Songfa Zhong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Songfa Zhong has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Soo Hong Chew
National University of Singapore
5 shared publications - Junjian Yi
National University of Singapore (NUS) - Department of Economics
2 shared publications Junsen Zhang
Department of Economics, Chinese University of Hong Kong, Hong Kong SAR, People’s Republic of China
2 shared publicationsRobin Chark
Marketing, National University of Singapore
2 shared publicationsSoo Hong Chew
Department of Economics, National University of Singapore, Singapore 119077
2 shared publicationsBin Miao
School of Economics, Renmin University of China
1 shared publication
Latest Publications
- 1 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (5) , 4231-4246
Individualism-Collectivism and Risk Perception Around the World
Understanding cultural differences in risk perception is critical in an increasingly uncertain world. Here we examine the relationship between the individualism-collectivism continuum and risk perception around the world using a data set from the Lloyd’s Register Foundation World Risk Poll. The data set contains rich information of a representative sample of 150,000 participants from 142 countries, and investigates risk perception in terms of perceived likelihood and personal experiences for a range of risks in daily life. We observe that participants from countries with a more individualistic culture perceive lower risk after controlling their personal experiences. We observe similar but weaker patterns when we adopt an epidemiological approach to investigate the individualistic cultural influence of first- and second-generation immigrants and use historical kinship tightness to proxy for individualism. Our study sheds light on the importance of culture in shaping risk perception and contributes to understanding global differences in behavioral traits. This paper was accepted by Ilia Tsetlin, behavioral economics and decision analysis. Funding: This research was supported by the Social Science Research Council (Singapore, grant number MOE2022-SSRTG-029), and National Natural Science Foundation of China (grant number 72425005). Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.01612 .
- 1 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (5) , 4231-4246
Individualism-Collectivism and Risk Perception Around the World
Understanding cultural differences in risk perception is critical in an increasingly uncertain world. Here we examine the relationship between the individualism-collectivism continuum and risk perception around the world using a data set from the Lloyd’s Register Foundation World Risk Poll. The data set contains rich information of a representative sample of 150,000 participants from 142 countries, and investigates risk perception in terms of perceived likelihood and personal experiences for a range of risks in daily life. We observe that participants from countries with a more individualistic culture perceive lower risk after controlling their personal experiences. We observe similar but weaker patterns when we adopt an epidemiological approach to investigate the individualistic cultural influence of first- and second-generation immigrants and use historical kinship tightness to proxy for individualism. Our study sheds light on the importance of culture in shaping risk perception and contributes to understanding global differences in behavioral traits. This paper was accepted by Ilia Tsetlin, behavioral economics and decision analysis. Funding: This research was supported by the Social Science Research Council (Singapore, grant number MOE2022-SSRTG-029), and National Natural Science Foundation of China (grant number 72425005). Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.01612 .
- Top Cited PapersOpen Access
- 12 December 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (51)
The emergence of economic rationality of GPT
As large language models (LLMs) like GPT become increasingly prevalent, it is essential that we assess their capabilities beyond language processing. This paper examines the economic rationality of GPT by instructing it to make budgetary decisions in four domains: risk, time, social, and food preferences. We measure economic rationality by assessing the consistency of GPT’s decisions with utility maximization in classic revealed preference theory. We find that GPT’s decisions are largely rational in each domain and demonstrate higher rationality score than those of human subjects in a parallel experiment and in the literature. Moreover, the estimated preference parameters of GPT are slightly different from human subjects and exhibit a lower degree of heterogeneity. We also find that the rationality scores are robust to the degree of randomness and demographic settings such as age and gender but are sensitive to contexts based on the language frames of the choice situations. These results suggest the potential of LLMs to make good decisions and the need to further understand their capabilities, limitations, and underlying mechanisms.
- 12 December 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (51)
Top Cited PapersOpen AccessThe emergence of economic rationality of GPT
As large language models (LLMs) like GPT become increasingly prevalent, it is essential that we assess their capabilities beyond language processing. This paper examines the economic rationality of GPT by instructing it to make budgetary decisions in four domains: risk, time, social, and food preferences. We measure economic rationality by assessing the consistency of GPT’s decisions with utility maximization in classic revealed preference theory. We find that GPT’s decisions are largely rational in each domain and demonstrate higher rationality score than those of human subjects in a parallel experiment and in the literature. Moreover, the estimated preference parameters of GPT are slightly different from human subjects and exhibit a lower degree of heterogeneity. We also find that the rationality scores are robust to the degree of randomness and demographic settings such as age and gender but are sensitive to contexts based on the language frames of the choice situations. These results suggest the potential of LLMs to make good decisions and the need to further understand their capabilities, limitations, and underlying mechanisms.
- Open Access
The Emergence of Economic Rationality of GPT
As large language models (LLMs) like GPT become increasingly prevalent, it is essential that we assess their capabilities beyond language processing. This paper examines the economic rationality of GPT by instructing it to make budgetary decisions in four domains: risk, time, social, and food preferences. We measure economic rationality by assessing the consistency of GPT's decisions with utility maximization in classic revealed preference theory. We find that GPT's decisions are largely rational in each domain and demonstrate higher rationality score than those of human subjects in a parallel experiment and in the literature. Moreover, the estimated preference parameters of GPT are slightly different from human subjects and exhibit a lower degree of heterogeneity. We also find that the rationality scores are robust to the degree of randomness and demographic settings such as age and gender, but are sensitive to contexts based on the language frames of the choice situations. These results suggest the potential of LLMs to make good decisions and the need to further understand their capabilities, limitations, and underlying mechanisms.
Open AccessThe Emergence of Economic Rationality of GPT
As large language models (LLMs) like GPT become increasingly prevalent, it is essential that we assess their capabilities beyond language processing. This paper examines the economic rationality of GPT by instructing it to make budgetary decisions in four domains: risk, time, social, and food preferences. We measure economic rationality by assessing the consistency of GPT's decisions with utility maximization in classic revealed preference theory. We find that GPT's decisions are largely rational in each domain and demonstrate higher rationality score than those of human subjects in a parallel experiment and in the literature. Moreover, the estimated preference parameters of GPT are slightly different from human subjects and exhibit a lower degree of heterogeneity. We also find that the rationality scores are robust to the degree of randomness and demographic settings such as age and gender, but are sensitive to contexts based on the language frames of the choice situations. These results suggest the potential of LLMs to make good decisions and the need to further understand their capabilities, limitations, and underlying mechanisms.
- Open Access
- 3 March 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (10)
A multifaceted poverty reduction program has economic and behavioral consequences
This paper examines the causal impact of poverty reduction interventions on the social preferences of the poor. A multifaceted poverty reduction program in China provides a setting for the use of a fuzzy regression discontinuity design. The design compares households with base-year income just below a preset criterion, who were more likely to receive the program treatment, with households just above the criterion. Five years after the program’s launch, we conducted a lab-in-the-field experiment to measure the distributional preferences of household heads. Combining quasi-random variation from program rules with administrative census and experimental data, we find both economic and behavioral consequences of the program: It increased household income by 50% 5 y later, increased consistency with utility maximization by household heads, and increased their efficiency preference while reducing selfishness and leaving equality preference unchanged. Our findings advance scientific understanding of social preferences formation and highlight a broad perspective in evaluating poverty reduction interventions.
- 3 March 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (10)
Open AccessA multifaceted poverty reduction program has economic and behavioral consequences
This paper examines the causal impact of poverty reduction interventions on the social preferences of the poor. A multifaceted poverty reduction program in China provides a setting for the use of a fuzzy regression discontinuity design. The design compares households with base-year income just below a preset criterion, who were more likely to receive the program treatment, with households just above the criterion. Five years after the program’s launch, we conducted a lab-in-the-field experiment to measure the distributional preferences of household heads. Combining quasi-random variation from program rules with administrative census and experimental data, we find both economic and behavioral consequences of the program: It increased household income by 50% 5 y later, increased consistency with utility maximization by household heads, and increased their efficiency preference while reducing selfishness and leaving equality preference unchanged. Our findings advance scientific understanding of social preferences formation and highlight a broad perspective in evaluating poverty reduction interventions.
- Open Access
- 1 January 2023
- Published by The Econometric Society in Quantitative Economics
- Vol. 14 (3) , 1133-1162
Ellsberg meets Keynes at an urn
- 1 January 2023
- Published by The Econometric Society in Quantitative Economics
- Vol. 14 (3) , 1133-1162
Open AccessEllsberg meets Keynes at an urn
- 1 January 2023
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 69 (1) , 475-490
Reference Dependence in Intertemporal Preference
Dynamic structures in intertemporal choices offer multiple channels through which reference dependence can occur. We examine intertemporal reference dependence when using choice lists to elicit time preference over outcome sequences. Reference point effects can arise endogenously within an outcome sequence and exogenously from the experimental environment of the choice list. We show that estimated discount factors in the choice list are biased by both kinds of reference points and propose a model to jointly account for both. Our experimental design with outcome sequences also enables us to jointly estimate the discount factor and utility curvature. We further discuss the implications for recent developments in the measurement of intertemporal preference. This paper was accepted by Manel Baucells, behavioral economics and decision analysis. Funding: S. Zhong was supported by the National University of Singapore [Singapore Ministry of Education Academic Research Fund Tier 1 Grant R-122-000-325-115]. Supplemental Material: The data files and online appendices are available at https://doi.org/10.1287/mnsc.2022.4348 .
- 1 January 2023
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 69 (1) , 475-490
Reference Dependence in Intertemporal Preference
Dynamic structures in intertemporal choices offer multiple channels through which reference dependence can occur. We examine intertemporal reference dependence when using choice lists to elicit time preference over outcome sequences. Reference point effects can arise endogenously within an outcome sequence and exogenously from the experimental environment of the choice list. We show that estimated discount factors in the choice list are biased by both kinds of reference points and propose a model to jointly account for both. Our experimental design with outcome sequences also enables us to jointly estimate the discount factor and utility curvature. We further discuss the implications for recent developments in the measurement of intertemporal preference. This paper was accepted by Manel Baucells, behavioral economics and decision analysis. Funding: S. Zhong was supported by the National University of Singapore [Singapore Ministry of Education Academic Research Fund Tier 1 Grant R-122-000-325-115]. Supplemental Material: The data files and online appendices are available at https://doi.org/10.1287/mnsc.2022.4348 .
- 1 April 2020
- Published by Oxford University Press (OUP) in Journal of the European Economic Association
- Vol. 18 (2) , 1009-1039
INDIVIDUAL PREFERENCE FOR LONGSHOTS
Results from studies on risk taking behavior suggest that people tend to be risk seeking when making choices over lotteries that involve longshots: small probabilities of winning sizable payoffs. To investigate preferences over longshots systematically, we conduct an incentivized experiment using state lotteries in China, each involving a single prize and fixed winning odds. This enables our construction of single-prize lotteries involving winning odds between 10-5 and 10-1 and winning prizes ranging from RMB10 (about USD1.60) to RMB10,000,000 (about USD1.60 million) across different expected payoffs. For lotteries with expected payoffs of 1 and 10, subjects exhibit heterogeneous preferences for longshots: some prefer the smallest winning probability whereas others favor intermediate winning probabilities. As the expected payoff increases to 100, subjects become predominantly risk averse, even for the lowest winning probability of 10-5. Our findings pose challenges for utility models of decision making under risk.
- 1 April 2020
- Published by Oxford University Press (OUP) in Journal of the European Economic Association
- Vol. 18 (2) , 1009-1039
INDIVIDUAL PREFERENCE FOR LONGSHOTS
Results from studies on risk taking behavior suggest that people tend to be risk seeking when making choices over lotteries that involve longshots: small probabilities of winning sizable payoffs. To investigate preferences over longshots systematically, we conduct an incentivized experiment using state lotteries in China, each involving a single prize and fixed winning odds. This enables our construction of single-prize lotteries involving winning odds between 10-5 and 10-1 and winning prizes ranging from RMB10 (about USD1.60) to RMB10,000,000 (about USD1.60 million) across different expected payoffs. For lotteries with expected payoffs of 1 and 10, subjects exhibit heterogeneous preferences for longshots: some prefer the smallest winning probability whereas others favor intermediate winning probabilities. As the expected payoff increases to 100, subjects become predominantly risk averse, even for the lowest winning probability of 10-5. Our findings pose challenges for utility models of decision making under risk.
- 28 September 2019
- Published by American Economic Association
Evaluating Longshot Preference in the Field
- 28 September 2019
- Published by American Economic Association
Evaluating Longshot Preference in the Field
- 1 September 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (8) , 3489-3500
Recover Overnight? Work Interruption and Worker Productivity
This paper investigates the effect of work interruption on workers’ subsequent productivity. We employ a data set of individual productivity and machine conditions, in which each worker faces the chance, on a daily basis, that her machine will break down randomly. Our analysis finds that compared to a workday with smooth production, experiencing a machine breakdown is associated with a 3.3 percentage point decline in the worker’s productivity the following day. We discuss possible explanations for the observed effect, including negative emotions, increased cautiousness in operating the machine, and proficiency loss. Our findings shed light on the importance of understanding and managing interruptions in the workplace, and contribute to a growing literature on the determinants of productivity at the micro level. Data and the online appendix are available at https://doi.org/10.1287/mnsc.2017.2792 . This paper was accepted by John List, behavioral economics.
- 1 September 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (8) , 3489-3500
Recover Overnight? Work Interruption and Worker Productivity
This paper investigates the effect of work interruption on workers’ subsequent productivity. We employ a data set of individual productivity and machine conditions, in which each worker faces the chance, on a daily basis, that her machine will break down randomly. Our analysis finds that compared to a workday with smooth production, experiencing a machine breakdown is associated with a 3.3 percentage point decline in the worker’s productivity the following day. We discuss possible explanations for the observed effect, including negative emotions, increased cautiousness in operating the machine, and proficiency loss. Our findings shed light on the importance of understanding and managing interruptions in the workplace, and contribute to a growing literature on the determinants of productivity at the micro level. Data and the online appendix are available at https://doi.org/10.1287/mnsc.2017.2792 . This paper was accepted by John List, behavioral economics.
- 1 September 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (8) , 3896-3910
Risk Aversion and Son Preference: Experimental Evidence from Chinese Twin Parents
We study the role of risk aversion underlying son preference in patriarchal societies, where sons serve as better insurance for old-age support than daughters. The implications of an insurance motive on son preference are twofold. First, prior to the birth of their children, more-risk-averse parents have a stronger preference for sons than for daughters. Second, after the birth of their children, parents with sons are more risk seeking, compared to parents with daughters. We adopt a within-twin-pair fixed-effects estimator with a weak identification assumption, which enables us to jointly identify these two effects. We further conduct an incentivized choice experiment to assess parental risk attitude in a sample of Chinese twins with children, and follow up with a second twin sample to examine the replicability of the findings. In both samples, we find that parents with higher risk aversion before the birth of their children are more likely to have sons through sex selection than parents with lower risk aversion. Additionally, having sons significantly decreases parental risk aversion. These results contribute to the literature on the sources of son preference and help shed light on the nature of gender inequality. Data are available at https://doi.org/10.1287/mnsc.2017.2779 . This paper was accepted by John List, behavioral economics.
- 1 September 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (8) , 3896-3910
Risk Aversion and Son Preference: Experimental Evidence from Chinese Twin Parents
We study the role of risk aversion underlying son preference in patriarchal societies, where sons serve as better insurance for old-age support than daughters. The implications of an insurance motive on son preference are twofold. First, prior to the birth of their children, more-risk-averse parents have a stronger preference for sons than for daughters. Second, after the birth of their children, parents with sons are more risk seeking, compared to parents with daughters. We adopt a within-twin-pair fixed-effects estimator with a weak identification assumption, which enables us to jointly identify these two effects. We further conduct an incentivized choice experiment to assess parental risk attitude in a sample of Chinese twins with children, and follow up with a second twin sample to examine the replicability of the findings. In both samples, we find that parents with higher risk aversion before the birth of their children are more likely to have sons through sex selection than parents with lower risk aversion. Additionally, having sons significantly decreases parental risk aversion. These results contribute to the literature on the sources of son preference and help shed light on the nature of gender inequality. Data are available at https://doi.org/10.1287/mnsc.2017.2779 . This paper was accepted by John List, behavioral economics.