2018 Shanghai Workshop
of Behavioral and Experimental Economics
School of Economics, Shanghai University of Finance and Economics
(Room 511, the Economics Building, 111 Wuchuan Rd., Yangpu District, Shanghai 200433)
June 14th, 2018
Greetings from the Dean
Session 1 Chair: Ninghua Du
James C. Cox, Georgia State University
Title: Rational Choice and Moral Monotonicity
Abstract: In this address I will explain why published data, from some dictator games with giving and taking opportunities, and from some payoff-equivalent public good and common pool games, are inconsistent with conventional rational choice theory. The data motivate development of moral monotonicity theory, an extension of rational choice theory that incorporates moral reference points. I will describe new experiments with dictator games, and with payoff-equivalent public good and common pool games, that provide data for which conventional rational choice theory is rejected in favor of moral monotonicity theory. I will also explain that the new theory can rationalize data from many experiments reported in previous literature including many types of dictator games and strategic games with contractions of feasible sets.
Bin Miao, Shanghai University of Finance and Economics
Title: Multiple-Switching Behavior in the Elicitation of Risk Preferences
Abstract: In the setting of the price-list elicitation of certainty equivalents, in which the lottery option is fixed while the sure options vary, we investigate the pervasive phenomenon of subjects switching between the fixed lottery option the range of sure options multiple times. Relating to Agranov and Ortoleva’s (2016) finding of subjects deliberately switching between two lotteries when they are told in advance that the same two lotteries will be presented three times in a row, this multiple-switching behavior (MSB) enables a discrimination in favor of models of deliberate randomization based on quasiconcavity in probabilities (Machina, 1985) rather than negative certainty independence (Dillenberger, 2010). In addition, we find an intriguing phenomenon of reference dependence in MSB – the frequency of MSB for mixed lotteries is higher than those for gain and loss oriented lotteries. We further show that this observed hump pattern in MSB is compatible with rank-dependent utility with a concave probability weighting function and a piecewise linear loss averse utility function.
Group Photo and Tea Break
Hang Ye, Zhejiang University & Zhejiang University of Finance and Economics
Title: Logical Demonstration and Empirical Demonstration – Behavioral Economics and (Generalized) Experimental Methods
Abstract: The current report discusses the generalized experimental methods of behavioral economics from the perspective of methodology, including behavioral experiments (Experimental economics), neural experiments (Neuroeconomics) and simulation experiments (Computational economics), which play an important role in behavioral economics studies. To be specific, a series of case studies concerning the solution to the social dilemma will be demonstrated in this report, revealing how the three kinds of experimental methods can provide a powerful technical analysis support for behavioral economics research.
Tingting Ding, Shanghai University of Finance and Economics Title: Bargaining and News: An Experimental Study
Abstract: Agreement may not be reached when asymmetric information exists between buyers and sellers. We conduct a series of laboratory experiments to examine behavior of subjects and efficiency in bargaining when a “news” process is introduced to mitigate the information gap. Our experiments implement the Daley and Green (2018) bargaining model in which an uninformed buyer makes frequent offers to an informed seller while gradually learning the seller’s type from an observable news process. Our results are consistent with theory in several respects: (i) The buyer’s ability to extract more surplus by leveraging the information is limited; (ii) Sellers with high values and high costs benefit from the buyer’s ability to learn; and (iii) The quality of news has a mixed effect on efficiency. However, the bargaining duration in the lab is much longer than theory prediction because sellers demand “fair” prices.
Session 2 Chair: Bin Miao
Erkut Ozbay, University of Maryland
Title: Unavailable Options and Irrelevant Attributes
Abstract: This paper experimentally investigates the effect of introducing unavailable alternatives and irrelevant information regarding the alternatives on the optimality of decisions in choice problems. We find that interaction between the unavailable alternatives and irrelevant information regarding the alternatives generates suboptimal decisions. Irrelevant information in any dimension increases the time costs of decisions. We also identify a pure “preference for simplicity” beyond the desire to make optimal decisions or minimize time spent on a decision problem. Our results imply that the presentation set, distinct from the alternative set, needs to be a part of decision making models.
Lan Yao, Shanghai University of Finance and EconomicsTitle: Public Discourse and Pro-Social Market Behavior
Abstract: Based on the experiments of Bartling, Weber, and Yao (2015), in which buyers and sellers trade products with a negative externality on helpless third party, we investigate formation of social norm and its roles in ethical market exchange by introducing pre-play communication before the market game. We ask subjects to discuss whether to exchange harmful product to the third party is socially appropriate. Our treatments differ in whether communication is in the veil of ignorance, where participants discuss without knowing their roles, and whether third party participates the discussion. We find significantly higher rates of fair trade in Zurich than Shanghai when pre-play communication is not allowed, a consistent replication of the results in BWY(2015). Communication in the Veil of Ignorance significantly improve the fair trade in both countries, with 96 and 60 percent of fair product transactions in Zurich and Shanghai, respectively. The results of fair trade in all communication treatments do not differ much in Zurich, while in Shanghai excluding the third party in the discussion displayed the least fair trade of all treatments.
Chun-Lei Yang, Nanjing Audit University
Title: Voluntary Separation as a Disciplinary Device for Long-Term Cooperation: Reconciling Theory with Evidence
Abstract: In this paper, we provide an evidence-based theory to explain how partners cultivate a long-term relationship when either party has the liberty to unilaterally terminate the match. We utilize laboratory methods to gather insights on the evolution of cooperation in a voluntarily separable repeated prisoner's dilemma game (VSRPD). We observe behavioral patterns that are at odds with out-for-tat which, based on the VSRPD literature, is a disciplinary device that helps facilitate mutual cooperation in the long run. Our Pro-Partnership Proposition is formulated to accommodate the observation that players in our experiment often prefer to use a stay-but-act-like-a-stranger move to punish defecting partners. A new class of equilibria, called the CoDe-indifferent equilibria, is introduced to address the within-match rewards and punishments found in the data.
Binglin Gong, East China Normal UniversityTitle: The Sequence Effect on Hybrid Auction-Lottery Mechanisms
Abstract: Hybrid mechanisms combing auction and lottery are now widely used to allocate publicly provided private goods, such as private car license plate, to balance efficiency and equality. A natural question is how to implement them. We compare the performance of two sequence designs: “first auction, then lottery” and “first lottery, then auction”, in a unified framework from perspectives of revenue, efficiency, and equality.
Our theoretical results suggest that given the proportion of goods being auctioned, lottery first is always more efficient than auction first. Efficiency increases in the proportion of goods being auctioned. When common distributions of players’ values, such as uniform, normal, lognormal, exponential and Cauchy, are not heavily right skewed, lottery first generates more revenue than auction first.
In a laboratory experiment, results show that Lottery first can indeed yield significantly higher efficiency, especially when more goods are auctions. And the more goods are auctioned, the higher the overall efficiency is. However, contrary to our theoretical prediction, auction first results in more revenue, because subjects do not shade their bids as much as predicted in auction first game, possibly due to risk aversion, and shade a lot in lottery first game, possibly due to inequality aversion.
Gergely Horvath, Xi'an Jiaotong Liverpool UniversityTitle: Network Structure and Conspicuous Consumption: An Experiment
Abstract: A laboratory experiment has been designed to study the dynamics of social competition under different social network structures. In the experiment, we employ a so-called social competition game, i.e., each individual maximizes her utility by allocating her endowment between two goods, a private one and a competitive one. A competitive good means that an individual can get higher payoffs if she consumes “more than the average of her neighbors in a network”, for instance, a status good or national-defense good. By spending more on the competitive good, however, everybody decreases their utility/payoffs. In the welfare-maximizing allocation everybody restricts themselves and spends a low amount on the competitive good. In the experiment, subjects simultaneously make consumption decisions and receive feedbacks about the payoffs and the consumption decisions of subjects with whom they are directly connected in the network. We study this game in four networks that differ in connectedness and the symmetry of network positions.
We found in most cases the results converge to the Nash predictions about the competitive good consumption levels and payoffs. In none of the network structures can participants fully coordinate on the welfare-maximizing outcome. The consumption choices of individuals are determined by their network position: by a specific measure of centrality, the so-called Katz-Bonacich centrality. The only exception happens in the network with only one central individual who tries to lower the intensity of competition and hence consumes less of the competitive good than the Nash equilibrium consumption level. We also study the impact of overall network structure, and find that in a more connected network, individuals spend more on the competitive good and earn lower payoffs. Further, we find evidence that the convergence to Nash equilibrium can be described by myopic best-response dynamics.
June 15th, 2018
Session 3 Chair: Tingting Ding
Antonio Guarino, University College London
Title: Non-Bayesian Updating in Social Learning: An Experiment
Abstract: We present a novel experimental design to study social learning in the laboratory. Two subjects, in sequence, have to predict the value of a good on which they receive private information. We elicit the second subject’s belief twice: first (“first belief”), after he observes his predecessors’ action; second (“posterior belief”), after he observes his private signal. We are, therefore, able to disentangle social learning from learning from private information. Our main result is that subjects update on their private signal in an asymmetric way. They weigh the private signal as a Bayesian agent would do when the signal confirms their first belief; they overweight the signal when it contradicts their first belief. We show that this way of updating, incompatible with Bayesianism, can be explained by multiple priors on the predecessor’s rationality and a generalization of the Maximum Likelihood Updating rule. In another experiment, we directly test this theory and find further support for it.
Peiyao Shen, Shanghai Tech University
Title: Market Power and Transparency in a Multi-unit Auction
Abstract: We study the effects of different information structures (complete information, asymmetric information and symmetric but incomplete information) and different levels (monopoly, duopoly, competitive) and types (pivotal power and limit pricing) of market power on equilibrium prices, efficiency and bidding behavior in uniform multi-unit auctions, using a supply function equilibrium approach and a novel experimental design. The implemented auction design is motivated by expected changes in available information when high shares of renewables are integrated in electricity markets. The experimental data strongly suggests that asymmetric information (cost uncertainty) leads bidders to coordinate on a much more competitive equilibrium, implying that both pivotal market power and limit pricing are substantially reduced. We find no such effect in case of demand uncertainty.
Jianbiao Li, Nankai University
Title: Using Non-Invasive Brain Stimulation to Test the Effect of Self-Control on Investor Behavior
Abstract: We test whether self-control causally affects investor behavior. Using non-invasive brain stimulation, we exogenously vary self-control. Subjects trade stocks in an experimental asset market while they receive anodal, cathodal, or sham stimulation over a brain area related to self-control. Anodal and cathodal stimulations are known to increase or decrease the neural excitability of the targeted regions, respectively; whereas sham stimulation mimics the peripheral effects and does not affect any neural processing. We document that cathodal stimulation significantly reduces the disposition effect by helping subjects exhibit a sufficient self-control, which is reflected by a high percentage of actual-ideal consistent decision. These results provide support for the role of self-control in disposition effect. Generally, we demonstrate that non-invasive brain stimulation can be helpful in causally testing the underlying mechanism of investor behavior.
Xiaolan Yang, Shanghai International Studies University
Title: Social Distance and Job Referral: An Experimental Study
Abstract: Job referral through social network plays an important role in labor markets worldwide. In this study, we investigate how social distance affects job referral decisions in a lab experiment. We set up a job referral game which includes three players, one worker, one advisor and one employer. The ability of the worker is only revealed to the worker and the advisor. The advisor has to decide whether to recommend the worker to the employer or not. The employer receives a noisy signal about the worker’s ability and decides whether to employ the worker or not after he/she receives the recommendation. Subjects are recruited from three different university halls at one university. We have three treatments which differ in whether the three players are from the same university hall. In the control treatment, subjects have no information about where their partners come from. In the in-group (out-group) treatment, all three players know that the worker and the advisor are from the same (different) university hall. Results show that advisors in the in-group treatment have the highest recommendation rate, even when such decisions could only benefit the workers. Besides, compared with advisors in the control treatment, the advisors in both the in-group and the out-group treatments are more inclined to recommend. This result suggests that providing background information may narrow the social distance between the worker and the advisor, hence increase the referral. There is no significant difference in employer’s behavior across three treatments.
Danyang Li, Hofstra University
Title: Liar Liar: Experimental Evidence of the Effect of Confirmation-reports on Dishonesty
Abstract: We identify the effect of confirmationreports on dishonesty using ‐data from an experiment where subjects are asked to roll a die and report its outcome using either a self-report or confirmation-report mechanism. We find that relative to self-reports, confirmation-reports have a positive effect on the share of subjects who report honestly. The effect on the magnitude of lies told depends greatly on the accuracy of the prefilled information on the confirmation-report. We argue that these results are driven by changes in the intrinsic costs of lying induced by the confirmation report.
Session 4 Chair: Zhengxing Huang
Zhiqiang Dong, South China Normal UniversityTitle: Impact of Being Left-behind on Competitive Behavior of Children in Rural China: A Field Experiment
Abstract: Since the progress of China's economic system reform and urbanization in last decades, millions of young parents migrated from countryside to cities, which induced lots of left-behind children. The absence of parents' upbringing caused different kinds of mental and physical problems of left-behind children. The competition plays a vital role in children's development and socialization by encouraging children's engagement, mastery of a task and a desire to achieve their best. We firstly explored the effect of left-behind experience on children's competitive behaviors through a field experiment in a rural area of central China. The results show that left-behind children are significantly reluctant to choose competition with others compared with non-left-behind children, even though their performances under competition have no difference. Father plays a more important role in the development of children's competition preference. Our study implicates that more encouragements should be provided to left-behind children to build their confidence and cultivate their good competition concepts.
Lingfang Li, Fudan University
Title: The Displaying Effect of Price History on Consumers’ Purchase Decisions
Abstract: As visualization has become increasingly common today, consumers’ perception is likely to be influenced by the displayed reality, rather than the reality. In this research, we explore the effects of the visualization of price history on consumer purchase decisions through a series of experiments. In addition to study the factors of price history, including last price changes, price change frequency and time window, we especially focus on the factors of displaying, including visual turbulence and length of price charts. Our findings suggest that (a) the factors of displaying affect consumers’ purchase decisions; (b) different factors of displaying may mitigate each other’s effect on consumers’ purchase decisions; (c) the factors of displaying and the factors of price history may offset each other’s effect on consumers’ purchase decisions.
Zhi Li, Xiamen University
Title: Generalized Serial Cost Sharing Mechanisms for the Provision of Non-excludable Threshold Public Goods: An Experimental Investigation
Abstract: Following Gailmard and Palfrey (J Pub E, 2005, GP thereafter), we introduce two uniform price mechanisms, the uniform price auction (UPA) mechanism to extend the direct serial cost sharing to non-excludable public goods, and the uniform price cap (UPC) mechanism to solve the inefficiency of serial cost sharing by a variation of residual serial cost sharing. Both cost sharing mechanisms are budget balanced, individually rational, anonymous, but are not strategy proof. We compare these two cost sharing rules with voluntary cost sharing with proportional rebates (PR) and with no rebates (NR, or PPM) as in GP for the provision of non-excludable public goods under incomplete information. We characterize the Bayesian Nash equilibria (BNE) of the two new mechanisms and conduct laboratory experiments to compare the performance of the four mechanisms. We find that UPA induces significantly higher value revelations than all the other three and UPC generates significantly more efficient allocations than all the others. In both uniform price mechanisms, that an individual's payment is not affected by the declarations of valuation higher than their own contributes to the higher value revelation and more efficient allocations by reducing the equilibrium size and facilitating coordination toward more efficient equilibria. Our results are robust to group size.
Yun Wang, Xiamen University
Title: Rebate Policies in Large Group Threshold Public Goods Experiment: Belief, Information, and Repetition
Abstract: This paper explores multi-round threshold public goods provision in large groups. We elicit subjects' beliefs regarding others' contributions and the project's cost; and we vary the amount of information between Round-0 and subsequent rounds to examine how the demand revelation changes with information, repetition, and cost. Our Round-0 decision-task replicates the large-group one-shot game of Rondeau et al. (1999) and Spencer et al. (2009), in which the project cost and group members' value distribution are unknown to subjects. In addition to the proportional rebate and three winner-take-all policies (Spencer et al., 2009), we introduce and compare two novel rebate rules: the uniform price cap (UPC) and the uniform price auction (UPA) mechanisms.
Our Round-0 result shows that, in the one-shot unknown-cost setup, all rebate policies induce group contributions higher than the actual cost. Nonetheless, only UPC and UPA achieve almost 100% demand revelation, while the demand revelation under all other rebate policies is significantly below 100%. This result is different from Rondeau et al. (1999) and Spencer et al. (2009), which suggest 100% or more demand revelation for most rebate policies under the one-shot large-group design. Interestingly, subjects' perceived cost is not significantly different from their induced values for all mechanisms but two winter-take-all policies. Subjects contribute significantly less than their perceived cost except for UPC and UPA treatments. Furthermore, in the subsequent rounds when the cost information is revealed and decisions are repeated, full demand revelation disappears. Group contributions increase with cost; yet the rates for successful provision decrease.
Yang Yang, Sun Yat-sen University
Title: On Language and Meaning: A Randomized Experiment
Abstract: We develop a method for random assignment of language to participants in a laboratory experiment. We use this approach to test the linguistic relativity hypothesis (also referred to as the Sapir-Whorf hypothesis). Linguistic relativity suggests that the structure of one’s language can influence one’s perceptions, interpretations and beliefs about the world around them. Although provocative, empirical evidence on this hypothesis has been elusive. A reason is that previous empirical studies typically rely on naturally occurring languages whose speakers differ in ways that correlate with grammar differences. Here we hypothesize that linguistic relativity can emerge when the same object resides in different semantic categories across different languages. To test this, we develop a novel extension of laboratory games within which grammars emerge endogenously. We show, first, that one can control the semantic categories of an emergent grammar by varying the game’s incentives. This finding enables random assignment of language to participants. Advantaged by this randomization, our data support the hypothesis that the same object’s meaning can vary according to its semantic category. Our methodological and substantive insights promise to be important in improving communication, cooperation and understanding across human societies.
Concluding Remarks: Lan Yao