model, is calibrated and validated using 60,000 shopping trips and purchase records. Specifically, the purchase-incidence model predicts the probability of an individual customer who purchases (and who does not purchase) from a given product category during a shopping trip. The no-purchase probability enables us to estimate lost sales due to assortment changes in the category. The brand-share model predicts which brand the customer chooses if a purchase incidence occurs in the category. Our brand-share model extends the classical Guadagni and Little model (1983) by utilizing three new brand-width measures that quantify the similarities among products of different brands within the same category. We illustrate how our modeling framework is used to reconfigure the category assortment in eight food categories for five stores in our data set. This reconfiguration exercise shows that a reconfigured category assortment can have a profit improvement of up to 25.1% with 32 products replaced. We also demonstrate how our modeling framework can be used to gauge lost sales due to assortment changes. We find the level of lost sales could range from 0.9% to 10.2% for a period of 26 weeks.

Berkeley: University of California, 2004

... Rajiv Sarin and Farshid Vahid, 2003, show that “spilling” over payoff reinforcement to neighb... more ... Rajiv Sarin and Farshid Vahid, 2003, show that “spilling” over payoff reinforcement to neighboring strategies can explain the rapid pace of learning in the Van Huyck et al. (1996) game. ... rule will be found that can always beat rules which adapt locally. ...

CORE Lecture STEADY STATE …, 1991

Page 1. Learning and Equilibrium in Games Drew Fudenberg Marshall Lectures Cambridge University O... more Page 1. Learning and Equilibrium in Games Drew Fudenberg Marshall Lectures Cambridge University October 2009 Page 2. Game Theory: The study of interdependent decisions where an agent's payoff can depend on the decisions of other agents. ...

International Series in Quantitative Marketing

The number of products or stock keeping units (SKUs) in most product categories has been growing ... more The number of products or stock keeping units (SKUs) in most product categories has been growing at a phenomenal rate. Even though the number of products increases, the average sale of products can decrease. Due to cannibalization, the sales of some products may even drop below a threshold that makes them unprofitable. This has spurred some firms to remove these under-performing products from their product lines. This ad-hoc “trim the lame duck” procedure can have an adverse effect on the firm’s profit for two reasons: first, the “lame duck” may not be the most substitutable product within the line, trimming it results in higher lost sales; second, the “lame duck” may be cheaper to keep with lower inventory cost due to less variability in sales. As an initial step to developing a better product elimination procedure, we use a model that explicitly captures product substitution phenomenon to examine various product portfolios. We compare the mean and the variance of the sales associated with two basic strategies: trimming and no trimming. Our results provide insight into when and which products could be trimmed.

Functional experience weighted attraction (fEWA) is a one-parameter theory of learning in games. ... more Functional experience weighted attraction (fEWA) is a one-parameter theory of learning in games. It approximates the free parameters in an earlier model (EWA) with functions of experience. The theory was originally tested on seven di®erent games and compared to four other learning and equilibrium theories, then four more games were added. Generally fEWA or parameterized EWA predict best out-of-sample, but one kind of reinforcement learning predicts well in games with mixed-strategy equilibrium. Of the learning models, belief learning models ¯t worst but ¯t better than noisy (quantal response) equilibrium models. The economic value of a theory is measured by how much more subjects would have earned if they followed the theory's recommendations. Most learning theories add value (though equilibrium theories often subtract value) and fEWA and EWA usually add the most value. \In nature hybrid species are usually sterile, but in science the reverse is often true"{ Francis Crick ...

... 3 Chong, Juin-Kuan, Teck-Hua Ho and Chris Tang (2001), “A Modeling Framework for Category Ass... more ... 3 Chong, Juin-Kuan, Teck-Hua Ho and Chris Tang (2001), “A Modeling Framework for Category Assortment Planning,” Manufacturing & Service Operations Management, 3(3), pp. 191-210. Fader, Peter S. and Bruce GS Hardie ...

Management Science

We study taxi markets in Singapore to understand market entry in the field. Although market-entry... more We study taxi markets in Singapore to understand market entry in the field. Although market-entry games in the laboratory consistently produce equilibrium outcomes, we show that a lack of market knowledge hinders the markets from consistently reaching equilibrium in the field. In Singapore, a small, 720-square-kilometre island city that can be divided into 29 taxi markets, full equilibrium is elusive: 68% of the market-entry decisions made by the 2,728 taxi drivers in our data could be improved. Using three months of earnings and detailed movement data from these taxi drivers, we find an average 20% gap in marginal wage rates across markets. We use dynamic programming to derive the optimal solution for more than 3 million search decisions and find that only 32% of the searches ended in an optimal market. Finally, we find that market knowledge developed in a given month explains an additional 3% variation of the earning losses in the 2.6 million decisions for the subsequent two month...

BMJ

Objective To investigate the association between popular football games played in Europe and the ... more Objective To investigate the association between popular football games played in Europe and the incidence of traffic accidents in Asia. Design Study based on 41 538 traffic accidents involving taxis in Singapore and 1 814 320 traffic accidents in Taiwan, combined with 12 788 European club football games over a seven year period. Setting Singapore and Taiwan. Participants The largest taxi company in Singapore, with fine grained traffic accident records in a three year span; all traffic accident records in Taiwan in a six year span. Exposure Days when high profile football games were played or not played. Main outcome measure Number of traffic accidents. Results Regression based and time series models suggest that days with high profile European football matches were more positively associated with traffic accidents than days with less popular European football matches. For an approximate €134.74m (£120.25m; $159.76m) increase in average market value for matches played on a given day...

Management Science

There has been a considerable amount of interest in the empirical investigation of social influen... more There has been a considerable amount of interest in the empirical investigation of social influence in the marketing and economics literature in the last decade or so. Among the many different empirical models applied for such investigations, the most common class of model is the linear-in-means model. These models can be used to examine whether social influence is truly a result of agents affecting each other through their choices simultaneously (endogenous effect) or of having similar taste and characteristics (homophily). However, the two effects are not separately identified in general in the standard linear-in-means model unless data on all members of an individual’s network are available. With data on a sample of individuals from a network, these effects are not identified. In this research, we leverage a very natural aspect of social settings, namely that consumers are usually part of multiple—as opposed to single—networks. We discuss the sufficient conditions for identificat...

Journal of Statistical Computation and Simulation

Proceedings of the National Academy of Sciences of the United States of America, Jan 21, 2017

Is there a link between the color of a taxi and how many accidents it has? An analysis of 36 mo o... more Is there a link between the color of a taxi and how many accidents it has? An analysis of 36 mo of detailed taxi, driver, and accident data (comprising millions of data points) from the largest taxi company in Singapore suggests that there is an explicit link. Yellow taxis had 6.1 fewer accidents per 1,000 taxis per month than blue taxis, a 9% reduction in accident probability. We rule out driver difference as an explanatory variable and empirically show that because yellow taxis are more noticeable than blue taxis-especially when in front of another vehicle, and in street lighting-other drivers can better avoid hitting them, directly reducing the accident rate. This finding can play a significant role when choosing colors for public transportation and may save lives as well as millions of dollars.

Games and Economic Behavior, 2016

Subjects in simple games often exhibit nonequilibrium behaviors. Cognitive Hierarchy (CH) and Lev... more Subjects in simple games often exhibit nonequilibrium behaviors. Cognitive Hierarchy (CH) and Level k (LK) are two prevailing structural models that can predict these behaviors well but they make quite different assumptions on players' beliefs of their opponents' actions. This paper develops a generalization of CH and shows that CH and a variant of LK belong to the same family. Under generalized CH (GCH), level k players best respond to level 0 to level k −1 but the perceived proportion of each lower level is obtained by weighting its actual frequency by a parameter α, reflecting stereotype bias well documented in social psychology literature. When α = 1, GCH reduces to CH; and when α = ∞, it becomes Level m (LM) in which level k best responds to only the modal level below k (and the modal level may be k−1). GCH also fixes prior ad-hoc assumptions about level 0 by developing a plausible model for it. GCH posits that non-strategic level 0 players are more likely to choose strategies that will never yield the minimum payoff in all possible outcome scenarios. This minimum-aversion tendency captures level 0's avoidance for dominated strategies and compromise effects welldocumented in the individual choice literature. Using fifty-five 2-player m x n games from four distinct datasets, we show that GCH describes and predicts behaviors better than CH and LK. Structural estimation results show that higher level players exhibit stereotype bias and level 0 players exhibit minimum aversion. Finally, we apply GCH to two new games and find that GCH is able to overcome CH's inadequacies and predict behavior remarkably well.

1This research was supported by NSF grant SES-0078911. This draft is not intended for publication... more 1This research was supported by NSF grant SES-0078911. This draft is not intended for publication. It simply summarizes a variety of results in a format accessible to interested readers through web-based archiving. Thanks to C. M¶onica Capra, Haitao Cui, Paul Glimcher, and Roger Myerson. Former math nerd Matthew Rabin directed our attention to the golden ratio. Ming Hsu and Brian Rogers provided excellent research assistance. Useful comments were received from seminars at Caltech, Chicago, New

Behavioral & Experimental Economics eJournal, 2021

This paper examines the sunk cost fallacy as a self-commitment device in mitigating self-control ... more This paper examines the sunk cost fallacy as a self-commitment device in mitigating self-control problems and analyzes its implications for contract design. The sunk cost fallacy can lead to over-consumption and escalation of commitment. We show that consumers anticipate the fallacy ex-ante, and can strategically use it to mitigate their self-control problem. Therefore, a firm's optimal pricing contract has to balance the demand for flexibility due to the sunk cost fallacy and the demand for commitment due to the self-control problem. We find that the optimal fixed fee for investment goods (e.g., gym attendance) has a U-shape relationship with the fallacy when the consumer has self-control problems: i.e., the optimal fixed fee first decreases and then increases with the sunk cost fallacy. We compare the optimal fixed-fee contract with a pay-per-use contract which does not induce the sunk cost effect. We also investigate two commonly-observed pricing schemes: a contract menu incl...

We are especially grateful to John Lynch for helping us to develop a behavioral underpinning for ... more We are especially grateful to John Lynch for helping us to develop a behavioral underpinning for our model. David Bell, Pete Fader, and Bruce Hardie generously provided the data. We also thank the editor and 3 anonymous reviewers for their invaluable suggestions. Direct correspondence to Teck-Hua Summary We develop a highly parsimonious model to describe and predict stock-keeping unit (SKU) choice in any frequently-bought consumer packaged good category. The model posits that a product category consists of a few salient attributes and that an SKU is an attribute-level combination. The proposed model has 21 parameters for a category with 3 salient attributes, and in general has 5 + 4 ¢ (K + 1) parameters for a K-attribute product category. This model parsimony is accomplished by neither discarding part of the data nor aggregating the level of analysis beyond the level of SKU. Since the number of parameters is neither a function of the number of SKUs nor a function of the number of at...

International Series in Operations Research & Management Science, 1998

Most brand managers believe that a brand that has more products should have a higher brand share ... more Most brand managers believe that a brand that has more products should have a higher brand share because it offers more options to customers. We utilize the underlying structure of the products (within a brand) to develop three measures of brand width: the number of stock keeping units (SKUs), the number of distinct feature levels, and the number of distinct products. To examine the impact of brand width on brand share, we develop a logit model and estimate the sensitivities of brand share to our brand width measures using panel data of eight food product categories. Our logit model suggests that the brand width measures provide explanatory and predictive power. In addition, our latent-class analysis implies that different segments have different (brand choice) responses to different measures of brand width. We also use the estimated model to simulate the impact of stock-out and delayed new product introduction on brand share. Our simulation results suggest: (1) stock-out will lower the brand share in a long run and its impact is more severe when the stock-out duration lengthens; and (2) delayed new product introduction will lower the brand share initially but has minimal impact on the brand share in a long run.

The Quarterly Journal of Economics, 2004

American Economic Review, 2003

Manufacturing & Service Operations Management, 2001

T he complexity of managing a category assortment has grown tremendously in recent years due to t... more T he complexity of managing a category assortment has grown tremendously in recent years due to the increased product turnover and proliferation rates in most categories. It is an increasingly difficult task for managers to find an effective assortment due to uncertain consumer preferences and the exponential number of possible assortments. This paper presents an empirically based modeling framework for managers to assess the revenue and lost sales implication of alternative category assortments. Coupled with a local improvement heuristic, the modeling framework generates an alternative category assortment with higher revenue. This framework, which consists of a category-purchase-incidence model and a brand-share model, is calibrated and validated using 60,000 shopping trips and purchase records. Specifically, the purchase-incidence model predicts the probability of an individual customer who purchases (and who does not purchase) from a given product category during a shopping trip. The no-purchase probability enables us to estimate lost sales due to assortment changes in the category. The brand-share model predicts which brand the customer chooses if a purchase incidence occurs in the category. Our brand-share model extends the classical Guadagni and Little model (1983) by utilizing three new brand-width measures that quantify the similarities among products of different brands within the same category. We illustrate how our modeling framework is used to reconfigure the category assortment in eight food categories for five stores in our data set. This reconfiguration exercise shows that a reconfigured category assortment can have a profit improvement of up to 25.1% with 32 products replaced. We also demonstrate how our modeling framework can be used to gauge lost sales due to assortment changes. We find the level of lost sales could range from 0.9% to 10.2% for a period of 26 weeks.

Berkeley: University of California, 2004

... Rajiv Sarin and Farshid Vahid, 2003, show that “spilling” over payoff reinforcement to neighb... more ... Rajiv Sarin and Farshid Vahid, 2003, show that “spilling” over payoff reinforcement to neighboring strategies can explain the rapid pace of learning in the Van Huyck et al. (1996) game. ... rule will be found that can always beat rules which adapt locally. ...

CORE Lecture STEADY STATE …, 1991

Page 1. Learning and Equilibrium in Games Drew Fudenberg Marshall Lectures Cambridge University O... more Page 1. Learning and Equilibrium in Games Drew Fudenberg Marshall Lectures Cambridge University October 2009 Page 2. Game Theory: The study of interdependent decisions where an agent's payoff can depend on the decisions of other agents. ...

International Series in Quantitative Marketing

The number of products or stock keeping units (SKUs) in most product categories has been growing ... more The number of products or stock keeping units (SKUs) in most product categories has been growing at a phenomenal rate. Even though the number of products increases, the average sale of products can decrease. Due to cannibalization, the sales of some products may even drop below a threshold that makes them unprofitable. This has spurred some firms to remove these under-performing products from their product lines. This ad-hoc “trim the lame duck” procedure can have an adverse effect on the firm’s profit for two reasons: first, the “lame duck” may not be the most substitutable product within the line, trimming it results in higher lost sales; second, the “lame duck” may be cheaper to keep with lower inventory cost due to less variability in sales. As an initial step to developing a better product elimination procedure, we use a model that explicitly captures product substitution phenomenon to examine various product portfolios. We compare the mean and the variance of the sales associated with two basic strategies: trimming and no trimming. Our results provide insight into when and which products could be trimmed.

Functional experience weighted attraction (fEWA) is a one-parameter theory of learning in games. ... more Functional experience weighted attraction (fEWA) is a one-parameter theory of learning in games. It approximates the free parameters in an earlier model (EWA) with functions of experience. The theory was originally tested on seven di®erent games and compared to four other learning and equilibrium theories, then four more games were added. Generally fEWA or parameterized EWA predict best out-of-sample, but one kind of reinforcement learning predicts well in games with mixed-strategy equilibrium. Of the learning models, belief learning models ¯t worst but ¯t better than noisy (quantal response) equilibrium models. The economic value of a theory is measured by how much more subjects would have earned if they followed the theory's recommendations. Most learning theories add value (though equilibrium theories often subtract value) and fEWA and EWA usually add the most value. \In nature hybrid species are usually sterile, but in science the reverse is often true"{ Francis Crick ...

... 3 Chong, Juin-Kuan, Teck-Hua Ho and Chris Tang (2001), “A Modeling Framework for Category Ass... more ... 3 Chong, Juin-Kuan, Teck-Hua Ho and Chris Tang (2001), “A Modeling Framework for Category Assortment Planning,” Manufacturing & Service Operations Management, 3(3), pp. 191-210. Fader, Peter S. and Bruce GS Hardie ...

Management Science

We study taxi markets in Singapore to understand market entry in the field. Although market-entry... more We study taxi markets in Singapore to understand market entry in the field. Although market-entry games in the laboratory consistently produce equilibrium outcomes, we show that a lack of market knowledge hinders the markets from consistently reaching equilibrium in the field. In Singapore, a small, 720-square-kilometre island city that can be divided into 29 taxi markets, full equilibrium is elusive: 68% of the market-entry decisions made by the 2,728 taxi drivers in our data could be improved. Using three months of earnings and detailed movement data from these taxi drivers, we find an average 20% gap in marginal wage rates across markets. We use dynamic programming to derive the optimal solution for more than 3 million search decisions and find that only 32% of the searches ended in an optimal market. Finally, we find that market knowledge developed in a given month explains an additional 3% variation of the earning losses in the 2.6 million decisions for the subsequent two month...

BMJ

Objective To investigate the association between popular football games played in Europe and the ... more Objective To investigate the association between popular football games played in Europe and the incidence of traffic accidents in Asia. Design Study based on 41 538 traffic accidents involving taxis in Singapore and 1 814 320 traffic accidents in Taiwan, combined with 12 788 European club football games over a seven year period. Setting Singapore and Taiwan. Participants The largest taxi company in Singapore, with fine grained traffic accident records in a three year span; all traffic accident records in Taiwan in a six year span. Exposure Days when high profile football games were played or not played. Main outcome measure Number of traffic accidents. Results Regression based and time series models suggest that days with high profile European football matches were more positively associated with traffic accidents than days with less popular European football matches. For an approximate €134.74m (£120.25m; $159.76m) increase in average market value for matches played on a given day...

Management Science

There has been a considerable amount of interest in the empirical investigation of social influen... more There has been a considerable amount of interest in the empirical investigation of social influence in the marketing and economics literature in the last decade or so. Among the many different empirical models applied for such investigations, the most common class of model is the linear-in-means model. These models can be used to examine whether social influence is truly a result of agents affecting each other through their choices simultaneously (endogenous effect) or of having similar taste and characteristics (homophily). However, the two effects are not separately identified in general in the standard linear-in-means model unless data on all members of an individual’s network are available. With data on a sample of individuals from a network, these effects are not identified. In this research, we leverage a very natural aspect of social settings, namely that consumers are usually part of multiple—as opposed to single—networks. We discuss the sufficient conditions for identificat...

Journal of Statistical Computation and Simulation

Proceedings of the National Academy of Sciences of the United States of America, Jan 21, 2017

Is there a link between the color of a taxi and how many accidents it has? An analysis of 36 mo o... more Is there a link between the color of a taxi and how many accidents it has? An analysis of 36 mo of detailed taxi, driver, and accident data (comprising millions of data points) from the largest taxi company in Singapore suggests that there is an explicit link. Yellow taxis had 6.1 fewer accidents per 1,000 taxis per month than blue taxis, a 9% reduction in accident probability. We rule out driver difference as an explanatory variable and empirically show that because yellow taxis are more noticeable than blue taxis-especially when in front of another vehicle, and in street lighting-other drivers can better avoid hitting them, directly reducing the accident rate. This finding can play a significant role when choosing colors for public transportation and may save lives as well as millions of dollars.

Games and Economic Behavior, 2016

Subjects in simple games often exhibit nonequilibrium behaviors. Cognitive Hierarchy (CH) and Lev... more Subjects in simple games often exhibit nonequilibrium behaviors. Cognitive Hierarchy (CH) and Level k (LK) are two prevailing structural models that can predict these behaviors well but they make quite different assumptions on players' beliefs of their opponents' actions. This paper develops a generalization of CH and shows that CH and a variant of LK belong to the same family. Under generalized CH (GCH), level k players best respond to level 0 to level k −1 but the perceived proportion of each lower level is obtained by weighting its actual frequency by a parameter α, reflecting stereotype bias well documented in social psychology literature. When α = 1, GCH reduces to CH; and when α = ∞, it becomes Level m (LM) in which level k best responds to only the modal level below k (and the modal level may be k−1). GCH also fixes prior ad-hoc assumptions about level 0 by developing a plausible model for it. GCH posits that non-strategic level 0 players are more likely to choose strategies that will never yield the minimum payoff in all possible outcome scenarios. This minimum-aversion tendency captures level 0's avoidance for dominated strategies and compromise effects welldocumented in the individual choice literature. Using fifty-five 2-player m x n games from four distinct datasets, we show that GCH describes and predicts behaviors better than CH and LK. Structural estimation results show that higher level players exhibit stereotype bias and level 0 players exhibit minimum aversion. Finally, we apply GCH to two new games and find that GCH is able to overcome CH's inadequacies and predict behavior remarkably well.